Journal of Personal Finance Tools, Techniques, Strategies, and Research to Aid Consumers and

Journal of Personal Finance

Tools, Techniques, Strategies, and Researchto Aid

Consumers and

Professional Financial Advisors

Volume 5, Issue 2

The Official Journal of the International Association ofRegistered Financial Consultants

©2006, IARFC All rights of reproduction in any form reserved.

Journal of Personal Finance2

CONTENTS

EDITOR’S NOTES .................................................................. 9John E. Grable, Ph.D., CFP®, RFC

NEWS & NOTES .................................................................. 14

PRACTICE MANAGEMENT

A Further Examination Of The Validity Of The Kansas MaritalSatisfaction Scale: Implications For Financial Consultants .............. 17John E. Grable, Ph.D., CFP®, RFCKansas State UniversitySonya Britt, M.S.Kansas State UniversityThe purpose of this study, using data from a survey of Midwestern marriedrespondents (N = 293), is to yield additional empirical evidence regarding thecriterion- and construct-related validity of the Kansas Marital SatisfactionScale. The KMS was significantly correlated with the Relationship Assess-ment Scale and with Type-A Personality, Self-Esteem, and Religiosity in thisstudy. These findings confirm the criterion-related validity of the KMS. Alogistic regression was used to test the construct validity of the KMS.Results support the construct-related validity of the KMS. Lower KMSscores were associated with an increased likelihood of thinking a marriagemight be in trouble. Age and employment status also were found to benegatively related to the thought of marital trouble. Since money managementis a large aspect of most marriages, the findings from this study support theuse of the KMS by financial consultants who believe that their clients may beexhibiting problematic financial and marital behaviors. Implications forpractitioners are provided to elaborate on this finding.

Volume 5, Issue 2 3

RESEARCH & THEORY

Mortgages And Asset Accumulation: A Ten YearComparison Of Homeowners .............................................................. 32Lance Palmer, Ph.D.The University of GeorgiaJean M. Lown, Ph.D.Utah State UniversityThe purpose of this study was to investigate the long-term relation betweenhousehold leverage through the use of mortgages, and changes in householdwealth using the theoretical framework of the life cycle income hypothesis.This study used the 1992 through 2002 waves of the Health and RetirementStudy. The characteristics of leveraged and unleveraged households werecompared in 1992 and 2002 as were changes in wealth during that period.Households without mortgages consistently had greater net assets relative tohouseholds with mortgages. Whether members of the household werecurrently working was consistently associated with carrying a mortgage.

The Effects Of Investment Education On Gender DifferencesIn Financial Knowledge ....................................................................... 55Ronald E. GoldsmithFlorida State UniversityElizabeth B. GoldsmithFlorida State UniversityThere is much concern that women in particular have less knowledge and skillin managing their personal finances than men do, despite its manifest impor-tance and the fact that so many women have this responsibility. Thesegender differences appear early in life: previous studies have suggested thatmale college students not only feel that they have more knowledge ofpersonal finance than female students, but that in fact they really do havemore knowledge on average. One avenue for improving financial skills isthrough education. Little research, however, has been done to evaluate theeffects of education on the financial skills of women versus men. Using datafrom 122 students, the present study confirmed that women knew less aboutinvestments than did men. However, the experiment showed that formalinstruction in personal finance not only increased both men’s and women’sinvestment knowledge, it closed the knowledge gap between the genders.



Consensus Market Forecasts: Too Risky To Rely On Experts .......... 70Seth Hammer, CPA, Ph.D.Towson UniversityOra Freedman, Ph.D.Villa Julie CollegeThis study examines whether it is feasible that consensus market forecasts,such as those published annually by Business Week can provide informationuseful for asset allocation decisions. The findings suggest that reliance onforecasts of the Business Week experts or any group of experts is highly riskybecause a) experts, similarly to laymen, demonstrate a proclivity to engage inthe anchoring and adjustment heuristic; b) experts must perform at a levelapproximately 20% above random simply to match a buy-and-hold strategy;and c) where taxes are a factor, experts must perform at an even higher level tocompensate for loss of compounding benefits and loss of preferred tax ratesapplicable to stock investments.

Covered Call Writing As A Strategy For Small Investors ................... 87Ki C. Han, Ph.D.Suffolk UniversityKamlesh DadlaniErnst & Young LLPThis study investigated, using Dow Jones Industrial Average (DJIA) compa-nies, whether writing a covered call provides a significant benefit overoutright stock investment. It was found that even a simple strategy usingout-of-the-money calls brought significant benefits to investors in terms ofrisk and reward. The results demonstrated that, during the period of 1990-2003, covered call writing reduced the volatility of returns significantly, butdid not significantly lower returns. This study also investigated two othercovered call strategies. Writing a covered call with half of the position in out-of-the-money calls and the other half in in-the-money calls worked as well aswriting covered calls with out-of-the-money calls only. In fact, the formerworked even better than the latter during the less bullish period. However,writing covered calls with in-the-money calls only did not work well.

Summary Of Author And Article Citations: Journal OfPersonal Finance Volumes 1, 2, 3, & 4 .............................................. 101Theresa KasperEditorial AssistantJohn E. Grable, Ph.D., CFP®, RFCEditor

Volume 5, Issue 2 5

Traditional Versus Roth 401(K): A Benefits Analysis ....................... 112G. Eddy Birrer, PhD, CPAGonzaga University

BOOK REVIEW

Blink: The Power of Thinking without Thinking ............................. 121Author: Malcom GladwellReviewer: Tim Griesdorn, MBA, CFM, CMA

INSTITUTIONAL PROFILE

Financial Planning At Western Carolina University ........................ 125

Journal of Personal FinanceGuidelines for Authors ....................................................................... 132



Write an article today! The Journal of Personal Finance is currentlyaccepting manuscripts and reviews for publication in future issues.

Practitioners, this is your opportunity to contribute to the professionby sharing your ideas and insights with others. Academicians, this isyour opportunity to add tothe body of literature in personal financialplanning through a rigorous peer reviewed process.

The Journal of Personal Finance is unique in its publication approach. TheJournal’s Research Policy Board is committed to publishing timely originalcontributions that offer readers applicable personal financial planning tools,techniques, and strategies.

The Journal is practitioner oriented. Approximately one-half of each issue isdevoted to practice management articles written by financial consultants andacademicians. Each issue also includes empirically based academic articles.Both qualitative and quantitative articles are acceptable. A blind peer reviewprocess is used to evaluate each manuscript. Contributors are encouraged tosubmit papers corresponding to the following topic areas:

� Client Relationship Management� Financial Planning Trends� Technology Issues� Planning for Special Needs� Regulation Overview� Ethics of Financial Planning� Practice Management Techniques� Interesting and Unique Planning Tools and Techniques� Investment Decision Management� Marketing Methods� Book Reviews and Letters� Attitude and Behavioral Measurement

The audience for the Journal consists primarily of practicing financialplanners, insurance advisors, other securities industry professionals,consultants, and academicians. Empirically based submissions shouldprovide a detailed discussion of findings that are directly related topractitioner implementation.

If you are a new author, or just thinking about the idea of writing for the firsttime, please feel free to contact Ruth H. Lytton, Ph.D. about your manuscriptideas. Details regarding the Journal’s manuscript submission process can befound at www.ksu.edu/ipfp/jpf.htm

Volume 5, Issue 2 7

JOURNAL OF PERSONAL FINANCE

VOLUME FIVE, ISSUE TWO

Steve Bailey, RFC, LUTCF, CEBAHB Financial ResourcesJoyce CantrellKansas State UniversityCharles Hatcher, Ph.D.University of WisconsinSo-Hyun Joo, Ph.D.Texas Tech UniversityRich Landsberg, RFC, J.D., LLMNationwide Financial ServicesMike MassrockWestern Reserve Life and ISIRuth H. Lytton, Ph.D.Virginia TechMike Lemieux, RFCU.S. Worldwide Financial Services, Inc.Jean Lown, Ph.D.Utah State UniversityConstance O. Luttrell, RFCCC & AssociatesBurnett Marus, RFCBurnett Marus AssociatesJerry Mason, Ph.D., CFP�, RFCEdelman Financial Services

Jim McCarty, RFCAmerican Express FinancialAdvisorsEdwin P. Morrow, RFC, CFP�, CLU,ChFCFinancial Planning ConsultantsBarbara O’Neill, Ph.D., CFP�, AFC,CHC, CFCSRutgers Cooperative ExtensionRobert Moreschi, Ph.D., RFCVirginia Military InstituteAnthony Sorrentino, RFCNational Financial ServicesRichard Salmen, CFP�, CRFA, EAG-TrustRonald Stair, Ph.D., RFC, MLT, AEPCreative Plan DesignsJing Xiao, Ph.D.University of Rhode IslandWilliam J. Trainor, Jr., Ph.D.Western Kentucky University

Mailing Address: Institute of Personal Financial Planning,School of Family Studies and Human Services, 318 Justin Hall,Kansas State University, Manhattan, KS 66506Phone: (785) 532-1486 or (785) 532-5510; Fax: (785) 532-5505;E-mail: [email protected] Site: www.ksu.edu/ipfp/jpf.htm

EDITOR

John E. Grable, Ph.D., CFP®, RFCEDITORIAL COORDINATOR

Jennifer King

EDITORIAL ASSISTANT

Theresa KasperEDITORIAL ADVISORY BOARD

© © © © © Copyright 2006. International Association of Registered FinancialConsultants. (ISSN 1540-6717)



Postmaster: Send address changes to Editorial Assistant, Journal of PersonalFinance, 303 Justin Hall, Family Studies and Human Services, Kansas StateUniversity, Manhattan, KS 66506.

Permissions: Requests for permission to make copies or to obtain copyrightpermissions should be directed to the Editor.

Certification Inquiries: Inquiries about or requests for information pertainingto the Registered Financial Consultant or Registered Financial Associatecertifications should be made to IARFC, Financial Planning Building, 2507North Verity Parkway, Middletown, Ohio 45042.

Disclaimer: The Journal of Personal Finance is intended to present timely,accurate, and authoritative information. The editorial staff of the Journal isnot engaged in providing investment, legal, accounting, financial, retirement,or other financial planning advice or service. Before implementing anyrecommendation presented in this Journal readers are encouraged to consultwith a competent professional. While the information, data analysis method-ology, and author recommendations have been reviewed through a peerevaluation process, some material presented in the Journal may be affected bychanges in tax laws, court findings, or future interpretations of rules andregulations. As such, the accuracy and completeness of information, data,and opinions provided in the Journal are in no way guaranteed. The Editor,Editorial Advisory Board, the Institute of Personal Financial Planning, and theBoard of the International Association of Registered Financial Consultantsspecifically disclaim any personal, joint, or corporate (profit or nonprofit)liability for loss or risk incurred as a consequence of the content of theJournal.

General Editorial Policy: It is the editorial policy of this Journal to onlypublish content that is original, exclusive, and not previously copyrighted.

Subscription Rates:Individual $ 55 U.S. Institution: $ 98 U.S.

$ 68 Non-U.S. $115 Non-U.S.

Single issue $19 U.S.

Send subscription requests and payment to:IARFCJournal of Personal FinanceThe Financial Planning Building2507 N Verity ParkwayMiddletown, OH 45042

Volume 5, Issue 2 9

Editor’s N

otesEDITOR’S NOTES

The year 2006 is well on its way, and what a great year it has been. Ihave been receiving wonderful feedback from readers about the papers thatwere published in Volume 4 Issue 4. The dual topics of predicting the futureof the profession and why we all should be writing really hit home withreaders. I only hope that we all can continue the dialogs that were started inthat issue into the future.

My guess is that after reading through this issue of the Journal I willbe receiving even more feedback. We have an all-star lineup of authors in thisIssue. The topics covered are diverse, but the underlying theme is the same,namely, information, tools, techniques, and strategies that can be used toimprove the practice of personal financial planning. If you are interested inwho is the most widely quoted author over the past five years of the Journal,I encourage you to look at the summary of author and article citations paper.The quality of references is an indication of the excellent research that isbeing conducted in the area of personal finance.

On a personal note I want to share some (and somewhat sad for me)news. This issue of the Journal of Personal Finance will be my last as editor.When Ed Morrow and I concluded that this Journal needed to be establishedI eagerly volunteered to get it off the ground, but I always had one stipula-tion, namely, I would stay on only long enough for the Journal to achievestability. That time has now come. The Journal has grown from one issue inlate 2001 to four issues yearly in 2006. The number of papers being submittedhas grown dramatically over that time, and our acceptance rate has continu-ally dropped (this is a good thing in the academic world). The Journal ofPersonal Finance can now be found in some of the nation’s largest and mostrespected libraries, on the Internet, and through indexing services includingProQuest and Cabal’s. It has been a pleasure to watch the Journal growduring my tenure as editor, but it is now time to allow the Journal to groweven larger under the leadership of someone who will bring a fresh perspec-tive and a dedication to sustained quality.

I am very pleased to announce that Dr. Ruth Lytton will be takingover editorial responsibilities starting with the Issue Three of Volume Five. Iam personally excited to know that the Journal will be in the hands of Ruth.She is, by far, one of the nation’s leading thinkers on personal finance topics.Students at Virginia Tech, under her guidance, have won national collegiatefinancial planning competitions, and she herself, has won several national andinternational research and teaching awards. So, while I am saddened that Iwill no longer have a day-to-day role in the Journal’s operations, I am thrilledto know that the Journal will continue to improve into the future. Please feelfree to contact Ruth at: [email protected].



10

Meet the Authors

The first paper in this Issue of the Journal of Personal Financedeals with a topic that, at first glance, may not appear to be related topersonal finance. Sonya Britt and I (John Grable) looked at how effectivelythe Kansas Marital Scale works in predicting relationship satisfaction. Theissue of marital stability is one that is growing in importance for practitioners,and it was concluded that the Kansas Marital Scale may be one tool thatfinancial consultants can use to assess their clients’ marital wellness. SonyaBritt, M.S. received her undergraduate degree in Personal Financial Planningand a master’s degree in Marriage and Family Therapy from Kansas StateUniversity. She is Licensed Marriage and Family Therapist who works as atherapist at a low-income agency and also at an independent agency. Inaddition to practicing marriage and family therapy, Sonya teaches anintroduction to personal and family finance course at Kansas StateUniversity. Dr. John Grable, CFP®, RFC holds the Vera Mowery McAninchProfessor of Human Development and Family Studies professorship atKansas State University. He received his undergraduate degree in economicsand business from the University of Nevada, an MBA from ClarksonUniversity, and a Ph.D. from Virginia Tech. He is the Certified FinancialPlanner™ Board of Standards Inc. and International Association ofRegistered Financial Consultants registered undergraduate and graduateprogram director at Kansas State University. Teams of undergraduatefinancial planning students mentored by Dr. Grable have won the NationalCollegiate Financial Planning Championship in 2000, 2003, 2005, and 2006. Dr.Grable also serves as the director of The Institute of Personal FinancialPlanning at K-State. Prior to entering the academic profession he worked asa pension/benefits administrator and later as a Registered InvestmentAdvisor in an asset management firm. Dr. Grable serves as the editor for theJournal of Personal Finance, a rigorous peer-reviewed research journal. Hisresearch interests include financial risk-tolerance assessment, financialplanning help-seeking behavior, and financial wellness assessment. He hasbeen the recipient of several research and publication awards and grants, andis active in promoting the link between research and financial planningpractice where he has published more than 60 refereed papers. Dr. Grableserved on the Board of Directors of the International Association ofRegistered Financial Consultants, as Treasure for the American Council onConsumer Interests (ACCI), and on the Research Advisory Council of theTake Charge America Institute (TCAI) for Consumer Education and Researchat University of Arizona. In 2004 he won the prestigious Cato Award forDistinguished Journalism in the Field of Financial Services.

Drs. Lance Palmer and Jean Lown have written an interesting paperthat compares mortgages and asset accumulation over a 10-year period

Volume 5, Issue 2 11

Editor’s N

otes

among homeowners. The paper’s lead author, Lance Palmer, Ph.D., is anAssistant Professor at the University of Georgia where he teachesintermediate family financial management, family retirement planning andemployee benefits, and family taxation courses. Dr. Palmer also serves as theCertified Financial Planner Board of Standards, Inc. Program Director at theUniversity of Georgia. His coauthor, Jean M. Lown, Ph.D., teachesundergraduate courses in the Family Finance Emphasis in the Family,Consumer, and Human Development major and graduate courses in familyresource management at Utah State University. Dr. Lown’s research interestsinclude women and financial planning attitudes and practices, especially inrelation to retirement planning. Her research asks how to motivate women totake more responsibility, prior to mid-life, for their financial security in laterlife. A second interest is credit overuse and consumer bankruptcy in Utahdue to the high filing rate. Specifically, why is the failure rate for Chapter 13repayment plans so high and how do successful Chapter 13 plans differ fromunsuccessful plans? To what extent does local legal culture contribute to theproblem of multiple filings and unsuccessful repayment plans?

The third paper in this Issue looks at the effects of education ongender differences in financial knowledge. The paper was written by adynamic couple, Drs. Ronald and Elizabeth Goldsmith. Ronald E. Goldsmith,Ph.D., is a marketing professor at Florida State University. His researchinterests include measurement and data quality issues, especially in the areasof diffusion on innovations and consumer involvement. Elizabeth B.Goldsmith, Ph.D., is a professor in the College of Human Sciences at FloridaState University. Her research interests focus on work and family, women andmoney, the functioning of the American home, and environmental issues.During a 1992 sabbatical, she conducted research at the White House andthe Smithsonian Institution. Since then, she has had grants from theSmithsonian, Duke University, the Hoover Presidential Library, and been aresearcher, policy advisor, and guest at the White House. During a 1999sabbatical, she conducted research in Ireland and at the JFK PresidentialLibrary in Boston. Dr. Goldsmith has published encyclopedia chapters,journal articles, two college textbooks entitled Resource Management forIndividuals and Families, 2nd edition, and Personal Finance both byWadsworth, and a monograph for the United Nations. She serves on theeditorial board of the Journal of Family and Economic Issues and on theUnited States Trustee/Consumer Educator Working Group on DebtorEducation in the Department of Justice. She served as Associate Editor forthe Journal of Family and Consumer Sciences and has given speeches inAustralia, Germany, Great Britain, Malta, Mexico, Northern Ireland, Sweden,and Finland.

The next paper looks at how well experts are able to consistentlyand accurately make market forecasts. The paper’s lead author is Dr. Seth



12

Hammer. Dr. Hammer is an associate professor of accounting at TowsonUniversity in Towson, Maryland. He is the author of CCH’s Investment TaxPlanning Guide (2004) and has written articles for the Journal ofRetirement Planning, The Tax Adviser, and other journals. He is active in theMaryland Association of Certified Public Accountants. His professionalactivities have included providing financial education programs for membersof the NFL’s Baltimore Ravens since 1996. Since assuming his position atTowson University in 1994, Dr. Hammer has made a number of intellectualcontributions in both applied and basic research. His success in integratingskills honed from his experiences in academia and professional accountancyhas contributed towards his becoming a leader in the field of tax education. Dr. Hammer’s significant contributions in applied research include articles inthe CPA Journal and the prestigious Tax Advisor. Dr. Hammer has alsoserved as a principal co-speaker at a Foundation for Accounting Educationseminar, sponsored by the New York State Society of Certified PublicAccountants. Dr. Hammer’s contributions in basic research includepresentations at an American Accounting Association’s annual meeting, aDecision Science Institute’s northeast regional meeting, and WashingtonAccounting Research Society conferences. He is currently developingarticles for journal submission, based on his tax research experiments. Hiscoauthor is Dr. Ora Freedman. Dr. Freedman is a professor of economics atVilla Julie College in Stevenson, Maryland. Prior to joining the faculty at VillaJulie she taught managerial economics at SUNY Binghamton in both theirtraditional and executive M.B.A. programs. Dr. Freedman has also taught inthe Technion, Israel. Her research interests are primarily in the area ofeconomics education.

IARFC members who actively manage client’s investment portfolioswill find the paper coauthored by Ki Han, Ph.D. and Kamlesh Dadlani veryinformative. The paper deals with ways covered call writing can be used bysmall investors and their advisors. Ki C. Han, Ph.D. is a professor of financeand the chairperson at Suffolk University in Boston. His areas of expertiseinclude firm ownership structures, stock distributions, international finance,value-driven management, and capital budgeting. He serves as a journalreferee for the Journal of Business Finance and Accounting, FinancialReview, and the International Journal of Finance. Dr. Han’s coauthor is aKamlesh Dadlani, MBA who is a Tax Consultant with Ernst & Young inBoston, Massachusetts.

The next paper in this Issue is a summary of author and articlecitations in the Journal of Personal Finance for Volumes 1, 2, 3, and 4. Thisis the second paper on the topic written by Theresa Kasper who is theEditorial Assistant for the Journal. This issue of the Journal also includes areview of the book titled “Blink: The Power of Thinking without Thinking.”


Editor’s N

otes

The review was written by Tim Griesdorn, MBA, who is currently pursuing aPh.D. at Texas Tech University.

The final paper was written by G. Eddy Birrer, Ph.D., CPA. Dr. Birrerhas served as an accounting professor at Gonzaga University since 1984,with a teaching emphasis on financial accounting and reporting. Hisscholarly work includes publications in periodicals ranging from The Journalof Accountancy to Personal Financial Planning and a co-authored book onfinancial planning and analysis. His research specialty is applications ofpresent value analysis to financial reporting and financial planning. TheIssue concludes with an Institutional Profile of Western Carolina University’soutstanding undergraduate degree program in financial planning. Readerswho are interested in obtaining more information about WCU’s program needto contact Dr. Grace Allen.

In summary, I want to thank everyone who has submitted a paperover the past five years. This ‘little’ journal emerged from a once-per-yearpublication to a quarterly in a short period of time. To me, this is proof thatpersonal finance, as an academic discipline, is in the right place at the righttime. I wish you all the best.

John E. Grable, Ph.D., CFP®, RFCEditor



NEWS AND NOTES

Information and Research of Interest to IARFC Members

Industry Bulletin

The 2006 National Collegiate Financial Planning Championship,sponsored by Amerpise Financial, Inc., was held between April 19th and 22nd inMinneapolis. Universities from across than nation competed for eight spotsat the finals by writing a comprehensive financial plan using Word and Excelfor a hypothetical client family. This year’s finalists included Kansas StateUniversity, Fort Hays State University, Minnesota State University, TexasTech University, Virginia Tech, University of Hawaii, Slippery Rock University,and Wright State University. Kansas State University won the best writtenplan segment of the competition, while Texas Tech University won the “HowDo You Know Challenge” and Fort Hays State University won the bestpresentation award. The winner of the overall competition was Kansas StateUniversity. This is the 4th national title victory in the past seven year for K-State.

The fall 2006 and spring 2007 conference season is almost here. Thefollowing is a list of conference readers of the Journal might be interested inattending:

� NAPFA West: September 14 – 17; Portlando This is small gathering includes presentations by leaders within

the Fee-Only financial planning movement. Few presentations arebased on empirical research.

� Academy of Financial Services Professionals: October 11 – 12; SaltLake Cityo This small group meets annually to discuss topics of interest to

financial planners. The membership is comprised primarily ofacademicians from small business schools and programs that offerfinancial planning degrees.

� Financial Planning Association: October 21 – 24; Nashvilleo This conference bills itself as the largest financial planning

conference in the country. Speakers tend to be well knownpractitioners; few discussions are based on academic research.

� NAPFA Northeast: October 26 – 28; Bethesda, Marylando This is small gathering includes presentations by leaders within


Volume 5, Issue 2

New

s & N

otes

15

� NAPFA South: November 2 – 4; San Antonioo This is small gathering includes presentations by leaders within


� Association for Financial Counseling and Planning Education(AFCPE): November 15 – 17; San Antonioo This growing organization’s conference is becoming the centerpiece

event for those interested in financial counseling topics andfinancial planning from a highly developed academic perspective.Attendees will not find high pressure sales strategies at thisconference. Instead, expect hands-on techniques that can be usedwith middle-America clients.

� American Council on Consumer Interests (ACCI): May 18 – 21; St.Louiso This small gathering includes the most rigorous research as it

pertains to consumer topics. Financial planning and personalfinance as a consumer interest play a major role in the conference.For those interested in hearing about cutting-edge research beforeit reaches a journal, this conference is a must.

� NAPFA National: May 2 – 6; St. Petersburgo This is the largest gathering of Fee-Only planners in the nation.

Presentations include a mix of practice management topics as wellas a few research based discussions.

� International Association of Registered Financial Consultants(IARFC): May 2007; Las Vegaso This conference, although small in numbers, is one of the absolute

best in terms of information, strategies, and techniques delivered.There are few conferences where a practitioner can sit down andtalk with speakers to discussion strategies. Las Vegas is a newvenue for the conference, so expect a large turnout.

Research Notes

Readers who are interested in the assessment of financial risk tolerance willfind the following recent article of importance:

� Ardehali, P. H., Paradi, J. C., & Asmild, M. (2005). Assessingfinancial risk tolerance of portfolio investors using dataenvelopment analysis. International Journal of InformationTechnology and Decision Making, 4, 491-519.



Do you have an interesting news story, or have you read an interestingresearch paper, that you feel would be useful to International Association ofRegistered Financial Consultant members? If yes, send a note to the editor [email protected].

Abstract: For some investors their own personal investment counsellors(sic) address their investment strategy; for others automated meansare used. To protect investors, the Canadian Government has enactedthe “Know Your Client” Act requiring that all investment dealers andvendors of securities must know their clients and advise them on theappropriate investment strategy. This paper uses Data EnvelopmentAnalysis (DEA) in a novel manner by applying it to a large data set ofanswers to a number of psychological questions. A Slacks Based Modelwas used to estimate investor risk tolerance. The model analyses therisk profile of the investor and can be used as a guide to match the riskrating of the investment vehicles for the client. Statistical comparisonswere also carried out to show how risk tolerance relates to variousdemographic variables. Finally, the DEA results were validated throughcomparisons with the commercial system already in use.

According to Ardehali et al., “DEA is a non-parametric LP optimizationtechnique, which is often used for measuring efficiency in servicesproduction units. What makes it appealing for risk tolerance assessmentis its ability to handle multiple variables at the same time. Just like whenit measures efficiency based on several inputs and outputs, it canmeasure risk tolerance based on several psychological questions, eachof them representing one dimension of risk tolerance. This is the firsttime DEA has been used for evaluating a psychologically orienteddataset, and the results have been quite encouraging.”


Practice M

anagement

A FURTHER EXAMINATION OF THE VALIDITY OF

THE KANSAS MARITAL SATISFACTION SCALE:IMPLICATIONS FOR FINANCIAL CONSULTANTS

John E. Grable, Ph.D., CFP®, RFCKansas State University

Sonya Britt, M.S.Kansas State University

ABSTRACT

The purpose of this study, using data from a survey of Midwesternmarried respondents (N = 293), is to yield additional empirical evidenceregarding the criterion- and construct-related validity of the KansasMarital Satisfaction Scale. The KMS was significantly correlated withthe Relationship Assessment Scale and with Type-A Personality, Self-Esteem, and Religiosity in this study. These findings confirm thecriterion-related validity of the KMS. A logistic regression was used totest the construct validity of the KMS. Results support the construct-related validity of the KMS. Lower KMS scores were associated with anincreased likelihood of thinking a marriage might be in trouble. Age andemployment status also were found to be negatively related to thethought of marital trouble. Since money management is a large aspect ofmost marriages, the findings from this study support the use of theKMS by financial consultants who believe that their clients may beexhibiting problematic financial and marital behaviors. Implications forpractitioners are provided to elaborate on this finding.

Introduction

The relationship between financial behaviors exhibited by individu-als and couples and relationship satisfaction is a topic of growing interest tomarriage and family therapists and financial consultants. Research suggeststhat the reasons for seeking marital therapy and financial planning are similar.It has been estimated that approximately one-third of all couples who seek thehelp of a financial consultant report having marital issues that cause conflict



in the relationship (Aniol & Snyder, 1997). In fact, money behaviors are oneof the primary causes for divorce (Amato & Rogers, 1997; Berry & Williams,1987; Terling-Watt, 2001). Relationship dissatisfaction can lead to behavioralactions that are counter productive to a couple’s financial situation. Unfortu-nately, it is not always easy for financial consultants to know whether aclient’s behaviors are being influenced by marital dissatisfaction. Thepurpose of this paper is to discuss the merits of using a short marital satisfac-tion scale as a tool to learn more about one’s married clients.

Marital Satisfaction

Burgess and Cottrell (1939) were among the earliest proponents whosought a better understanding of the factors that influence marriage satisfac-tion. Their early work on marital relationships set the stage for later researchdesigned to measure marital satisfaction. In an effort to meet the need for avalid and reliable measure of marital satisfaction, researchers developed a briefscale of marital satisfaction in 1977 – the Kansas Marital Satisfaction Scale(KMS). The concurrent and discriminant validity of the KMS was laterdescribed in the Journal of Marriage and the Family. Since the late 1970s thescale has been widely used by both researchers and practitioners who areinterested in assessing three distinct factors of marital quality: (a) satisfactionwith a persons’ marriage as an institution; (b) satisfaction with the relation-ship (i.e., intimacy and quality of communication); and (c) satisfaction withhusband or wife as a spouse (Mitchell, Newell, & Schumm, 1983). Financialmanagement and decision making is one aspect of marriage, so it is nosurprise that one’s relationship satisfaction and financial satisfaction arehighly related (Amato & Previti, 2003; Kerkmann, Lee, Lown, & Allgood,2000). The KMS is unique because it uses only three items to assess thesethree dimensions (i.e., satisfaction with a persons’ marriage as an institution;satisfaction with the relationship; and satisfaction with husband or wife as aspouse) of marital quality. Over the years several tests have been conductedto benchmark the scale’s reliability and validity. In general, the reliability ofthe scale, using Cronbach’s alpha, is high, often exceeding .90 and rarelyfalling below .75 (e.g., Calahan, 1997; Eggeman, Moxley, & Schumm, 1985;Grover, Paff-Bergen, Russell, & Schumm, 1984; Hatch, James, & Schumm,1986; Schumm et al., 1986).

The validity of the KMS has been widely reported in the literature.For example, the KMS appears to be highly correlated with the DyadicAdjustment Scale (Kurdek, 1992), the Locke-Wallace Marital Adjustment Test(White, Stahmann, & Furrow, 1994), and the Quality Marriage Index (Karney &Bradbury, 1997). In addition to a baseline level of concurrent validity, thecriterion validity of the KMS has also been established. In general, distressed


Practice M

anagement

spouses tend to have lower scores than well adjusted spouses (Moxley,Eggeman, & Schumm, 1986; Shek, Lam, Tsoi, & Lam, 1993). Construct validityhas also been measured. The KMS tends to be associated with religiosity(Schumm, Bollman, & Jurich, 1982; Mitchell et al., 1983), locus of control(Bugaighis, Schumm, Bollman, & Jurich, 1983), household income (Grover etal., 1984), life satisfaction (Morris & Blanton, 1994), and emotional intimacy(Hatch et al., 1986) among other factors. Although all of these items areimportant components of marital satisfaction, household income and lifesatisfaction are likely the most prevalent for financial consultants.

The joint committee of the American Psychological Association(APA), American Educational Research Association (AERA), and NationalCouncil on Measurement in Education (NCME) recommends that the accuracyof any scale be measured by assessing content-related, criterion-related, andconstruct-related validity (Ary, Jacobs, & Razavieh, 1990). The content-related validity of the KMS is already well established, as evidenced by themethodology used in the scale development process and the scale’s contin-ued wide use.

The quest to both quickly and accurately assess marital satisfactionis an activity that has been pursued since the 1930s (Burgess & Cottrell,1939). The development of the KMS in the 1970s furthered this quest. TheKMS has a long history of use as an effective measurement of satisfaction,but little recent research has been published regarding the scale’s validity.1

Given that marital satisfaction has been linked to marital difficulties (Vaughn &Baier, 1999), the need for marital counseling (Terling-Watt, 2001), intimacy(Hatch, James, & Schumm, 1986), and life stress management (Aniol & Snyder,1997), and that fact that practitioners and researchers continue to use theKMS, further examination of the validity of the scale is appropriate. As such,the purpose of this study is to yield additional empirical evidence regardingthe criterion- and construct-related validity of the scale to determine if thescale warrants use by financial consultants who are interested in learningmore about their married clients.

Method

Data for this study were collected from a survey conducted duringspring 2005. The survey was sent to randomly selected individuals from fourdatabases owned by the research team. Of the 1,318 surveys originally mailed(using the U.S. postal service), 548 were returned. Thirty-six were returned as

1 For a review of papers published in the Journal of Marriage and the Family withreferences related to the validity of the KMS go to: http://www.jstor.org/search/AdvancedSearch?si=1&hp=25&All=&Exact=kansas+marital+satisfaction&One=&None=&Search=Search&sd=&ed=&jt=&node.Sociology=1&ic=00222445%7C08857059%7C15381420



undeliverable, while three were returned with missing data. Nine surveys werenot opened. The useable return response rate was 39.37%.

The sample was delimited to include only married respondents (N =293). The average respondent was 45.23 years of age, which compared to aspouse’s age of 45.69 years. The typical respondent had been married 20.50years, with a standard deviation of 12.82 years. Eighty-nine percent ofrespondents were employed full-time, while only 63% of respondents’spouses were employed full-time. The average respondent reported house-hold income between $60,000 and $70,000. In general, respondents were welleducated. Fifty-eight percent reported having at least a college degree levelof education or higher. The sample consisted of 95% non-Hispanic Whites,with the remainder being African-American, Hispanic, or Asian. Finally,approximately 89% of respondents owned their own homes. Demographicdata for the sample are shown in Table 1.

Table 1Demographic Characteristics of SampleSample Characteristic Mean/Standard DeviationAge 45.23/11.47Spouse’s Age 45.69/11.95Gender (1 = Male) .32/.47Years Married 20.50/12.82Number of Children in Household 1.03/1.19Household Income 5.21/2.45

1 = Less than $20,0002 = 20,001 - $30,0003 = $30,001 - $40,0004 = $40,001 - $50,0005 = $50,001 - $60,0006 = $60,001 - $70,0007 = $70,001 - $80,0008 = $80,001 - $90,0009 = $90,001 - $100,00010 = More than $100,000

Employed Full Time (1 = Yes) .89/.31Spouse Employed Full Time (1 = Yes) .63/.48Race/Ethnic Background (1 = Non-Hispanic White) .95/.23Level of Education (1 = College or Higher) .58/.49Homeownership (1 = Own Home) .89/.32


Practice M

anagement

Measures

In this study, the KMS (Table 2) was measured using a five-pointLikert-type scoring system, with 1 = Extremely Dissatisfied, 3 = Mixed, and 5 =Extremely Satisfied. The mean score of respondents was 12.55, with astandard deviation of 2.72. The reliability estimate, as measured withCronbach’s alpha, was .97.

Table 2Kansas Marital Satisfaction Scale ItemsItem Mean/Standard

Deviation1. How satisfied are you with your marriage? 4.16/.952. How satisfied are you with your marriage withyour husband/wife? 4.16/.923. How satisfied are you with your husband/wifeas a spouse? 4.24/.95

Relationship Assessment Scale (RAS). The RAS was developed inthe 1980s to measure a person’s subjective evaluation of a close relationship,which may or may not be a marriage (Vaughn & Baier, 1999). Research usingthe scale suggests that it is highly correlated with the Dyadic AdjustmentScale, and as such, a good predictor of relationship stability and quality(Hendrick, 1988). The current study is the first to assess the relationshipbetween the RAS and KMS. In this study, the seven items in the scale (Table3) were measured using a five-point Likert-type measurement system, with 1 =Extremely Dissatisfied, 3 = Mixed, and 5 = Extremely Satisfied. The mean scorefor respondents was 29.30, with a standard deviation of 5.63. Cronbach’salpha for the scale was .93.

Table 3Relationship Assessment Scale ItemsItem Mean/Standard

Deviation1. How well does your spouse meet your needs? 4.01/.902. In general, how satisfied are you with yourrelationship? 4.12/.973. How good is your relationship compared to most? 4.19/.934. How often do you wish you hadn’t gotten intothis relationship? 4.42/.965. To what extent has your relationship met youroriginal expectations? 3.90/1.02



6. How much do you love your spouse? 4.63/.757. How many problems are there in your relationship? 4.03/1.08Note: Items 4 and 7 were reverse coded

Type-A Personality. It was hypothesized that respondents whoexhibited signs of Type –A behavior (e.g., strict, rigid, perfectionist, etc.)would be less satisfied with their marriage. Type-A Personality was measuredwith six items, as shown in Table 4. These items were adapted from Eaker andCastelli (1988). The scale was scored as follows: 1 = Not at all; 2 = Somewhat;3 = Fairly well; and 4 = Very well. The overall mean and standard deviation forthe Type-A scale was 13.91 and 3.15, respectively.

Table 4Type-A Personality ItemsItem Mean/Standard

Deviation1. Being bossy or dominating 1.84/.692. Having a strong need to excel (be best) in mostthings 2.72/.923. Usually feeling pressured for time 2.55/.884. Being hard driving and competitive 2.22/.915. Eating too quickly 2.48/1.076. Upset when have to wait on anything 2.13/.79

Sensation Seeking. It was hypothesized that respondents scoringhigh in terms sensation-seeking attitudes would be less satisfied with theirmarriage. Sensation Seeking was measured using five items (Table 5). Thescale used was similar to one originally designed by Arnett (1994). The scalewas scored as follows: 1 = Not at all; 2 = Somewhat; 3 = Fairly well; and 4 =Very well. The mean for this scale was 11.68 and the standard deviation was2.58. Items four and five were reverse coded.


Practice M

anagement

Table 5Sensation Seeking ItemsItem Mean/Standard

Deviation1. It’s fun and exciting to perform or speak beforea group. 1.86/.992. I would prefer to ride on the roller coaster orother fast rides at an amusement park. 1.97/1.123. I would like to travel to places that are strangeand far away. 2.64/1.054. I think it’s best to order something familiar wheneating in a restaurant. 2.61/.895. If I have to wait in a long line, I am usuallypatient about it. 2.60/.87

Self-Esteem. The relationship between marital satisfaction and self-esteem was hypothesized to be positive, with those exhibiting higher levels ofself-esteem also indicating a higher level of marital satisfaction. Self-Esteemwas measured using 10 items. Self-Esteem items are shown in Table 6 (notethat items 2, 6, 7, 8, 9, and 10 are reverse coded). The scale used here is basedon a scale originally developed by Rosenberg in 1965 and later revised byDidato (2003). The scale has shown high levels of reliability and validity inprevious studies. In the current study, scores ranged from a low of 10 to ahigh of 40 with a mean and standard deviation of 30.77 and 2.58, respectively.

Table 6Self-Esteem ItemsItem Mean/Standard

Deviation1. I am usually comfortable and poised amongstrangers. 2.53/.822. I am often jealous or envious of others. 3.47/.603. I always accept compliments without feelingembarrassed. 2.25/.744. I openly show recognition and appreciationwhen others do something noteworthy. 3.20/.675. I can almost always accept disagreementswithout feeling “put down.” 2.42/.696. I strongly seek recognition and praise. 3.00/.847. I am known as one who is hard to please. 3.48/.678. I am often miffed if the opinions of others differfrom mine. 3.56/.54



9. I am sometimes embarrassed in public by thoseclose to me. 3.51/.6410. I judge my worth by comparing myself to others. 3.32/.75

Locus of Control. As with self-esteem, the hypothesized relationshipbetween locus of control and marital satisfaction was expected to be positive.Specifically, those with a strong internal locus of control were expected to bemore satisfied with their marriage (Bugaighis et al., 1983). Locus of Controlwas assessed using a 10-item scale based on work conducted by Didato(2003). The Locus of Control items are shown in Table 7. Items 2, 4, 5, 7, 8,and 9 were reverse scored. Scores on the scale ranged from a low of 10 to ahigh of 40, with a mean and standard deviation of 29.82 and 2.92, respectively.

Table 7Locus of Control ItemsItem Mean/Standard

Deviation1. When I am certain that I am right I canconvince others. 2.38/.702. It’s probably silly to think that I can changesomeone’s basic attitudes. 2.83/.893. Success in school or work is due mainly to myown efforts and frame of mind. 3.14/.734. Whether I make a lot of money in life ismostly a matter of luck. 3.36/.725. There’s not much that a disadvantaged personcan do to succeed in life unless he or she is educated. 3.07/.916. Assuming there are two teams of equal skill, thecheering of the crowd is more important than luck indetermining the winner. 2.32/.867. Most problems work themselves out. 2.59/.788. I sometimes get a feeling of being lucky. 3.07/.819. I own a good luck charm. 3.89/.4610. It’s better to be smart than lucky. 3.17/.82

Religiosity. Schumm et al. (1982) and Anderson et al. (1983) foundthat the amount of religiosity someone exhibits is associated with maritalsatisfaction. In general, those who are more religious in their daily activitiestend to be more satisfied with their marriages. In this study, respondents wereasked to rate how much their religious beliefs influence their daily life.Response categories included: 1 =Very Much; 2 = Quite a Bit; 3 = Some; 4 =Little; 5 = None. The mean and standard deviation level of religiosity amongsurvey respondents was 2.26 and 1.19, respectively.


Practice M

anagement

Psychosocial characteristics, such as Type-A personality, sensationseeking, self esteem, locus of control, and religiosity, are factors also knownto affect they way in which people manage their personal finances. Forexample, Grable and Joo (2004) and Wong and Carducci (1991) documentedthe significant impact that these types of variables have in influencing riskpreferences among those engaged in everyday risk taking activities. Usingthese variables as controlling factors within the modeling process is one wayto isolate the role of marital satisfaction as a predictor of marital instability.Furthermore, assuming a relationship between and among marital satisfactionand these psychosocial factors exists, it may be possible, in future research,to link conclusively marital satisfaction with financial satisfaction.

Results

Criterion-related validity reflects the relationship between a scale andone or more criterion believed to be representative of the attribute or behaviorunder study (Huck & Cormier, 1996). Most often a Pearson product-momentcorrelation coefficient is calculated to reflect the association between a scaleand the criterion measure. The concurrent validity (i.e., a special form ofcriterion-related validity) of the KMS was assessed by calculating thestatistical significance of correlation coefficients between the KMS and sixitems thought to be associated with marital quality: The Relationship Assess-ment Scale, Type-A Personality, Sensation Seeking, Self-Esteem, Locus ofControl, and Religiosity. Each of these factors is described below.

Table 8 displays the Pearson product moment correlations betweenthe KMS and the criteria used in the analysis. The KMS was highly statisti-cally correlated with the RAS, and moderately statistically correlated withType-A Personality, Self-Esteem, and Religiosity. These findings support theconcurrent validity (i.e., criterion-related validity) of the KMS.

Table 8Validity CoefficientsCriterion Correlation with KMSRelationship Assessment Scale .88**Type-A Personality -.22**Sensation Seeking -.03Self-Esteem .20**Locus of Control .01Religiosity -.18***p < .01 **p < .001

A second test was undertaken to evaluate the construct-relatedvalidity of the KMS. Construct-related validity assesses the extent to which



a scale reflects the psychological construct it purports to measure. For a scaleto meet an adequate level of construct-related validity the items included inthe tool must represent the larger elements that comprise the construct, andthe items that make up the scale must be an appropriate measurement of thoseelements. Empirically, the KMS should be highly associated with an indepen-dent observation of a person’s own assessment of their marriage or a maritalbehavior (e.g., divorce). A logistic regression was used to test the constructvalidity of the KMS.

The dependent variable used was based on responses to thefollowing question: “Even people who get along quite well with their spousesometimes wonder whether their marriage is working out. Have you thoughtyour marriage might be in trouble within the last 3 years?” Those whoanswered occasionally, often, or very often were coded 1, while those whoanswered never were coded 0. Approximately 49% of respondents indicatedthat they thought their marriage might be in trouble.

The regression model included all of the demographic variablesshown in Table 1 (excluding spouse’s age) in addition to scales for MaritalSatisfaction (KMS), Type-A Personality, Self-Esteem, and religiosity. Fiveadditional variables were included in the model. Financial satisfaction wasmeasured by asking respondents to use a 10-point scale to state how satisfiedthey were with their present financial situation. Inclusion of this variable wasbased on previous research showing a relationship between financial satisfac-tion and marital satisfaction (Aniol & Snyder, 1997). Mean and standarddeviation scores were 5.90 and 1.94, respectively, indicating a higher thanaverage level of satisfaction. The age difference between the respondent andspouse was also used. Parent’s marital status when the respondent wasmarried was also measured and used in the analysis. Those respondentswhose parents were married when the respondent was married were coded 1,otherwise 0. Financial stressors were also included (see Aniol & Snyder). Inthis study, respondents were presented with a list of 24 financial events thatare known to cause stress (e.g., changing jobs, becoming disabled, incurring amajor household repair expense, etc.). Respondents were asked to check eachevent that had occurred over the past year. These events were then summedinto a financial stressor index. The mean score was 2.16, with a standarddeviation of 1.44. All independent variables were entered simultaneously. Itwas hypothesized that in order to confirm the construct validity of the scale,the KMS should be statistically significantly associated with marriage troubleresponses. Specifically, those with lower KMS scores should tend to havethought, on average and holding all other factors constant, that their marriagewas in trouble. The logistic regression results are shown in Table 9.


Practice M

anagement

Results from the logistic regression support the construct-relatedvalidity of the KMS. Holding all other factors constant, lower KMS scoreswere associated with an increased likelihood of thinking a marriage might be introuble. Two other factors were also found to be associated with maritaltrouble. Both age and employment status were negatively related to thethought of trouble. Younger respondents and those that were employed full-time were more likely to think their marriage was in trouble. Together, thevariables were able to accurately categorize respondents into known groupswith a success rate of approximately 75%.

Discussion

A test of the KMS’s concurrent validity found that the scale is highlycorrelated with the reliable and valid Relationship Assessment Scale (RAS),Type-A personality, self-esteem, and religiosity. These findings supportsimilar results from other validity studies found in the literature (e.g., Kurdek,1992; Schumm et al., 1982; Schumm et al., 1986). The relationship between theKMS and the RAS and self-esteem scales was positive, suggesting that thosewho were more satisfied with their relationship and those with higher levels ofself-esteem also scored higher in terms of marital satisfaction. Type-Apersonality was inversely related to the KMS. Those respondents thatexhibited Type-A behaviors were found to have lower levels of maritalsatisfaction. Finally, religiosity was negatively related; however, this findingshould be interpreted with caution given the way the variable was coded.Low scores indicated higher levels of religious values. As such, those thatwere more religious tended to be more satisfied with their marriage.



Table 9Construct-Related Validity Logistic Regression ResultsVariable B S.E. Wald Exp(B)KMS -.62 .10 41.60 .54***Financial Satisfaction -.07 .10 .54 .93Type-A Personality .07 .06 1.78 1.08Self-Esteem -.06 .05 1.23 .94Gender (1 = Male) -.43 .41 1.11 .65Age -.03 .02 3.15 .97*Age Difference -.01 .05 .09 .99Children in Household .13 .15 .81 1.14Religiosity .15 .15 1.07 1.17Household Income -.16 .46 .12 .85Parent’s Marital Status When Married(1= Married) -.04 .36 .01 .96Employment Status (1 = Full Time) -1.28 .67 5.16 .28**Spouse’s Employment Status (1 = Full Time) .06 .37 .02 1.06Race/Ethnic Background (1 = Non-HispanicWhite) -.41 .77 .29 .66Educational Background (1 = College orHigher) -.24 .38 .41 .79Housing Status (1 = Own Home) -.22 .60 .13 .81Financial Stressors -.12 .13 .86 .88

*p < .10 **p < .05 ***p < .01

The KMS was also found to be associated with respondents’feelings of marital troubles. This finding is noteworthy because the result wasbased on a multivariate logistic regression. Holding all other factors constant,the lower a respondent’s KMS score the more likely they were to think that, atsome point in the past three years, their marriage was in trouble. The higherthe KMS score the less likely someone was to think their marriage was introuble.

The findings from this study suggest that the criterion-related andconstruct-related validity of the KMS is sufficiently strong to be used bypractitioners and researchers who are interested in measuring their client’smarital quality. When considering personality factors, basic demographicdifferences, and financial attitudes and behaviors, the KMS is able to differen-tiate among those who have indicated marital instability within the past threeyears. So to reiterate, this instrument offers potential for both practitionersand researchers.


Practice M

anagement

However, two particular research needs continue to be of importance.First, a predictive criterion-related validity test is needed. Recall that thesurvey used in this study was cross-sectional, which limited the criterion-related test to a concurrent validity analysis. A longitudinal or panel study isneeded to determine the predictive usefulness of the scale. Second, the KMSought to be included in ongoing national surveys of marital quality. The scaleoffers several advantages. The scale is short. It has a consistently highalpha, and a sufficiently strong level of criterion- and construct-relatedvalidity. If included in national studies, results could then be used to normassessment scores to representative national and sub-sample cohorts.Ultimately, this information could be used to better educate practitioners inthe use and interpretation of the scale.

The results from this study indicate that the KMS can providefinancial consultants with an easy to implement and interpret measure of theirclient’s marital satisfaction. It is not uncommon for married clients who areexperiencing relationship distress to take financial actions that are opposite totheir best interests. Using this tool, while not appropriate in all client situa-tions, does offer a way to determine why a married client is making financialdecisions that run counter to financial planning recommendations. Withoutan accurate assessment of marital satisfaction some financial consultants maybe led to believe that their client’s actions are the result of a breakdown in theplanner/client relationship. Arriving at this conclusion can lead to recommen-dations that take a client further from the core cause of their distress. Usingthe KMS can help a financial consultant make a very difficult recommenda-tion, namely, referring a distressed client to a marriage and family therapist. Atthe very least, the use of the KMS can be used as a starting point in discus-sions with a client about what is driving negative financial behaviors.

References

Amato, P. R. & Previti, D. (2003). People’s reasons for divorcing: Gender, socialclass, the life course, and adjustment. Journal of Family Issues, 24,602-626.

Amato, P. R., & Rogers, S. J. (1997). A longitudinal study of marital problemsand subsequent divorce. Journal of Marriage and the Family, 59,612-624.

Anderson, S. A., Russell, C. S., & Schumm, W. R. (1983). Perceived maritalquality and family life-cycle categories: A further analysis. Journal ofMarriage and the Family, 45, 127-139.

Aniol, J. C., & Snyder, D. K. (1997). Differential assessment of financial andrelationship distress: Implications for couples therapy. Journal ofMarital and Family Therapy, 23, 347-352.



Arnett, J. (1994). Sensation seeking: A new conceptualization and a new scale.Personality and Individual Differences, 16, 289-296.

Ary, D., Jacobs, L. C., & Razavieh, A. (1990). Introduction to research ineducation. Orlando, FL: Holt, Rhinehart and Winston, Inc.

Berry, R. E., & Williams, F. L. (1987). Assessing the relationship betweenquality of life and marital and income satisfaction. Journal ofMarriage and the Family, 49, 107-116.

Bugaighis, M. A., Schumm, W. R., Bollman, S. R., & Jurich, A. P. (1983). Locusof control and marital satisfaction. Journal of Psychology, 114, 275-279.

Burgess, E. W., & Cottrell, L. S. (1939). Predicting success or failure inmarriage. New York: Prentice-Hall.

Calahan, C. A. (1997). Internal consistency, reliability, and concurrent validityof the Kansas Marital Satisfaction scale and the Quality MarriageIndex. Psychological Reports, 80, 49-50.

Didato, S. V. (2003). The big book of personality tests. New York: Black Dogand Leventhal.

Eaker, E. D., & Castelli, W. P. (1988). Type A behavior and coronary heartdisease in women: Fourteen-year incidence from the Framinghamstudy. In B. K. Houston & C. R. Snyder (Eds.), Type A behaviorpattern: Research, theory, and intervention. New York: John Wiley& Sons.

Eggeman, K., Moxley, V., & Schumm, W. R. (1985). Assessing spouses’perceptions of Gottman’s temporal form in marital conflict.Psychological Reports, 57, 171-181.

Grable, J. E., & Joo, S-H.(2004). Environmental and biopsychosocial factorsassociated with financial risk tolerance. Financial Counseling andPlanning, 15 (1), 73-88.

Grover, K. J., Paff-Bergen, L. A., Russell, C. S., & Schumm, W. R. (1984). TheKansas Marital Satisfaction Scale: A further brief report.Psychological Reports, 54, 629-630.

Hatch, R. C., James, D. E., & Schumm, W. R. (1986). Spiritual Intimacy andmarital satisfaction. Family Relations, 35, 539-545.

Hendrick, S. S. (1988). A generic measure of relationship satisfaction. Journalof Marriage and the Family, 50, 93-98.

Huck, S. W., & Cormier, W. H. (1996). Reading statistics and research. NewYork: HarperCollins College Publishers, Inc.

Joo, S., & Grable, J. E. (2004). An exploratory framework of the determinants offinancial satisfaction. Journal of Family and Economic Issues, 25,25-50.

Karney, B. R., & Bradbury, T. N. (1997). Neuroticism, marital interaction, andthe trajectory of marital satisfaction. Journal of Personality andSocial Psychology, 72, 1075-1092.


Practice M

anagement

Kerkmann, B. C., Lee, T. R., Lown, J. M., & Allgood, S. M. (2000). Financialmanagement, financial problems and marital satisfaction amongrecently married university students. Financial Counseling andPlanning, 11 (2), 55-64.

Kurdek, L. A. (1992). Dimensionality of the Dyadic Adjustment Scale:Evidence from heterosexual and homosexual couples. Journal ofFamily Psychology, 6, 22-35.

Mitchell, S. E., Newell, G. K., & Schumm, W. R. (1983). Test-retest reliability ofthe Kansas Marital Satisfaction Scale. Psychological Reports, 53,545-546.

Morris, M. L., & Blanton, P. W. (1994). Denominational perceptions of stressand the provision of support services for clergy families. PastoralPsychology, 45, 345-364.

Moxley, V., Eggeman, K., & Schumm, W. R. (1986). An evaluation of the“Recovery of Hope” program. Journal of Divorce, 10, 241-261.

Schumm, W. R., Bollman, S. R., & Jurich, A. P. (1982). The “maritalconventionalization” argument: Implications for the study ofreligiosity and marital satisfaction. Journal of Psychology andTheology, 10, 236-241.

Schumm, W. R., Paff-Bergen, L. A., Hatch, R. C., Obiorah, F. C., Copeland, J.M., Meens, L. D., & Bugaighis, M. A. (1986). Concurrent anddiscriminant validity of the Kansas Marital Satisfaction Scale.Journal of Marriage and the Family, 48, 381-387.

Shek, D. T. L., Lam, M. C., Tsoi, K. W., & Lam, C. M. (1993). Psychometricproperties of the Chinese version of the Kansas Marital SatisfactionScale. Social Behavior and Personality, 21, 241-249.

Terling-Watt, T. (2001). Explaining divorce: An examination of the relationshipbetween marital characteristics and divorce. Journal of Divorce andRemarriage, 35, 125-145.

Vaughn, M. J., & Baier, M. E. M. (1999). Reliability and validity of therelationship assessment scale. The American Journal of FamilyTherapy, 27, 137-147.

White, M. B., Stahmann, R. F., & Furrow, J. L. (1994). Shorter may be better: Acomparison of the Kansas Marital Satisfaction Scale and the Locke-Wallace Marital Adjustment Test. Family Perspective, 28 (1), 53-66.

Wong, A., & Carducci, B.J. (1991). Sensation seeking and financial risk takingin everyday money matters. Journal of Business and Psychology, 5,525-530.

Contact Information: John E. Grable, Ph.D., CFP®, RFC, Program Director,Institute of Personal Financial Planning, 318 Justin Hall, Family Studies andHuman Services, Manhattan, KS 66506; Phone: (785) 532-1486; Fax: (785) 532-5505; E-mail: [email protected]



ABSTRACT

The purpose of this study was to investigate the long-term relationbetween household leverage through the use of mortgages, andchanges in household wealth using the theoretical framework of the lifecycle income hypothesis. This study used the 1992 through 2002 wavesof the Health and Retirement Study. The characteristics of leveragedand unleveraged households were compared in 1992 and 2002 as werechanges in wealth during that period. Households without mortgagesconsistently had greater net assets relative to households with mort-gages. Whether members of the household were currently working wasconsistently associated with carrying a mortgage.

MORTGAGES AND ASSET ACCUMULATION:A TEN YEAR COMPARISON OF HOMEOWNERS

Lance Palmer, Ph.D.The University of Georgia

Jean M. Lown, Ph.D.Utah State University

Introduction

Consumers looking for financial advice regarding whether theyshould pay-off their mortgages ahead of schedule or keep the mortgages andinvest the extra money, find contradictory opinions in both the popular andprofessional press. Many in the financial community argue that mortgages,with their low cost and favorable tax treatment, provide excellent capital forinvesting. Others counter, arguing that the interest rate charged on mort-gages is a high hurdle for the average risk-averse investor to overcome (Goff& Cox, 1998; Orman, n.d.; Storms, 2000; Tomlinson, 2002).

The economy has provided borrowers with several incentives tocarry mortgages. During the late 1990s and early years of the 2000s, mortgageinterest rates were at or near historical lows and home owners benefited fromsignificant appreciation of home values which resulted in large increases inhome equity. By refinancing at lower interest rates, homeowners could


Research &

Theory

potentially cash-out some of their accumulated equity and simultaneouslylower their monthly payment (Coy & Keenan, 2003).

In addition to low interest rates, the tax system in the United Statesallows households that itemize their deductions to include mortgage interestin their income tax deduction calculation. Mortgage interest is deductiblewhen the household has itemized deductions in excess of the standarddeduction. According to the United States Department of the Treasury (1996),approximately 30% of households itemize their deductions. The itemizeddeduction of interest on home-secured debt reduces the after-tax cost ofmortgages by the amount of the excess interest deduction above the standarddeduction multiplied by the borrower’s marginal tax rate. This tax break hasthe potential of creating an artificially low cost of debt for some households.Favorable tax treatment of mortgage interest encourages households to carrylarger mortgages than they otherwise would. Consequently, many house-holds have reallocated their debt portfolios to increase their mortgages andreduce other forms of debt (Dunsky & Follain, 2000; Stango, 1999).

While incentives abound, the lack of a consensus among financialwriters and planners has left consumers without a clear understanding ofwhether mortgages, aside from the purchase of a home, can be beneficial in ahousehold’s financial portfolio. This question has been approached from atheoretical perspective using an expected utility model, Monte Carlo simula-tions, and other bootstrap statistical models (Palmer, 2002; Tomlinson, 2002;Waggle & Johnson, 2003), and also from a practitioner’s perspective with theuse of case studies and hypothetical scenarios (Goff & Cox, 1998; Storms,1996, 2000). Both methods inadequately address the long-term consequencesof keeping or eliminating mortgages, since neither method addresses actualhousehold behavior, nor provides a means for a retrospective analysis of thedecision.

An actual examination of households with and without mortgages isnecessary to understand homeowner’s behavior and whether either choiceyields positive economic benefits to the household. This study investigatedthe long-term relation between household leverage, through the use ofmortgages, and changes in household wealth. The results will help support orrefute current positions regarding household leverage through mortgages.The specific objectives of this study are to a) compare and contrast thecharacteristics (i.e., debt, assets, income, portfolio allocation, and demograph-ics) of leveraged households (households with mortgages) and unleveragedhouseholds (households without mortgages) in 1992 and 2002 and b) discussthe general implications of carrying mortgages for consumers, financialprofessionals, and educators.



Review of Literature

The Choice to Leverage

Carrying a mortgage and investing any additional money – thatwould otherwise be paid toward early retirement of the loan – in investmentsthat yield a higher after-tax rate of return than the after-tax interest rate paid onthe mortgage is a commonly promoted strategy (Edelman, 2001; Johnston,2000; Storms, 2000). This strategy diversifies the assets in a household’sportfolio by reducing the amount of money held as home equity, and couldenhance the financial wealth of individuals and in turn may increase theiroverall life time consumption. Since investment returns are uncertain whilemortgage payments are certain, this strategy is risky. Furthermore, mortgagesare generally secured by the individual’s primary residence, making the choiceto carry a mortgage for investment purposes a potentially emotional decision.

When evaluating the choice to leverage an individual’s assets with amortgage, a common and popular comparison used is the historical return onequity investments versus the investor’s current interest rate on his or hermortgage. While this is a convenient comparison, most investors experiencerates of return well below historical market rates of return, nullifying theappropriateness of this comparison (Dalbar, Inc., 2001).

According to a recent study by Dalbar, Inc. (as cited in Clements,2004), the average annual return on equity mutual funds for the 19 yearsending December 2002 was 11.8%. However, over that same period theaverage annual return realized by equity mutual fund investors was only 2.6%.The reason cited for this large disparity was mutual fund owners’ relativelyshort holding period, approximately 2.6 years. Dalbar, Inc. (2001) suggestedthat mutual fund investors appeared to be switching between funds fre-quently, rather than employing a long-term, buy-and-hold strategy. However,Dalbar’s findings are subject to debate. Clements pointed out a bias inDalbar’s methodology which, when corrected, increased annual investorreturns to 8.2% and reduced the gap between actual investor returns and themarket’s performance to 3.4 percentage points. Under the revised methodol-ogy, individual investors appeared to perform better, yet still lagged behindthe overall market.

Comparisons using the historical rate of return in the equity marketto current mortgage rates are also inappropriate because the average investordoes not allocate 100% of their portfolio to stocks. Waggle and Johnson(2003) examined optimal portfolio allocations using a mean variance analysisand expected utility model and found that the optimal portfolio allocation formoderately risk-averse households with relatively large mortgages does notinclude a substantial allocation to stocks. For households with a high loan tovalue ratio and relatively modest financial asset holdings, optimal portfoliosincluded as little as 12% equity allocation.


Research &

Theory

Historical rates of return on equities, such as those published byIbbotson Associates (2002), are calculated using broad market indices basedon a buy-and-hold strategy. Given the major difference between mutual fundinvestors’ behavior, as cited by Dalbar, Inc. (2001) and the method forcalculating historical rates of return, it does not seem appropriate for indi-vidual investors to use historical rates of return on equities to compare thealternatives of mortgage prepayment and investing—unless the investor hasconsistently used a buy-and-hold strategy and invested the majority of his orher assets in a market representative portfolio of equities.

Factors Associated with Wealth

Many factors affect household net worth. Some factors discussedhere include inheritances, health, risk tolerance, and household behaviorssuch as having a regular savings plan (Kennickell and Starr-McCluer, 1997).Kotlikoff and Summers (1981) estimated that 80% of the stock of U.S. wealthwas a result of inheritances from older generations, while only 20% wasaccounted for by current savings. Modigliani (1986) argued that the percentof inherited wealth is much less, and based on a survey of research resultsand methods, estimates the amount of bequeathed wealth at no more than25% of households’ current asset holdings. Nevertheless, bequeathed wealthis a significant factor.

Recent research on the relation between health and wealth hashighlighted a strong correlation between the two concepts (Adams, Hurd,McFadden, Merrill, & Ribeiro, 2003; Meer, Miller, & Rosen, 2003). Meer et al.(2003), using the Panel Study of Income Dynamics (PSID) and instrumentalvariable methodologies, showed that the dominant path is from health towealth rather than from wealth to health, especially over short-term periods.Adams et al. (2003) also arrived at similar conclusions using the Asset andHealth Dynamics of the Oldest Old (AHEAD). They found no evidence of alink between wealth and mortality or the sudden onset of acute disease.Adams et al. did find evidence of a causal link from health conditions to totalwealth changes. The effect of health on wealth appears to be well establishedin the literature, while the effect of wealth on health results in insignificant,mixed, or unsubstantial effects.

Spencer and Fan (2002) suggested that a household’s willingness toincur debt is dependent upon its risk tolerance. Risk tolerance is also animportant aspect of wealth accumulation because it is a major determinant ofhow an individual’s portfolio is allocated among different assets, and thusdetermines the rate of return on assets (Gutter, 2000). Several studies haveexamined the determinants of risk tolerance and its importance in assetallocation decisions (Grable & Lytton, 1998; Schooley & Worden, 1996; Sung& Hanna, 1996; Wang & Hanna, 1997).



The characteristics of borrowers are also an important considerationwhen examining household leverage. While interest rates, asset prices, andtax incentives affect a household’s willingness to borrow, other demographicfactors are also important. Households with outstanding debt tend to be morelikely to be young, non-white, employed, less formally educated, homeownerswith lower net worth, larger household sizes, and higher incomes (Chen &Jensen, 1985; Crook, 2001; Salandro & Harrison, 1997; Spencer & Fan, 2002;Zhu & Meeks, 1994). Maki (1996) reported similar results except that marriedhouseholds and more educated households were more likely to carry largermortgages.

Theoretical Framework

According to the life cycle income hypothesis (Ando & Modigliani,1963), individuals seek to smooth their consumption over their lives byborrowing and saving at different stages of the life cycle, thus affecting thehousehold’s current portfolio of assets, debts, and net worth. The fundamen-tal idea of the life cycle income hypothesis is that individuals base theirconsumption on total life resources and not on current income. Total re-sources include current net worth, current income, and the present value offuture earned income and other cash flows, like Social Security. From theseresources, a permanent income flow is estimated and the individual’s con-sumption, a proxy for utility or satisfaction, is based on this permanent flow ofincome.

Hanna, Fan, and Chang (1995) used the life cycle income hypothesisto model household borrowing, saving, and net worth over the adult years.Under circumstances of rising real income, their model predicted that rationalconsumers would borrow to increase consumption in early years, repay theborrowed funds, and then accumulate wealth. Hanna et al. made the simplify-ing assumption that individuals can borrow and save at the same interest rateand ignored the effect of taxes. The model results in a “hump shape” ofwealth accumulation because households borrow, save, and then spend downwealth during retirement. One assumption of this theory is that individualswill have debt at a time when they begin saving. If the interest rate on debtand savings is the same, as is the case in Hanna and colleagues’ (1995) model,then it makes no difference whether the individual saves through debt payoffacceleration or through separate savings accounts.

In reality, individuals face a multitude of interest rates when borrow-ing and saving and often find that, as a result, there is a difference in the rateat which funds can be borrowed, and saved or invested. The differences ininterest rates may lead individuals to simultaneously borrow and save, bymaintaining relatively low-cost debt and saving transitory income (thedifference between current income and permanent income) in higher yieldingaccounts.


Research &

Theory

Hypotheses

Based on the review of literature, the primary hypotheses arepresented. The life cycle income hypothesis would result in householdskeeping mortgages as long as possible and investing money, even if it is atthe same pre-tax interest rate as their mortgage, rather than retire theirmortgage early. Given this theoretical premise, leveraged households shouldhave larger positive changes in wealth than unleveraged households.However, if households used mortgage proceeds to supplement consumption,and did not invest extra money that could be used to pay down the mortgagebut instead spent it, then leveraged households would have smaller, ornegative changes in assets compared with unleveraged households. Specifi-cally, it was hypothesized that:

1) leveraged households would have greater assets thanunleveraged households;2) leveraged households would experience greater increases inassets than unleveraged households; and3) leveraged households would have a higher allocation of assets torisky assets than unleveraged households.

The U.S. utilizes a progressive income tax structure so that higherincome households pay taxes at higher rates. As a result of this tax structure,higher marginal tax bracket households realize greater tax savings frominterest deductions (Stango, 1999). Consequently, households in highermarginal tax brackets have the potential of garnering the greatest benefitsfrom leveraging themselves with mortgages. Maki (1996) found evidence thatonly high-income, financially sophisticated households showed evidence ofshuffling their debt holdings in response to the 1986 tax law changes.Therefore, it was also hypothesized that 4) leveraged households would havehigher household income than unleveraged households.

Based on the literature, compared to unleveraged householdsleveraged households were hypothesized to be younger, working, in betterhealth, married, non-white, less likely to plan on leaving an inheritance, andmore likely to receive an inheritance. Leveraged households were alsohypothesized to have higher non-mortgage debt than unleveraged house-holds.



Methodology

Chi-square and t-test statistics were used to address the objective ofthis study, which was to compare and contrast the characteristics (i.e., totalresources, income and savings, portfolio allocation, health, and demograph-ics) of leveraged households with unleveraged households in 1992 and 2002.Mean housing debt by age cohort was also analyzed to identify potentialcohort effects on carrying a mortgage.

The Health and Retirement Study (HRS) is an ongoing nationallongitudinal survey conducted every two years by the Survey ResearchCenter at the University of Michigan. The analysis for this study used datagathered in the 1992, 1994, 1996, 1998, 2000, and 2002 waves of the Health andRetirement Study. The original HRS sample consisted of individuals and theirpartners, if applicable, who were between the ages of 51 and 61 at the time ofthe first wave in 1992. The sample size for the initial wave of the HRS con-sisted of 12,654 individuals. The analysis for this study limited the sample tostable households–households that did not experience a change in maritalstatus during the period of observation, were interviewed in each wave of thestudy, and reported owning their home in 1992. The final sample consisted of2,770 households. Household leverage—as measured by all mortgagesdivided by total assets—income and savings, health status, and somedemographic variables were calculated over the 1992 to 2002 time period. Alldollar calculations were adjusted to reflect constant 2002 dollars.

The HRS includes imputed values for missing financial information.Missing values for some other variables associated with pensions were alsoimputed. In order to preserve the sample size for this study, imputed valuesfor missing information were used.

Total household assets included: bank accounts, CDs, stocks,bonds, mutual funds, IRAs, Keoghs, cash value life insurance, annuities,defined contribution retirement plans, collectibles, vehicle equity, homeequity, other real estate, and business holdings. Total resources included allof total assets plus the present value of anticipated Social Security benefits,Veterans Administration benefits, defined benefit pension plan benefits, andother annuitized cash flows.

Results

For sample statistics, means testing, and chi-square tests, the 1992household weights –included with the dataset – were applied to the house-holds. Average household assets increased substantially over the period ofobservation. Mean assets increased $405,510. The median household’sassets increased $179,250. Based on the observed standard deviation for theresults, there was substantial variation among households in terms of both


Research &

Theory

absolute and percent increases. In contrast to household assets, averagetotal resources available to the household decreased by $270,780. Similar tohousehold assets, substantial variation across households was observed.These results are summarized in Table 1.

Table 1Descriptive Statistics for Changes in Assets and Total Re-sources from 1992 to 2002 (Weighted)Variables Mean (Median) SD %Household leverage 1992 Housing debt (000s) 44.45 (15.40) 81.82 Mortgage debt to assets (x 1,000) 148.05 (57.44) 203.18 Change in housing debt Paid off 25.49 Kept or borrowed 43.95 No housing debta 30.56Income and work 1991 income (000s) 70.80 (56.46) 64.75 Work trend Working 1992: not working 2002 41.36 Not working 1992: working 2002 2.05 Not working 1992 or 2002 14.36 Working 1992 and 2002a 42.23Initial wealth and portfolio Assets (000s) 1992 401.97 (220.24) 651.03 2002 807.48 (465.94) 1,452.98 Change in assets 405.51 (179.25) 1,289.55

Total resources (000s)

1992 1,544.50 (1,087.16) 2,043.34 2002 1,273.71 (793.36) 2,254.67 Change in total resources -270.78 (-272.28) 1,313.92Variables Mean (Median) SD % Risky assets to total assets (x 100) 34.69 (30.77) 30.79 1992 other debts (000s) 3.29 (0.00) 14.56



Table 1 (continued)Descriptive Statistics for Changes in Assets and Total Re-sources from 1992 to 2002 (Weighted)Variables Mean (Median) SD %Inheritance Received inheritance 21.89 Plan to leave sizable estate Definitely 14.59 Probably 19.38 Possibly 16.28 Not likelya 49.55Health 1992 self-reported health status Excellent 14.20 Very Good 40.81 Good 31.33 Fair or Poora 13.66 Change in self-reported health No change 49.10 Improved 11.86 Declineda 39.04Demographics 1992 household size 2.40 (2.00) 1.03 Change in household size Increased 11.39 Decreased 28.73 Constanta 59.88Variables Mean (Median) SD % 1992 Age 57.59 (57.00) 4.55 1992 Education 12.82 (12.50) 2.60 Coupled status Single female 15.95 Single male 5.56 Married or partnereda 78.49Race Black or African American 6.82 Hispanic 3.94 Other 1.86 Non-Hispanic Whitea

87.38a Reference category.


Research &

Theory

The increase in assets and the simultaneous decrease in totalresources is consistent with the life cycle income hypothesis, in that prior toretirement, households accumulate assets; however, their human capital–measured by the present value of future earnings–declines as a result of feweranticipated years of work. The present value of public and private definedbenefit pensions also decreased as the households aged because they hadfewer years left to draw on life pensions.

Just less than one third of the sample reported not having a mort-gage in 1992 and 2002. Over the period of observation 43.95% of householdskept or increased their mortgages while 25.49% decreased their mortgages.

Average household income from all sources in 1991 was $70,796.This high income was likely a result of the sample selection process. Theaverage allocation of non-housing assets to risky assets was 34.69%. Themedian percentage of assets allocated to risky investments was 30.77. Morethan one fifth of the households received an inheritance during the period ofobservation and 14.59% of households, when asked in 1992, definitelyplanned to leave a sizable estate to their heirs, whereas 49.55% thought that itwas not likely that they would leave a sizable estate. Median householdconsumer debt was zero in 1992.

The majority of households in the sample reported having “Excel-lent” or “Very Good” health with only 13.66% reporting “Fair or Poor” health.Self-reported health status in 2002 was unchanged for 49.11% of the sample.A large percentage, 39.04%, of the sample reported lower self-rated health in2002 than in 1992. The reported decrease in health was not surprising giventhat these households were 10 years older and the high percentage ofhouseholds reporting “Excellent” or “Very Good” health in 1992.

The average household size was 2.40 individuals per household.The relatively large household size for this age group may have resulted fromthe restriction placed on the ending sample that marital status be constantduring the time period. The average age of the household in 1992 was 57.59years old. The majority of the sample was married while single men accountedfor only 5.56% of the group. The sample was overwhelmingly non-HispanicWhite.

Figure 1 depicts selected age cohorts and the average mortgageamount carried by each household for each year of observation. The age ofthe cohort was as of 1992. The mortgage amount fell for each cohort duringthe first years of observation. The oldest cohort’s debt continued to decreasewhile the younger cohorts’ average mortgage increased and decreased moresporadically over the remainder of the period. The youngest cohort consis-tently had the highest or near highest debt loads relative to the other cohortswhile the oldest cohort consistently had the lowest mortgage amounts. Thedifferences between cohorts are consistent with the life cycle income hypoth-esis in that younger households carry larger mortgages than older house-holds.



Yea

r

Rat

io

19

92

19

94

19

96

19

98

20

00

20

02

010

20

30

40

50

60

70

80

51

-52

53

-54

55

-56

57

-58

59

-60

61

-62

63

-64

Fig

ure

1M

ean

hous

ing

debt

(00

0s)

by a

ge o

f ol

dest

ind

ivid

ual

in 1

992

from

199

2 to

200

2 (w

eigh

ted)


Research &

Theory

Younger cohorts appeared to be more responsive to changes ininterest rates than older cohorts. Historically low interest rates from 2000 to2002 appeared to have motivated younger households to increase theirmortgages in real terms. Consistent with the life cycle income hypothesis, theoldest households appeared to be less responsive to changes in interest ratesthan younger households. However, for some older households, the fallinginterest rates appeared to have slowed the rate at which mortgages weredeclining. In general, the younger cohorts increased or maintained mortgageholdings, while the two oldest cohorts decreased or maintained mortgagebalances.

Comparison of Unleveraged and Leveraged Households

The sample was divided into two subgroups: those householdswithout mortgages and those with mortgages in 1992 and 2002. Table 2contains the results of the independent t-tests comparing 1992 and 2002group means along selected continuous variables. Unleveraged householdsaccounted for 36.10% of the total sample in 1992 and 56.06% of the sample in2002. Significant differences existed between leveraged and unleveragedhouseholds in 1992 and 2002. Leveraged households in 1992 and 2002 weresignificantly younger than unleveraged households and had significantlyhigher household incomes, education, consumer debt, and household sizethan unleveraged households.

Leveraged households experienced significantly larger decreases intotal resources over the subsequent period of observation. Unleveragedhouseholds reported significantly higher assets in both 1992 and 2002. Thehigher income, younger age, and higher total resources in 1992 of the lever-aged group may be indicative of a larger percentage of the subgroup working.



Table 2Results of Independent t tests Comparing the Leveraged andUnleveraged Households by Continuous Variables in 1992 and forthe Period from 1992 to 2002 (Weighted)

1992 Households 2002 HouseholdsUnleveraged(36.10%)Leveraged(63.90%) Unleveraged(56.06%)

Leveraged(43.94%)Variable Mean Mean t score Mean Mean t score

(SD) (SD) (SD) (SD)Assets (000s) 438.05 381.59 2.30* 861.61 738.43 2.32*

(693.79) (624.85) (1,717.49) (1,017.10)Change in assets 349.85 436.94 -1.79 444.01 356.41 1.86($, 000s) (1,178.80) (1,347.36) (1,500.07) (954.35)Total resources (000s) 1,314.85 1,674.18 -4.68*** 1,264.67 1,285.25 -0.25

(2,396.33) (1,801.70) (2,682.83) (1,546.12)Change in total -151.65 -338.06 3.77*** -123.00 -459.26 7.07***resources ($, 000s) (1,033.58) (1,444.38) (1,401.45) (1,166.65)Mortgage debt (000s) 0.00 69.55 -24.68*** 0.00 79.76 -30.34***

(0.00) (93.43) (0.00) (108.65)1991 household 57.06 78.55 -8.91*** 62.11 80.41 -5.17***income (000s) (57.98) (67.06) (97.72) (96.17)Risky asset 33.43 35.40 -1.70 33.98 35.60 -1.45allocation (%) (30.49) (30.95) (30.39) (31.29)Other debt (000s) 1.72 4.17 -4.49*** 2.53 4.37 -2.76**

(11.93) (15.79) (19.12) (17.25)1992 household size 2.26 2.49 -5.80*** 2.08 2.31 -6.93***

(0.93) (1.08) (0.80) (1.06)Age 58.83 56.88 11.60*** 58.54 56.36 13.53***

(4.80) (4.25) (4.65) (4.11)Education level 12.20 13.17 -10.12*** 12.53 13.19 -7.04***

(2.70) (2.48) (2.62) (2.53)*p < .05., **p < .01., ***p < .001.

In 2002, 56.06% of the sample did not have a mortgage. In 2002, therewere no significant differences in the amount of total resources or change inassets over the preceding period of observation between the two groups.Several of the differences observed in 1992 remained in 2002. Leveragedhouseholds continued to be statistically significantly younger and also havehigher household incomes, education, and household size. Leveraged house-holds had significantly more consumer debt in 2002 than unleveraged house-holds.

Similar to the analysis performed for continuous variables in 1992 and2002, chi-square tests of independence were performed comparing unleveragedand leveraged households in 1992 and 2002 for categorical variables. The resultsshown in Table 3 are similar to those shown in Table 2 in that significant differ-ences between unleveraged and leveraged households existed.


Research &

Theory

Leveraged households differed from unleveraged households in 1992and 2002 based on their work trend over the period of observation, initial totalresources, bequest expectations, self-rated health, changes in household size,and race. Consistent with the comparison of total resources in Table 2,leveraged households were more likely to be working in 1992 and 2002 thanunleveraged households. Over half of the leveraged households in 2002reported working in 1992 and 2002 and only 33.90% of unleveraged house-holds were working in 1992 and 2002.

In 1992, statistically significant differences in initial total resourcesalso distinguished the two groups. Higher proportions of leveraged house-holds were observed in the highest two total resources percentile brackets,whereas, larger proportions of unleveraged households were observed in thelowest percentile category. Similar to differences observed in 1992, in 2002larger proportions of leveraged households belonged to higher initial totalresources percentiles than unleveraged households. This was largely a resultof leveraged households being much more likely to be working compared tounleveraged households, thus having higher present values of futureearnings.

In 1992, leveraged households were much more likely thanunleveraged households to receive an inheritance during the subsequent 10year period. However, in 2002 there was no statistical difference in receivingan inheritance between leveraged and unleveraged households during thepreceding time period. While leveraged households generally had greatertotal resources, bequest expectations were more likely to be higher amongunleveraged households.

Table 3Results of Chi-Square Tests of Independence Comparing theLeveraged and Unleveraged Households by Categorical Vari-ables in 2002 and for the Period from 1992 to 2002 (Weighted)

1992 Households 2002 HouseholdsVariable Unlever. Lever. 2, df Unlever. Lever. 2, df

(36.10%) (63.90%) (56.04%)(43.96%)Change in mortgage debt 2,408.94***,23,046.00***, 2

Kept or borrowed 15.29 60.14 0.00 100.00Paid off 0.00 39.86 45.46 0.00No debta 84.71 0.00 54.54 0.00

Work status 98.40***, 3 123.59***, 3Working to not working 43.31 40.27 46.43 34.88

Not working to working 2.46 1.80 1.76 2.39Not working to not working21.38 10.37 17.92 9.86Working to workinga 32.85 47.56 33.90 52.88



Initial wealth percentile 109.00***, 4100.92***, 4

25th to 50th 24.91 24.09 27.05 20.9950th to 75th 21.45 24.55 22.72 24.3575th to 90th 10.64 15.97 10.71 18.3090th to 100th 5.45 12.69 7.20 13.820 to 25th a 37.55 22.70 32.32 22.55

Received inheritance 18.91 23.57 8.95**, 1 21.49 22.40 0.37, 1Likelihood of leaving an estate 21.83***, 3 12.07*,3

Definitely 16.11 13.71 15.47 13.45Probably 21.11 18.39 20.21 18.31Possibly 17.65 15.51 16.81 15.62Definitely or probably nota45.13 52.39 47.51 52.62

Health status in 2002 62.92***, 321.57***, 3

Excellent 9.91 16.64 4.34 6.27Very good 36.91 43.04 30.70 34.80Good 34.91 29.28 41.42 41.22Fair or poora 18.27 11.04 23.55 17.70

Change in health status 6.19*, 2 3.37, 2Maintained health 51.59 47.71 50.15 47.80Health improved 12.10 11.71 12.24 11.35Health declineda 36.31 40.58 37.61 40.85

Change in household size 30.26***, 250.93***, 2

Increased 9.91 12.22 10.60 12.41Decreased 23.73 31.54 24.06 34.68No changea 66.36 56.24 65.34 52.91

Coupled status 1.61, 2 6.92*, 2Single female 17.09 15.31 16.04 15.83Single male 5.45 5.60 6.50 4.33Marrieda 77.45 79.10 77.46 79.84

Race 17.30**, 3 16.52**,3African American 5.18 7.76 5.51 8.51Hispanic 5.09 3.29 3.93 3.96Other 1.18 2.26 1.35 2.46Whitea 88.55 86.70 89.22 85.06

a Reference category.*p < .05., **p < .01., ***p < .001.


Research &

Theory

Complementing the household’s work status was its ability to work.Higher proportions of leveraged households reported “Excellent” or “VeryGood” health relative to unleveraged households in 1992 and 2002. In 1992,unleveraged households were more likely to maintain or improve their healthduring the subsequent period of observation. Looking back over the 10 yearperiod in 2002 there was no statistical difference in health changes betweenunleveraged and leveraged households.

Leveraged households were more fluid than unleveraged householdsin 1992 and 2002, with higher proportions reporting changes in household sizeover the period. Looking back over the period from 2002, statistically signifi-cant differences in coupled status were also observable with higher percent-ages of single households categorized as unleveraged and married house-holds belonging to the leveraged group. Such differences did not exist in1992.

Significant differences in the racial and ethnic composition ofleveraged and unleveraged households were observed in 1992 and 2002.African Americans and other non-white households were consistently overrepresented among leveraged households relative to unleveraged house-holds, while higher percentages of Hispanics and non-Hispanic Whites wereobserved among unleveraged households.

Based on the results of independent t-tests and chi-square tests, thegeneral difference between the two groups was that greater proportions ofleveraged households were working in 1992 and 2002 compared tounleveraged households. This observation was supported by the differencesin earned income, work status trends, age, total resources, and changes intotal resources. With respect to asset holdings and changes in assets thefindings were mixed. Subgrouping households based on leverage status in1992 and 2002 resulted in unleveraged households having significantlygreater assets. Total resources were significantly higher for leveragedhouseholds in 1992, and categorically, leveraged households were moreheavily weighted to the higher percentile categories for total resources. Whileleveraged households generally had greater total resources, bequest expecta-tions were more likely to be higher among unleveraged households.

The first hypothesis, that leveraged households would have greaterassets than unleveraged households, was not supported and the results wereopposite of what was hypothesized. Hypothesis two, that leveraged house-holds would experience greater increases in assets than unleveraged house-holds, was also not supported by the results. Similarly, the third hypothesis,that leveraged households would have a higher allocation of assets to riskyassets than unleveraged households, was not supported. The fourthhypothesis, that leveraged households would have higher household incomethan unleveraged households, was supported by the results.



Discussion of Results

Leveraged and unleveraged households were significantly differentfrom each other in several aspects in 1992 and 2002. A key distinguishingfactor of the two groups was whether they were working in 1992 and 2002 andtheir resulting human capital. In 1992, leveraged households had higherincome, health, and total resources. Consistent with Grossman’s (1973)findings regarding health and work, a larger proportion of the leveragedhouseholds were working in 1992 and 2002 and also reported higher levels ofself-rated health. This contributed to the higher amount of total resourcesamong leveraged households.

The leveraged households in 1992 and 2002 were also significantlyyounger, which placed them closer to their peak earning years; they also hadlarger households. The younger age and larger household size of theleveraged households are consistent with the life cycle income hypothesis, aswell as the findings of Hanna and Rha (2000) and Chen and Jensen (1985).The leveraged households, as a result of their larger initial total resources andhuman capital, experienced a much larger reduction in total resources thanunleveraged households. However, the more abundant human capital amongthe leveraged households provided them with resources which could beconverted to financial capital. It does appear that households that movedfrom being leveraged to unleveraged were effective in converting some of thehuman capital to other forms of capital.

In 1992, leveraged households were more likely to expect to receivean inheritance while unleveraged households were more likely to expect toleave a sizable estate. Looking back over the period of observation in 2002and categorizing the households based on leverage status in 2002, there wasno statistically significant difference between the two groups regarding thereceipt of an inheritance or the household’s expectation to leave a sizableestate.

One possible explanation of the lack of difference between the twogroups in 2002 in regard to receiving an inheritance could be that inheritancemonies were used to pay off housing debt. Some households holdingmortgages in 1992 were perhaps borrowing against the expected proceeds ofan anticipated inheritance. Such households, after receiving the anticipatedinheritance, then eliminated their debt holdings. These households, in theabsence of the expected inheritance, may have chosen not to have mortgagesduring the period of observation. Similarly, the anticipated inheritance mayalso have served as the expected bequest among this group of households.Further research would be necessary to determine whether this was the case.

Unleveraged households were also more likely to expect to leave asizable inheritance to their heirs. The unleveraged house represents a largenon-liquid asset that can be bequeathed to heirs during life or upon death.


Research &

Theory

This finding was consistent with Kao and colleagues’ (1997) finding thathouseholds with non-liquid assets were more likely to expect to leave aninheritance.

The results were also consistent with Maki’s (1996) findings thatmore educated higher earners were more likely to have mortgages because ofincome tax advantages. In this study, leveraged households had substantiallyhigher household income and education than unleveraged households.Higher income generally results in higher tax rates for an individual; thus, thededuction of mortgage interest on personal income taxes would also be ofgreatest benefit to those individuals with the highest tax rates.

The statistically significant difference in consumer debt in 1992 and2002 may suggest a greater preference for current period consumption amongleveraged households. In 1992 and 2002 leveraged households had moreconsumer debt than unleveraged households. Greenspan’s (2003) delineationof the uses of extracted home equity indicates that a large amount of mortgageproceeds is used for current consumption.

The significant difference in total resources in 1992 and the lack ofdifference in total resources in 2002, in combination with the assets andchange in assets over the time period, may suggest that as the householdsage and the potential for earned income decreases, households transition to amore conservative financial position of being unleveraged. The ability toleverage earnings, not simply assets, appears to be a dominant factor whencomparing leveraged and unleveraged households. Furthermore, the substan-tially larger total resources among the leveraged households in 1992 appearsto have been transformed primarily into consumption rather than assets sincechange in assets were not different for leveraged and unleveraged house-holds and total assets remained greater for unleveraged households. Furtherresearch is necessary in order to make firmer conclusions.

Leveraged households at the beginning of the period had substan-tially more total resources than unleveraged households which overshadowedthe differences in assets. However, by the end of the period, leveraged andunleveraged households had approximately the same amount of total re-sources, yet unleveraged households continued to have more assets. Oneexplanation could be the type of income sources available to the household.If contractual and fixed amount sources of income, such as Social Security, VAbenefits, and defined benefit pension benefits were the primary resourcesthen the household may have been more willing to leverage those resourcessince there was less uncertainty with respect to income. However, if thehousehold had relatively few of these contractual sources of income, andincome was primarily derived from investments, the certainty of the incomewould have been much less and the household may have been much lesswilling to leverage those assets.



In light of all of the differences taken together, the major underlyingdifference between leveraged and unleveraged households appears to bework status and human capital of the household. Those households stillworking were more inclined to have mortgages than those households whichwere not working, or stopped working, and have relatively lower amounts ofhuman capital and total resources. By the end of the period of observation,with total resources approximately equal and unleveraged households havinggreater assets, leveraged working households became obligated to continueworking in order to maintain their total resources at or near those ofunleveraged households.

Implications

Some key implications for consumers and financial professionalsworking with clients can be drawn from the results. Consumers or profession-als working with consumers nearing retirement can have some confidence thateliminating mortgages appears to be a positive way of building assets andpreserving total resources. Households appear to be ineffective in usingleverage to achieve greater asset gains; as a result, the increased risk ofleverage may be unwise. The individual’s specific situation with respect totypes of income sources during retirement and comfort with debt shoulddictate how mortgages are used. Higher income households are more likely tobe leveraged in order to take advantage of the larger potential tax savingsfrom the interest deduction. The effectiveness of this practice in accumulat-ing assets is questionable based on the comparison of leveraged andunleveraged households. From this analysis, it does not appear that less riskaverse individuals are more or less likely to be leveraged.

Based on this analysis it does not appear that leveraged householdsin general are able to accumulate more assets than unleveraged households.Therefore, financial professionals should refrain from making general recom-mendations, such as in books, popular press literature, on radio and TV talkshows, that would encourage the average household to keep mortgagesrather than eliminate them. As the prevalence of defined benefit plansdecreases and income during retirement is increasingly dependent uponinvestment returns of the individual’s portfolio, a more prudent recommenda-tion to general audiences may be to eliminate mortgages prior to retirement.

Caution should be used when applying the results of empiricalstudies to specific individuals. Consequently, consumers and financialprofessionals working with these results should carefully evaluate the client’srisk tolerance and capacity to successfully leverage their portfolios withmortgages; a decision should then be made based on specific analysis of thesituation. As with the results of any empirical research, exceptions exist.However, the decision to use mortgages for investment purposes should becarefully analyzed.


Research &

Theory

Consumers and financial professionals working with clients shouldalso consider how much of the borrowed funds would be used for investmentpurposes, rather than consumption, and how those funds would be invested.The most appropriate expected rates of return for comparison would be theindividual’s actual investment returns, based on their asset allocation mix.Hypothetical returns on portfolios not currently utilized by the individualshould not be used in comparing alternatives.

While not included in this study, some implications may be drawnrelating to the Baby Boom generation. First, Baby Boomers may be morecomfortable with the responsibility of managing their own assets in a 401(k)plan and consistent with Engen and Gale’s (1997) findings, may leverage their401(k) accounts with mortgages. Second, given increasing life expectancies,the concept of retirement continues to change, particularly for Baby Boomerswho have time to plan and make arrangements for self-defined retirement.Thus, historical work patterns may no longer be relevant to the Baby Boomersand the rapid decrease in human capital, as measured by the present value offuture earnings, may not be as pronounced in their cohort. Working house-holds in this study were more likely to carry mortgages and if Baby Boomersadapt a retirement concept that includes some work, larger mortgages maymore frequently be carried by older adults.

References

Ando, A., & Modigliani, F. (1963). The “life cycle” hypothesis of saving:Aggregate implications and tests. The American Economic Review,53(1), 55–84.

Adams, P., Hurd, M. D., McFadden, D., Merrill, A., & Ribeiro, T. (2003).Healthy, wealthy, and wise? Tests for direct causal paths betweenhealth and socioeconomic status. Journal of Econometrics, 112(1),3–56.

Chen, A., & Jensen, H. H. (1985). Home equity use and the life cyclehypothesis. The Journal of Consumer Affairs, 19(1), 37–56.

Clements, J. (2004, March 31). Maybe investors aren’t stupid after all: Oft-cited performance study is revised. The Wall Street Journal, D1.

Coy, P., & Keenan, F. (2003). Consumer debt takes a holiday: Buyers are wary–but helped by refinancing. Business Week, 3822, 40.

Crook, J. (2001). The demand for household debt in the U.S.A.: Evidence fromthe 1995 Survey of Consumer Finance. Applied FinancialEconomics, 11, 83–91.

Dalbar, Inc. (2001, June). Dalbar issues 2001 update to “Quantitativeanalysis of investor behavior” report [Electronic version].Toronto:Author.



Dunsky, R. M., & Follain, J. R. (2000). Tax-induced portfolio reshuffling: Thecase of the mortgage interest deduction. Real Estate Economics,28(4), 683–718.

Edelman, R. (2001). Financial security in troubled times. New York: HarperCollins.

Engen, E. M., & Gale, W. G. (1997). Debt, taxes, and the effects of 401(k) planson household wealth accumulation. Retrieved May 15, 2004, fromhttp://papers. ssrn.com/sol3/papers.cfm?abstract_id=40180.

Goff, D. C., & Cox, D. R. (1998). 15-year versus 30-year mortgage: Which is thebetter option? Journal of Financial Planning, 11(2), 88–94.

Grable, J. E., & Lytton, R. H. (1998). Investor risk tolerance: Testing theefficacy of demographics as differentiation and classifying factors.Financial Counseling and Planning, 9(1), 61–73.

Greenspan, A. (2003, March 4). Home mortgage market. Speech given to theIndependent Community Bankers of America 2003 AnnualConvention. Retrieved December 30, 2003, from http://www.federalreserve.gov/boarddocs/speeches/2003/20030304/default.htm.

Grossman, M. (1973). The correlation between health and schooling, Centerfor Economic Analysis of Human Behavior and Social InstitutionsWorking Paper No. 22. New York: National Bureau of EconomicResearch, Inc.

Gutter, M. S. (2000). Human wealth and financial asset ownership. FinancialCounseling and Planning, 11(2), 9–20.

Hanna, S., Fan, J., & Chang, Y. R. (1995). Optimal life cycle savings. FinancialCounseling and Planning, 6, 1–15.

Hanna, S., & Rha J.-Y. (2000). The effect of household size changes on credituse: An expected utility approach. Consumer Interest Annual, 46.Retrieved April 12, 2004, from http://www.consumerinterests.org/public/articles/index.html?cat=250

Ibbotson Associates. (2002). Stocks, bonds, bills, and inflation 2002yearbook. Chicago: Author.

Johnston, M. A. (2000). Home mortgages can be more than just debt. Manage,52(1), 6–7.

Kao, Y. E., Hong, G.-S., & Widdows, R. (1997). Bequest expectations: Evidencefrom the 1989 Survey of Consumer Finances. Journal of Family andEconomic Issues, 18(4), 333–356.

Kennickell, A. B., & Starr-McCluer, M. (1997, April). Household saving andportfolio change: Evidence from the 1983–1989 SCF panel (FederalReserve Board of Governors. Retrieved March 5, 2004, from http://www. federalreserve.gov/ pubs/oss/oss2/papers/p8389.pdf.


Research &

Theory

Kotlikoff, L. J., & Summers, L. H. (1981). The role of intergenerational transfersin aggregate capital accumulation. Journal of Political Economy,89(4), 706–732.

Maki, D. (1996). Portfolio shuffling and tax reform. National Tax Journal,49(3), 317–329.

Meer, J., Miller, D. L., & Rosen, H. S. (2003). Exploring the health-wealth nexus.Journal of Health Economics, 22(5), 713–730.

Modigliani, F. (1986). Life cycle, individual thrift and the wealth of nations.The American Economic Review, 76(3), 297–313.

Modigliani, F., & Miller, M. H. (1958). The cost of capital, corporation financeand the theory of investment. The American Economic Review,48(3), 261–297.

Myers, S. C. (2001). Capital structure. Journal of Economic Perspectives,15(2), 81–102.

Orman, S. (n.d.). The sane retirement plan: Change the prescription of yourretirement glasses. In Yahoo: Money Matters. Retrieved April 12,2004, from http://biz. yahoo.com/pfg/ e04retire/art021.html

Palmer, L. (2002). The mortgage prepayment decision: A different perspective.Proceedings of the Association for Financial Counseling andPlanning Education, Scottsdale, AZ, 123–125.

Salandro, D., & Harrison, W. B. (1997). Determinants of the demand for homeequity credit lines. The Journal of Consumer Affairs, 31(2), 326–345.

Schooley, D. K., & Worden, D. D. (1996). Risk aversion measures: Comparingattitudes and asset allocation. Financial Services Review, 5(2), 87–99.

Spencer, H. L., & Fan, J. X. (2002). Savers, debtors, and simultaneous debtorsand savers. Financial Counseling and Planning, 13(2), 25–38.

Stango, V. (1999). The tax reform act of 1986 and the composition of consumerdebt. National Tax Journal, 52(4), 717–739.

Storms, P. (1996). Using mortgage credit to achieve client objectives. Journalof Financial Planning, 9(5), 77–86.

Storms, P. (2000). Mortgage mythology. Journal of Financial Planning,13(4), 46–48.

Sullivan, T. A., Warren, E., & Westbrook, J. L. (2000). The fragile middle class:Americans in debt. New Haven: Yale University Press.

Sung, J., & Hanna, S. (1996). Factors related to risk tolerance. FinancialCounseling and Planning, 7, 11–20.

Tomlinson, J. A. (2002). Advising investment clients about mortgage debt.Journal of Financial Planning, 15(6), 100–108.

United States Department of the Treasury. (1996). 1996 Individual income taxreturns (IRS Publications No. 1304). Washington, DC: Author.Retrieved May 21, 2001, from http://www.irs.gov/taxstats/display/0,,i1%3D40%26genericId% 3D16883,00.html.



Waggle, D., & Johnson, D. T. (2003). The impact of the single-family home onportfolio decisions. Financial Services Review, 12(3), 201–218.

Wang, H., & Hanna, S. (1997). Does risk tolerance increase with age?Financial Counseling and Planning, 8(2), 27–31.

Zhu, L. Y., & Meeks, C. B. (1994). Effects of low income families’ ability andwillingness to use consumer credit on subsequent outstanding creditbalances. The Journal of Consumer Affairs, 28(2), 403–422.

Contact Information: Lance Palmer, Ph.D., University of Georgia, Departmentof Housing and Consumer Economics, 111 Dawson Hall, Athens, GA 30602-2622; Phone: (706) 542-4916; Fax: (706) 542-4397; E-mail: [email protected] M. Lown, Ph.D., Utah State University, Family, Consumer, & HumanDevelopment, 2905 Old Main HillLogan, UT 84322-2905; Phone: (435) 797-1569; Fax: (435) 797-3845; E-mail:[email protected]


Research &

Theory

THE EFFECTS OF INVESTMENT EDUCATION ON

GENDER DIFFERENCES IN FINANCIAL KNOWLEDGE

Ronald E. GoldsmithFlorida State University

&Elizabeth B. GoldsmithFlorida State University

ABSTRACT

There is much concern that women in particular have less knowledgeand skill in managing their personal finances than men do, despite itsmanifest importance and the fact that so many women have thisresponsibility. These gender differences appear early in life: previousstudies have suggested that male college students not only feel thatthey have more knowledge of personal finance than female students,but that in fact they really do have more knowledge on average. Oneavenue for improving financial skills is through education. Littleresearch, however, has been done to evaluate the effects of educationon the financial skills of women versus men. Using data from 122students, the present study confirmed that women knew less aboutinvestments than did men. However, the experiment showed that formalinstruction in personal finance not only increased both men’s andwomen’s investment knowledge, it closed the knowledge gap betweenthe genders.

Introduction

Money plays a central role in our lives, exerting more power than anyother commodity (Furnham, 1998; Oleson, 2004). With levels of bankruptcyand credit card debt reaching all time highs in the United States, it is apparentthat there is a need for more financial literacy and an examination of whicheducational strategies work best (Hogarth, 2002). Financial literacy can bedefined in many ways (Hogarth, 2002), but for the purposes of the presentstudy it was taken to mean being knowledgeable about managing money andmaking sound financial decisions. Government outreach programs such asCredit Abuse Resistance Education (CARE, www.careprogram.us) and the



U.S. Treasury Department’s financial literacy education program addressthese issues. Similar concerns are apparent in other countries including theUnited Kingdom (Brennan & Ritters, 2004). The emphasis in the presentstudy was on investing as one integral and important part of financial literacy(Bell & Lerman, 2005; Hilgert, Hogarth, & Beverly, 2003; Lyons et al., 2005).

Because young people are especially important in this regard, severalresearch studies have examined the role of credit and financial attitudes incollege students’ lives (e.g., Pinto, Marsfield, & Parente, 2004; Yang, Spinella,& Lester, 2004). Studies of personal finance attitudes and behaviors amongyoung people reveal an important sex difference. Women feel that they knowless than men do about financial topics such as managing money (Hira &Mugenda, 2000), financial analysis (Webster & Ellis, 1996), and investing(Goldsmith & Goldsmith, 1997). Although there is evidence that womenemploy more prudent investing strategies than men do (Barber & Odean,2001), their lack of knowledge and confidence likely contribute to theirdissatisfaction with their financial situations (Hira & Mugenda, 2000).

One avenue to improving both women’s and men’s financial abilitiesis formal financial education (Lyons et al., 2005). Improved financial knowl-edge is assumed to lead to beneficial financial decision making (Bell &Lerman, 2005; Hilgert et al., 2003). Although “most empirical studies supportthe view that financial education positively affects financial behavior andfinancial outcomes” (Lyons et al., 2003, p. 1), there is little academic researchattesting to the efficacy of this approach to improving financial knowledge.Especially lacking are studies demonstrating changes in behavior associatedwith increased knowledge (Hogarth, 2002) as well as experimental studiesaimed at verifying cause and effect, so that longitudinal (pre- and post-testmethods) designs are needed (Fox, Bartholomae & Lee, 2005; Lyons et al.,2003). The present study did not address the former issue, but does directlycontribute to the latter concern. Thus, the purposes of the study were (1) toattempt to replicate previous findings regarding differences between male andfemale students’ self-assessments of their level of investment knowledge and(2) to test the proposition that formal education in personal finance couldincrease financial literacy, and especially close the knowledge gap betweenmen and women.

Literature Review

One aspect of financial literacy is whether there is a gender effect ornot: do women and men behave and feel differently about money? Theevidence suggests that they do. For example, Hira and Mugenda (2000) in anadult study found that there are differences in the way men and womenperceive financial issues, and women were more likely than men to be dissatis-fied with their current financial situation. Women were less satisfied than men


Research &

Theory

about their savings level, yet at the same time women reported spending morethan they should. Women admitted that spending habits were creating chaosin their lives. Hira and Mugenda (2000) concluded that gender is a socialphenomenon and that educational programs need to focus on genderdifferences in behaviors and their consequences. Others confirm thisconclusion because research indicates that financial behaviours are notnecessarily influenced by economic factors only, but reflect social andpsychological issues (Faber, 1992). Webster and Ellis (1996) showed thatwomen were less confident in their ability to perform financial analysis thanmen. Goldsmith and Goldsmith (1997) and Goldsmith et al. (1997) surveyedcollege students and showed that the men students felt they knew more aboutinvesting and actually did know more about investing than the womenstudents did. Barber and Odean (2001) used actual trading data to show thatin the area of finance, men were more overconfident than women, leading themto trade more often (thereby reducing their net returns). Batjelsmit andBernasek (1996) reported that research usually shows that women invest moreconservatively than men.

A second aspect of the concern for financial literacy involves waysto increase it, especially among women (Goldsmith, 2005). Lown (2004)conducted eight focus groups of baby boomer-age women and found thatsome women preferred attending a group activity involving learning aboutfinancial planning for retirement whereas others preferred to learn individually.She found that younger women preferred independent learning and lessonsover the Internet vs. older women who preferred class or seminar settings.These findings indicate that a woman’s age might be a determinant in howfinancial information can be put across most effectively. In an earlier study,DeVaney (1995) found that older and younger boomers are truly differentcohorts and thus, age or life stage should be considered when designingeducational programs and approaches.

Academic researchers are not the only ones interested in howwomen respond to financial information or in ascertaining how much theyknow. Investment firms, financial planners, and governments here and abroadhave expressed growing interest in improving consumer literacy in the area ofpersonal finance. One way of achieving this is through programs of consumereducation: “Consumer education empowers people to use informationeffectively and to be more confident about the decisions they make whenpurchasing and/or using goods and services. There is discussion in theliterature of the importance of both information and the skills required toconvert that information into effective knowledge” (Brennan & Ritters, 2004,p. 106). One example of interest in consumer education is illustrated by Xiao,O’Neill, Prochaska, Kerbel, Brennan, and Bristow (2004), who describe asuccessful consumer education program focused on savings and debtreduction. Another example is the Wi$eUp program of the U. S. Dept. of Labor



which provides financial information to GEN X women ages 22 - 35. TheWi$eUp program answers questions that GEN X women send in to the Dept.of Labor website and sponsors programs and teleconferences where youngwomen call in questions once a month to financial experts. The transcripts ofthese monthly question and answer sessions are available online. TheWi$eUp Program is just one among many financial and employment programsthat the Dept. of Labor offers to women of all ages under the umbrella of theWomen’s Bureau.

Thus, there is considerable interest in many different realms aboutgetting personal financial information to women and related to this is theconcern about how effective education is in improving financial knowledge.There is a correlation, but the direction of causation is not clear (Bell &Lerman, 2005; Hilgert et al., 2003). However, there is little direct researchverifying the effectiveness of investment education (Lyons et al., 2003; Medill,2003). Our study was motivated by the need to replicate previous studies ofinvestment knowledge among college students and by curiosity about theeffectiveness of formal instruction in improving this knowledge, especiallyamong women students.

Method

The study was a quasi-experiment using a non-equivalent controlgroup design (Huck & Cormier, 1996, p. 615) because subjects could not berandomly assigned to the experimental groups. Two undergraduate classes ata large southeastern U.S. university participated. Students are appropriatesubjects because this age group is an important target of financial literacy(Braunstein & Welch, 2002). The first group or treatment group was a class inthe College of Human Sciences entitled Family Financial Analysis. Thesestudents (14 men; 57 women) received formal instruction in personal finance,including investing. It was expected that the subject material taught in thiscourse would increase both their subjective and real levels of knowledge.Whether this instruction could close the gap between the men and womenwas the major research question of the study. The second group was amarketing class in the business school (18 men; 33 women). These studentswere chosen on a convenience basis to provide the control group. Owing torestrictions on registration it was highly unlikely that students in the controlgroup had either taken or were taking the class in personal finance class, sothey were not exposed to the same information as the students in the treat-ment group.

Both the treatment and control groups were administered theidentical questionnaire assessing their self-reported levels of investmentknowledge (subjective knowledge) and a test of their actual investmentknowledge (real knowledge) at the start of the semester. The control group


Research &

Theory

received no special instruction in personal finance, and fifteen weeks later, thesame questionnaire was again administered to both groups. Using the samequestionnaire rules out instrumentation error (changes or differences in themeasures used) as a rival explanation of the results (Campbell & Stanley, 1963,p. 9). The marketing class acted as a control to the experimental treatmentgroup to remove the rival explanations of history (events in the externalenvironment beyond the experimenters control) and maturation (the experi-mental subjects maturing or changing overtime) (Campbell & Stanley, pp. 7-8).

There is no agreement in the field of financial education about whatmeasures should be used to assess financial knowledge and there are nouniversally accepted, standardized measures (Lyons et al., 2005). Bothsubjective and objective measures are common (Hilgert et al., 2003; Hogarth,2002). Consequently, the present study measured both subjective andobjective knowledge. The questionnaires contained six-item measures ofsubjective and real investment knowledge (see Appendix) identical to thatused by Goldsmith and Goldsmith (1997) and Goldsmith et al. (1997). Theformer was chosen because it has been shown to be a reliable, valid measureof subjective knowledge in consumer research (Flynn & Goldsmith, 1999), andthe latter, which is comparable to other objective knowledge tests (see Hilgertet al., 2003) in order to make the results comparable to previous studies.

When the subjective knowledge items from the first administrationwere factor analyzed using the principal axis factor method, the resultsshowed that a single factor accounted for the majority of the variance (58.6%)in the item correlation matrix. The largest eigenvalue was 3.52, with thesecond eigenvalue being only .879, suggesting the scale was unidimensional.All the item loadings on the factor were greater than .3. The scale had goodinternal consistency (coefficient alpha = .85) and so the items were summed toform a subjective knowledge score for each participant. When the same scalewas administered the second time it again proved to be unidimensional withalpha of .91. The summed scores on this measure of subjective investmentknowledge could range from 6 to 30 with a theoretical midpoint of 18.

Six multiple-choice questions measured real investment knowledge.These were the same as those used by Goldsmith and Goldsmith (1997) andGoldsmith et al. (1997). The number of correct answers served to measure realinvestment knowledge. These scores could range from 0 to 6.

Other questions asked how many credit cards the students pos-sessed, how much they owed on their cards, and what their five-year planswere for investing (opening a savings account, buying stocks, and buyingbonds). Comparisons using t-tests showed that women owned more cardsand carried more debt than men (consistent with Hira & Mugenda, 2000), butthe two groups were not statistically different in their future investment plans.



Results

The class and gender group means on the measures of subjectiveand real knowledge served as the dependent variables instead of individualstudent scores (see Tables 1 and 2 and Figures 1 and 2) because it wasimportant to protect subject identity on this sensitive topic. Thus, individualsubjects were not identified, and a repeated measures analysis of variancecould not be used. T-tests were used to assess the statistical significance ofspecific comparisons.1

Table 1Mean Score on Subjective and Real Knowledge (BusinessStudents)

Subjective Knowledge Real Knowledgen Mean SD n Mean SD

First MeasureMen 18 16.33 4.98 18 3.56 1.46Women 33 14.55 4.00 33 2.91 1.47Total 51 15.18 4.41 51 3.14 1.48Second MeasureMen 15 17.20 5.10 15 3.40 1.23Women 25 13.56 4.70 25 2.84 1.72Total 40 14.93 5.11 40 3.05 1.57

Table 2Mean Score on Subjective and Real Knowledge (Human Sci-ences Students)

Subjective Knowledge Real Knowledgen Mean SD n Mean SD

First MeasureMen 14 16.64 4.18 14 2.93 1.14Women 57 12.44 4.01 58 2.03 1.11Total 71 13.27 4.35 72 2.21 1.16Second MeasureMen 8 20.38 3.42 8 5.38 0.92Women 41 21.24 3.66 39 4.97 0.81Total 49 21.10 3.60 47 5.04 0.83

For the first time the measures were taken, the mean subjectiveknowledge score of all the 32 men in both classes (M = 16.5, SD = 4.6) wassignificantly higher than the mean subjective knowledge score of all the 90women in both classes (M = 13.2, SD = 4.1; t = 3.7, p < .001). This result is


Research &

Theory

consistent with previous studies suggesting that men are more confident intheir financial knowledge and ability than women are (Barber & Odean, 2001)and specifically replicates findings of Goldsmith and Goldsmith (1997) and ofGoldsmith et al. (1997) for students.2

When real investing knowledge was measured the first time, all the32 men in both classes scored an average of 3.3 (SD = 1.4) (out of a perfectscore of 6) versus a mean score for all the 91 women of 2.4 (SD = 1.3). Thisdifference was statistically significant (t = 3.4, p = .001), again replicatingfindings of previous studies (Goldsmith & Goldsmith, 1997; Goldsmith et al.1997).3 Although the men scored higher than the women on the test of realknowledge, neither their average score of 55% nor the women’s average of40% suggest adequate levels of financial literacy, but are similar to perfor-mance on measures of financial literacy by these age groups reported in otherstudies (see Braunstein & Welch, 2002, p. 453; Hogarth, 2002, p. 18).

In Figures 1 and 2, the focus is not on the gender differences or classdifferences, but on the change in the means of the Human Sciences studentscompared with the business students. It is evident from examining the meanscores that the subjective knowledge of the marketing students (the controlgroup) did not improve dramatically over the course of the fifteen-weeksemester. They were given no formal instruction in investing or personalfinance, so this was expected. The means of the Human Sciences students,however, show a dramatic increase in subjective knowledge after completionof the Family Financial Analysis course, and this increase was evident forboth men and women. It makes sense that these students felt they knew moreabout investing after taking a course containing formal instruction in thissubject matter. After taking a course in Family Financial Analysis Finance,both men (M = 20.4, SD = 3.4) and women (M = 21.2, SD = 3.7) felt that theyknew more about investing (see Figure 1), and the original gap in male/femalescores virtually disappeared. Comparison with the control group rules out theeffects of history and maturation on these changes (Campbell & Stanley,1963). The most parsimonious conclusion is that the content of the FamilyFinancial Analysis class accounted for the change in subjective knowledge.



Estimated Marginal Means of KNOW

GROUP

PF 2PF 1MAR 2MAR 1

Est

imat

ed M

argi

nal M

eans

22

20

18

16

14

12

10

SEX

male

female

Note: MAR 1 = first data collection for Marketing StudentsMAR 2 = second data collection for Marketing StudentsPF 1 = first data collection for Personal Finance StudentsPF 2 = second data collection for Personal Finance Students

Figure 1: Subjective Knowledge as a Function of Gender andClass

Likewise, when the mean real knowledge scores were examined, asimilar pattern appeared (see Figure 2). The real investment knowledge of thebusiness students appeared not to change, but the Family Financial Analysisstudents showed a dramatic increase in real knowledge, mirroring their changein subjective knowledge. Both the men (M = 5.4, SD = .9) and women (M =5.0, SD = .8) experienced substantial increases in real investment knowledge(90% and 83%), and the male/female difference narrowed, as the difference inthese means is not statistically significant.


Research &

Theory

Estimated Marginal Means of REAL

GROUP

PF 2PF 1MAR 2MAR 1

Est

imat

ed M

argi

nal M

eans

6

5

4

3

2

1

SEX

male

female

Note: MAR 1 = first data collection for Marketing StudentsMAR 2 = second data collection for Marketing StudentsPF 1 = first data collection for Personal Finance StudentsPF 2 = second data collection for Personal Finance Students

Figure 2. Real Knowledge as a Function of Gender and Class



Discussion

The study replicated previous studies’ findings that men are moreconfident about their knowledge of investing than women are and they alsoappear to know more about investing. Moreover, the experiment showed thatconsumer education improved both men’s and women’s subjective and realinvestment knowledge. While these findings are not surprising, they areencouraging for those concerned with the important public policy issues ofconsumer education and women and finance. They suggest that deficienciesin women’s subjective and real knowledge of investing can be overcome byconsumer education, and in general, that such education potentially benefitsall consumers in this aspect of their lives. This coincides well with the Lown(1994) conclusion that both the financial planning industry and women wouldbenefit from more effective educational outreach geared to women.

Researchers need to be cautious when interpreting these findingsbecause the study, however, has many limitations. The first limitation of thestudy has to do with the fact that the courses were heavily weighted towardswomen. This reflects the reality of enrollments in these colleges, which areboth majority female and the fact that random assignment of subjects totreatments was not possible. Thus, there are likely difference in the personali-ties, abilities, and interests of both the women and the men who enrolled inthese courses and in the reasons for taking the courses. For example, thestudents in Human Sciences started the experiment with lower scores on themeasure of real investing knowledge than the business students’, so theywere not equivalent on the dependent variable. Unfortunately, it is notfeasible to randomly assign subjects to courses in a college setting, so thisresearch paradigm will always suffer from this inherent limitation. On theother hand, the gender differences revealed replicate those shown in severalother studies, the initial levels of the dependent variables were similar acrossthe classes, and one would be hard pressed to offer a pre-existing differencein the subjects that systematically improved investment knowledge of onlythe Human Sciences subjects, thereby accounting for the results. Neverthe-less, attempts should be made to overcome this difficulty in future studies.4

The second limitation comes from the fact that generalizability of theresults is limited by the convenience nature of the sample. Subjects wouldneed to be selected randomly from some population of interest to provideaccurate point and interval estimates of the dependent variables for thatpopulation. Moreover, since the subjects were college students, the resultsmight not generalize to different age and educational groups.

The third limitation lies in the subject mortality between the first andsecond data collections. Several students were unable or unwilling toparticipate in the second survey. These students, however, would have to beonly the poorly performing students in the Human Sciences class to account


Research &

Theory

for the increase in the overall mean of that group between time one and timetwo. The fourth limitation is the fact that the data were compared as pooled,group averages rather than as a repeated-measures analysis of individualscores. Ideally, subjects would be randomly assigned to the control andtreatment groups to create a true experiment. The fifth limitation lies in thespecific measures used; they are consistent with previous research on thistopic, but it is possible that other measures of real and subjective knowledgemight not yield the same results. Finally, it is not known whether differentinstruction in personal finance would produce the same results.

Another concern may be that investment education itself may havelittle influence on actual investment behaviour for either men or women(Medill, 2003, pp. 946-953). According to Braunstein and Welch (2002, p. 445),“improved financial behavior does not necessarily follow from increasedfinancial information.” Whether improving investment knowledge leads toimproved investment choices needs to be demonstrated, but this is outsidethe scope of the present study. Although men feel more confident (greatersubjective knowledge) than women about their financial abilities, this appar-ently leads them to trade more often than women, thereby reducing the netreturn on their portfolios (Barber & Odean, 2001). If this is the case, simplyimproving subjective knowledge may be counterproductive to the intendedoutcome of improving consumer financial well-being. Whether improved realknowledge leads to improved outcomes thus needs to be demonstrated (Foxet al., 2005).

Despite these limitations, the study makes a contribution to theoverall topic of women and finance. It can be concluded that male studentsboth feel they know more about investing than female students, and theyreally do know more than female students, but formal instruction in personalfinance not only enhances both men’s and women’s investment knowledge, itcloses the gap in this knowledge. Thus, the study successfully replicatedprevious findings regarding gender differences in financial knowledge andshowed the efficacy of formal education in improving it. The challenge lies indemonstrating that improving knowledge leads to improved investmentdecision and improved financial health for both men and women. An impor-tant concern in financial literacy is whether educational programs have a“differential impact” on men versus women (Braunstein & Welch, 2002, p.207). Our study suggests that, in so far as the subjective and objectivemeasures we used, the impact was largely similar in that both men and womenreported greater subjective and objective knowledge after taking the personalfinance course.



References

Barber, B. M., & Odean, T. (2001). Boys will be boys: Gender, overconfidence,and common stock investment. The Quarterly Journal of Economics,116 (1), 261-292.

Batjelsmit, V. L., & Bernasek, A. (1996). Why do women invest differently thanmen? Financial Counseling and Planning, 7, 1-9.

Bell, E., & Lerman, R. I. (2005). Can financial literacy enhance asset building?Opportunity and Ownership Project, The Urban Institute, 9 (6), 1-7.

Braunstein, S., & Welch, C. (2002). Financial literacy: An overview of practice,research, and policy. Federal Reserve Bulletin, November: 445-457.

Brennan, C., & Ritters, R. (2004). Consumer education in the UK: Newdevelopments in policy, strategy, and implementation. InternationalJournal of Consumer Studies, 28 (2), 97-107.

Campbell, D. T., & Stanley, J. C. (1963). Experimental and Quasi-experimentalDesign for Research. Chicago: Rand McNally.

DeVaney, S. (1995). Retirement preparation of older and younger boomers.Financial Counseling and Planning, 6, 25-33.

Faber, R. (1992). Compulsive Buying. American Behavioral Scientist, 35, 802-819.

Flynn, L. R., & Goldsmith, R. (1999). A short, reliable measure of subjectiveknowledge. Journal of Business Research, 46, 57-66.

Fox, J., Bartholomae, S., & Lee, J. (2005). Building the case for financialeducation. Journal of Consumer Affairs, 39 (1), 195-214.

Furnham, A. (1984). Many sides of the coin: The psychology of money usage.Personality and Individual Differences, 5, 501-509.

Goldsmith, E. B. (2005). Women and estate planning. Journal of PracticalEstate Planning, 7 (1), 25-28.

Goldsmith, E. B., & Goldsmith, R. E. (1997). Gender differences in perceivedand real knowledge of financial investments. Psychological Reports,80, 236-238.

Goldsmith, R. E., Goldsmith, E. B., & Heaney, J.-G. (1997b). Sex differences infinancial knowledge: a replication and extension. PsychologicalReports, 81, 1169-1170.

Hilgert, M. A., Hogarth, J. M., & Beverly, S. G. (2003). Household financialmanagement: The connection between knowledge and behaviour.Federal Reserve Bulletin, July: 309-322.

Hira, T., & Mugenda, O. (2000). Gender differences in financial perceptions,behaviors and satisfaction. Journal of Financial Planning, 13 (2),86-92.

Hogarth, J. M. (2002). Financial literacy and family & consumer sciences.Journal of Family and Consumer Sciences, 94 (1), 14-28.


Research &

Theory

Huck, S. W., & Cormier, W. H. (1996). Reading Statistics and Research (2nded.). New York: HarperCollins.

Lown, J. (2004). Women’s preferences for learning about financial planning.Journal of Personal Finance, 3 (4), 49-58.

Lyons, A. C., Hogarth, J., Schuchardt, J., Smith, T., & Toussaint-Comeau, M.(2003). Evaluating outcomes of personal financial education.Consumer Interests Annual, 49.

Lyons, A. C., Palmer, L., Jayaratne, K. S. U., & Scherpf, E. (2005). Are wemaking the grade? A national overview of financial education andprogram evaluation. University of Illinois Chicago. Working PaperNo. 012.

Medill, C. E. (2003). Challenging the four “truths” of personal social securityaccounts: evidence from the world of 401(K) plans. North CarolinaLaw Review, 81, 901-976.

Oleson, M. (2004). Exploring the relationship between money attitudes andMaslow’s hierarchy of needs. International Journal of ConsumerStudies, 28 (1), 83-92.

Pinto, M.B., Marsfield, P., & Parente, D. H. (2004). Relationship of creditattitude and debt to self-esteem and locus of control in college-ageconsumers. Psychological Reports, 94, 1405-1418.

Webster, R. L., & Ellis, T. S. (1996). Men’s and women’s self-confidence inperforming financial analysis. Psychological Reports, 79, 1251-1254.

Yang, B., Spinella, M., & Lester, D. (2004). Credit card use and prefrontal cortexdysfunction: a study in neuroeconomics. Psychological Reports,94, 1267-1268.

Xiao, J. J., O’Neill, B. O., Prochaska, J. M., Kerbel, C. M., Brennan, P., &Bristow, B. J. (2004). A consumer education programme based on thetranstheoretical model of change. International Journal ofConsumer Studies, 28 (1), 55-65.

AppendixQuestionnaire Items used to Measure Subjective and Real

Knowledge

Subjective Knowledge

SD D N A SAI know pretty much about investing. 1 2 3 4 5I do not feel very knowledgeable about investing. 1 2 3 4 5Among my circle of friends, I’m one of the“experts” on investments 1 2 3 4 5Compared to most other people, I know lessabout investing. 1 2 3 4 5



I have heard of most of the new investmentsthat are around. 1 2 3 4 5When it comes to investing, I really don’tknow a lot. 1 2 3 4 5

Yes, This is a Test!Please circle the correct response.

A bond coupon gives the holder an option to buydiscount bonds. True False Don’t know

Which of the following has the highest risk: a. US Government bondsb. Preferred stockc. Cashd. Don’t know

What is the 20 year average return on stocks? a. 9.7%b. 12%c. 23.4%d. Don’t know

Interest earned on savings accounts is tax-free. True False Don’t know

When interest rates rise, bond prices: a. riseb. fallc. no changed. Don’t know

Today, cash is backed by: a. treasury bondsb. faithc. goldd. Don’t know


Research &

Theory

1 Age and the number of credit cards possessed were initially used ascovariates in the analysis, but they had no effects on the group differences,and so analyses were performed without covariates, and the means reportedare unadjusted.

2 For the first time the measure was taken, the mean subjective knowledgescore for the 51 men and women business students was 15.18 (SD = 4.41) andthe mean subjective knowledge score for the 71 men and women in personalfinance was 13.27 (SD = 4.35). The difference, however, was not due to thedifference in the men’s scores but in the difference between the women’sscores and the fact that the personal finance class had a higher percentage ofwomen than did the business class (see Table 1). Business students might beexpected to feel they were more knowledgeable than other students aboutinvesting owing to their possible exposure to this information in their classes.The statistical significance of the difference in these means was not testedbecause it has no theoretical or practical significant, we just point it out to bethorough.

3 When the real investing knowledge of the business students of bothgenders was compared with the real investing knowledge of the HumanSciences students of both genders, the business students’ mean (3.14, sd =1.48) exceeded that of the Human Sciences students (M = 2.21, sd = 1.16).Although this again betrays the larger proportion of men business students, itmight be expected because it is possible that business students might be moreexposed to investing information in their classes than other students are. Theinterest of the present study was in the results of formal instruction on thesemeasures as shown by the comparisons between the genders and classesafter the Human Sciences students completed their course.

4 The authors want to thank an anonymous reviewer for suggesting thislimitation of the study.

Contact Information: Dr. Ronald E. Goldsmith, College of Business, FloridaState University, Tallahassee, FL 32306; Phone: (850) 644-4401; Fax: (850) 644-4098; E-mail: [email protected]



ABSTRACT

This study examines whether it is feasible that consensus marketforecasts, such as those published annually by Business Week canprovide information useful for asset allocation decisions. The findingssuggest that reliance on forecasts of the Business Week experts or anygroup of experts is highly risky because a) experts, similarly to laymen,demonstrate a proclivity to engage in the anchoring and adjustmentheuristic; b) experts must perform at a level approximately 20% aboverandom simply to match a buy-and-hold strategy; and c) where taxes area factor, experts must perform at an even higher level to compensate forloss of compounding benefits and loss of preferred tax rates applicableto stock investments.

CONSENSUS MARKET FORECASTS:TOO RISKY TO RELY ON EXPERTS

Seth Hammer, CPA, Ph.D.Towson University

Ora Freedman, Ph.D.Villa Julie College

Introduction

A key issue for practitioners and their clients is whether or not toadjust investment portfolios based on opinions of experts. If, for example, anexpert predicts that the stock market in the next year will provide significantlylower than average returns, should investors reduce their stock holdings andincrease their holdings of bonds and/or bills? In cases where the opinion ofone or a small group of experts may not inspire sufficient confidence, would itmake sense to adjust one’s portfolio based on the consensus opinion of alarge and broad set of experts?

Usage of consensus predictions allows for the avoidance of whathas been described as the “cross-sectional problem” (Taler, 2001), the problemof not knowing whether to attribute an outcome or set of outcomes to skill orchance. This cross-sectional problem can arise, for example, in attempting todetermine whether the individual who was most successful in predicting lastyear’s market results achieved that success through skill, luck, or a combina-tion of the two.


Research &

Theory

Consensus predictions, when generated by experts, can potentiallyprovide significant benefits, as was demonstrated in the case of forecasts offuture gross domestic gross product contained in the Federal Reserve Board’sBlue Chip Economic Indicator Survey. The study found that no individualforecaster, over time, provided better average scores than the consensusforecast (Bauer, Eisenbeis, & Waggoner, 2003).1 In other cases, however, forexample the Real Estate Research Corporation’s (RERC) quarterly consensuspredictions for most to least desirable real estate category investments, thecorrelation between performance and RERC consensus expert predictions wasactually negative (Ling 2005).

This study provides a multi-faceted examination of the potentialbenefits/risks of adjusting investment portfolios based on consensus expertpredictions of the following year’s market closing prices. While focusinglargely on the consensus predictions contained in Business Week’s annualyear-end investing issues, its emphasis is the broader issue of whether it isfeasible to expect that any set of consensus expert market forecasts, notsimply Business Week’s, can provide information useful for investors.Business Week was selected as the central model for review because a) theirprognosticators are well established and regarded in their field; b) thenumbers of prognosticators were sufficiently large each year to provide abroad sample; and c) the magazine’s weekly circulation, almost 1 million andthe largest of any business magazine, gives it the power to influence thedecisions of a very substantial number of investors. The Dow Jones Indus-trial Average was chosen because of its high visibility to the investing public.

In order to determine whether Business Week or other consensuspredictions of future market results have and/or are likely to generate usefulinformation, five issues were considered:

1. Have the Business Week experts demonstrated an ability to predictfollowing year-end market results for the Dow Jones Industrial Average?2. Can a simple formula based on prior performance match or exceedthat of the Business Week experts?3. Have the Business Week experts demonstrated an ability to avoid aproclivity to engage in the heuristic referred to as “anchoring andadjustment,” whereby initial judgments serve as an anchor for insuffi-ciently adjusted future judgments?4. How much better than random would expert forecasts have toperform in order to simply match the results from the statistically basedEinhorn (2000) method?

1 The authors’ described their finding as a “reverse Lake Wobegone” (p. 27) effect, theopposite of performance in the fictional town of Lake Wobegone where all the childrenare above average.



5. How do tax policies, including the reduction in taxes applicable todividends and capital gains, enacted under the Jobs and Growth TaxRelief Reconciliation Act of 2003, affect the required threshold for expertperformance?

Expert Judgment

According to Einhorn (2000), an important attribute of experts is theability to identify cues that might not be visible to others. An expert should,for example, be able to identify a weak signal from a noisy background.Further, after cues are identified an expert should be able to measure, cluster,weigh and combine them in order to form a global judgment. Research shows,however, that whereas “experts” typically propose that their judgments arederived from a careful weighing of many variables, in practice, only a fewvariables actually influence their determination (Faust & Ziskin 1988; Fisch,Hammond, & Joyce, 1982; Gillis & Moran 1981). Ironically, there may even bean inverse relationship between the confidence and skill of the expert (Faust& Ziskin).

Casting a further pale on the viability of expert judgment is that avariety of studies have shown that statistical evaluation, based on weightedlinear models or often simply unweighted (improper) models provide resultssuperior to those generated by human experts (Dawes, 1986; Meehl, 1954).Meehl summarized approximately 20 studies comparing “expert” judgmentwith statistical evaluation, based on weighted linear models. In all of thestudies, the statistical model either exceeded or tied the performance of theclinician.

Years later, Dawes (1986) went a step further conducting a series ofstudies comparing the performance of randomly weighted (i.e., improper)linear models with “expert” judgments in three scenarios: a) diagnoses ofneurosis of psychosis of 860 patients, using the Minnesota MultiphasicPersonality Inventory; b) first-year graduate graduate-school grade pointaverages at the University of Illinois, using 10 variable; and c) faculty ratingsof graduate students at the University of Oregon, using undergraduate grade-point averages, GRE scores, and selectivity of their undergraduate schools.Dawes found that, on average, random linear models accounted for 150%more variance than the holistic judgments of the experts.

The relevant issues, here, with respect to the prior research are 1)have financial experts, in this case the Business Week experts, demonstratedan ability to reliably forecast future market closings, and 2) would a simpleformula provide comparable or perhaps even better results?


Research &

Theory

Heuristic Bias: Anchoring and Adjustment

An important attribute for experts predicting future financial marketsresults is the ability to adjust their estimates to account for the high level ofyear-to-year volatility. During the twenty-year period of 1985 – 2004, forexample, while the average annual change in the Dow Jones IndustrialsAverage showed an increase of 12.5%, in only three years were actual returnswithin a 15-percentage point range of 5.0% to 20.0%. Results from the other17 years included four years with negative returns and nine years with returnsin excess of 20.0%.

Despite the requirement that predictions should vary in accordancewith their underlying volatility, research has shown that such judgments areoften made in accordance with the heuristic process referred to as anchoringand adjustment, whereby an initial judgment serves as an anchor for subse-quent judgments. The essential shortcoming of the anchoring and adjustmentmethod is that there is insufficient adjustment away from the initial judgment(Tversky & Kahnemann, 1974).

If experts are to be successful in making their predictions theyshould be able to avoid the proclivity to engage in the anchoring andadjustment bias. In practice, they could potentially avoid this bias byemploying techniques such as using multiple anchors, convergent validationtechniques, and different assessment techniques (e.g., fractiles and directprobability estimation) (Edwards & von Winterfeldt 1986). This studyattempts to evaluate whether, in fact, the Business Week experts demonstratean ability to avoid the common proclivity to engage in anchoring and adjust-ment

Minimum Required Threshold of Accuracy and the “Einhorn”Standard

The intimate tie between prediction and decision-making makes itincumbent that it be evaluated in the context of how it provides guidance foraction (Einhorn, 1974). Evaluation of the Business Week consensus predic-tions, therefore, should consider and perhaps center upon how effective theexperts are in providing guidance for readers who are determining whether orto what extent they should adjust their investment portfolios. If, for example,the experts predict stocks will generate a negative or below average rate ofreturn, investors might be expected to adjust their portfolios by decreasingstock holdings and increasing bond holdings. Article titles, such as “TheWay to Play this Market” (Vickers 2002), contained in one of the annualBusiness Week investments guide suggests that these predictions areprovided largely for that purpose.



A great challenge for the experts, to have their predictions serve as auseful barometer for asset allocation, however, is that they must not only bebetter than random, but sufficiently better to overcome the hurdle of predict-ing from an unbalanced universe of outcomes. Einhorn (1986), in his well-titled article, “Accepting Error to Make Less Error,” summarized this chal-lenge. If, as he explains through an example, we know a universe has 60% redoutcomes and 40% green outcomes, and we always choose red, we will beright 60% of the time. If, however, in an attempt to improve upon thatstandard, we choose red 60% of the time and green 40% of the time, we canexpect to achieve, at random, a success rate of only 52%.2 In this universe, asubject would have to be able to achieve success at a rate of approximately15% better than random in order to approximately match the results of theEinhorn method.3

In cases where readers use the Business Week expert forecasts as ameans of determining whether to move between stock and bond positions, therisk and potential for harm would seem to be real. If, for example, similarly toEinhorn’s (1986) red ball/green ball scenario, the investment universe stocksoutperform bonds 60% of the time, failure to provide performance that is atleast 15% above random, will result in less success than would have occurredhad stock been selected 100% of the time.

The feasibility of prognosticators matching the Einhorn (1986)threshold may depend largely on the expected outcomes within the universe.Should it be determined, for example, that a threshold of 55% stock selectionsis appropriate, an amount that is only slightly above random, then thefeasibility would be much greater than in cases where the distribution ofexpected outcomes is substantially more skewed. This study, using historicaldata, attempts to determine what would be a realistic universe of expectedoutcomes and, combined with the findings of Business Week predictiveaccuracy, evaluate whether it is feasible to expect that any consensus set ofexpert forecasts can serve as a reliable indicator of future market results.

Taxes

The effect of taxes may be expected to increase the required thresh-old of accuracy for prognosticators who, through their forecasts, implicitly orexplicitly (e.g., Business Week article, “The Way to Play this Market”)

2 Einhorn (1986) used an example of an urn containing 60 red balls and 40 green balls.If the subject guesses red 60 times, he or she would be expected to be correct 36 times(60 red balls x .60 probability). Similarly, if the subject guesses green 40 times, he orshe would be expected to be correct 16 times (40 balls x .40 probability). The sum of36 + 16 =52.3 Approximate matching would be achieved if the subject were to be correct 69% (1.15x 60) for the red balls and 46% (1.15 x 40) for the green balls: (60 balls x .69 probabil-ity) + (40 green balls x .46 probability) = 59.8%


Research &

Theory

advocate that investors adjust their portfolios based on their perceptions offuture market results. Stocks are tax-advantaged in that both dividends andcapital gains are subject to a maximum tax rate of 15%. Prior to passage of theJobs Growth and Tax Relief Reconciliation Act of 2003, capital gains weresubject to a maximum tax rate of 20% and dividends were taxed at ordinaryincome rates.

Taxes may be expected to increase the challenge for the prognostica-tors in at least the following three ways. First, the expected distribution ofpreferred investment may become increasingly skewed towards stocks, ratherthan bonds, potentially raising the required threshold of accuracy to meet thestatistically based Einhorn (1986) standard.

In an earlier section it was noted that if a distribution was expected tofavor stocks 60% of the time, then a prognosticator would have to perform ata rate approximately 15% better than random to match the Einhorn (1986)standard. If, however, because of taxes the expectation for stocks as preferredchoice increases, for example, to 65%, then the prognosticator must increaseaccuracy to 19% above random to meet that same standard.

Second, conversion of stocks into bonds can result in prematurerecognition of income, reducing the benefits of compounding. As an example,a $100,000 stock investment yielding an annual return of 12%, subject to acombined state and local tax rate of 20% (e.g., 15% federal and 5% state), andheld for 20 years would generate after-tax income of $691,700. Alternatively, ifthat same investment was taxed at the same rate and held for the same period,but taxed annually, rather than at redemption, the after-tax return would be$525,500, resulting in a reduction in income of $166,200.

Third, conversion of stocks into bonds can diminish or eveneliminate the ability to altogether avoid capital gains taxes by holdingappreciated property until death. Under current law, no capital gains tax isapplied to investments containing unrealized appreciation at an investor’sdeath.

These three issues carry on added significance because a strategicinvestor is more likely to hold stocks rather than bonds in taxable accounts.This occurs because a) stocks held in a tax-deferred or tax-exempt accountlose their tax-advantages, b) no compounding benefit is available for interest-bearing investments held in a taxable account, and c) no tax advantage isavailable for interest-bearing accounts held until death.

Another potential tax factor potentially increasing the attractivenessof stocks is that while state income taxes, regardless of how generated (e.g.,dividends or interest), are add backs for the alternative minimum tax (AMT),their timing of inclusion may be much easier managed for stocks than forbonds. An investor in stocks, unlike the investor in bonds whose incomemust be recognized yearly, therefore, may have the ability to time the recogni-tion of income in a year when he or she is not expecting to be subject to thealternative minimum tax.



Issues and Evaluation

Have the Business Week experts demonstrated an ability to predictfollowing year-end market results for the Dow Jones Industrial Average?

1. The Business Week consensus expert predictions demonstrated no abilityto predict future market returns. Comparisons between the Business Weekcomposite predictions and actual changes are shown in Table 1.

Table 1Comparison and Correlation of Predicted vs. Actual MarketChanges

Business Week ActualPredicted Market MarketChange Change

1996 + 6.1% + 26.0%1997 + 2.2% + 22.6%1998 + 7.0% + 16.1%1999 + 4.2% + 25.2%2000 +5.7% - 6.2%2001 +11.4% - 7.1%2002 +10.7% - 16.8%2003 +18.3% +25.3%2004 + 2.0% + 3.1%Correlation = -0.09

Changes based on difference between current year-end closing andBusiness Week experts’ predictions of the following year-end closing marketprice.

The correlation results suggest that the Business Week compositepredictions provide no benefit in assessing future market performance. Alimitation of this finding, however, is that while the composite predictionsincluded several hundred individual predictions, results are limited to only thenine years that the forecasts have been in existence.

2. Can a simple formula based on prior performance match or exceed that ofthe Business Week experts?

While the research of Dawes (1954), Meehl (1986), and otherssuggest that simple formulas, even unweighted ones, are likely to providepredictions that will be at least as accurate as those of experts, development


Research &

Theory

of a simple formula for forecasting future stock market returns may besomewhat more problematic than the ones contained in their analyses. Inpredicting future student performance, for example, at least two factors areimmediately evident: grade point averages and graduate school aptitude tests.In the case of predicting future market results only one factor, however, priorperformance stands out as the one that is universally used. Therefore, acomparison was made between predictions based simply on prior perfor-mance, using a ten-year rolling average, and those of Business Week experts(Table 2). Ten-year rolling averages were selected because that time frame istypically included in reports of mutual fund performance.

Table 2Comparison of Annual Absolute Differences Between ActualDow Jones Industrial Averages Year-End Market Closing andPredicted Market Closings

BW 10 Yr. RollActual BW Actual/ 10 Yr. Roll Actual/Market Forecast Forecast Forecast ForecastClosing Closing Difference Closing Difference (a) (b) (c) (d) (e)

a – b = c a – d = e1996 6,448 5,430 1,018 5,798 6501997 7,908 6,587 1,321 7,331 5771998 9,181 8,467 717 9,150 311999 11,497 9,567 1,930 10,659 8382000 10,787 12,154 1,367 13,325 2,5382001 10,022 12,015 1,993 12,491 2,4692002 8,342 11,090 2,748 11,325 2,9832003 10,454 9,871 583 9,251 1,2032004 10,783 10,665 118 11,936 936

——— ———11,795 12,225

Average annual difference 1, 311 1,358

Test for equality of variances, p = 0.25

While the average absolute differences for the Business Weekpredictions were slightly superior to that of the ten-year rolling average, thedifferences between the two as shown above, were statistically insignificant.Similar results were also obtained using a 20-year rolling average, with anabsolute average difference of 1,394 (test for equality of variances, p = 0.36).



The results, therefore, indicate no evidence that forecasts generated by eitherthe Business Week experts or a one-factor prior performance test providesvalid predictions of future market results.

3. Have the Business Week experts demonstrated an ability to avoid aproclivity to engage in the heuristic referred to as “anchoring and adjustment,whereby initial judgments serve as an anchor for insufficiently adjusted futurejudgments?”

Although no specific standard exists for evaluation of this character-istic, indications that the subjects are employing this strategy can be foundby a) signs of data clustered around a potential anchor, and b) a comparisonbetween the standard deviations of the expert’s predictions and the actualstandard deviation.

A review of Table 1 shows that predictions were clustered relativelyclosely together during the five-year period 1996-2000 and also for the two-year period 2001-2002. Only in 2003 and 2004 is their significant variation, buteven these numbers may exhibit less variation on the experts’ part than isshown in the year to year prediction. The Business Week issues are pub-lished prior to year-end and, therefore, reflect predictions for the next year,compiled before the end of the current year.

The predictions, shown in Table 1, based on changes between years,accordingly, may not reflect the thought processes of experts a few weeksbefore year-end. The consensus year-end predictions, for 2004, for example,while based on 2003 year-end totals, show a predicted change of 2.0%. If,however, the predictions are based on the DJIA market closing on December10, 2003, three weeks prior to year-end, the predicted change would be 7.5%,more consistent with recent predictions. The Business Week article describ-ing the forecasts stated that the experts predicted an increase of 6.2%, a datethat would be consistent with the DJIA’s closing price on December 12, 2003.

The article accompanying the 2000 predictions, similarly, stated thatthe experts predicted an 8.3% increase in the Dow, consistent with a December10th closing date. Table 3 provides a comparison between Business Weekpredictions based on a December 10th closing date and actual market changes.The correlation for the December 10th closing date, similar to the December 31closing date, is also negative.


Research &

Theory

Table 3Comparison of Predicted Market Changes, at December 10th

Closing Date and Actual Year-to-Year ChangesBusiness WeekPredicted Market ActualChange, as of MarketDecember 10th Change

1996 + 4.7% + 26.0%1997 + 1.0% + 22.6%1998 + 6.1% + 16.1%1999 + 8.2% + 25.2%2000 + 8.3% - 6.2%2001 +12.0% - 7.1%2002 +11.8% - 16.8%2003 +15.1% * + 25.3%2004 + 7.5% + 3.1%

Correlation = -0.3736

* If the predictions were based on market closings 1 week earlier, December 3,2002, the predicted market change for 2003 would have been 12.9%.

This clustering of predictions, whether based on exact one yearchanges (Table 1) or on ones that more likely reflect the prognosticatorsthought processes (Table 3) are in stark contrast to the actual results. Theactual mean change for the period 1996-2004 was 9.8% per year with nochanges within the range greater than 3.1% and less than 16.1%.

The standard deviation for the actual changes was 16.7%, while only5.2% for the consensus Business Week predictions, based on year-to-yearchanges and even lower for the predictions based on the December 10th

closing, 4.2%. The combined findings of clustering and inordinately lowstandard deviations strongly suggest that the Business Week experts’judgments may be generated in accordance with the anchoring and adjust-ment heuristic. Table 4 provides a summary of the means and standarddeviations for the actual results and two forms of prediction.



Table 4Anchoring and AdjustmentStandard Deviations in Forecasts of DJIA Closings [1996 –2004]Comparisons of Actual and PredictedActual Standard Deviation 16.7% [mean change 9.8%]Business Week Forecasts

based on year-end [Table 2] 5.2% [mean forecast 7.5%]based on December 10th [Table 3] 4.2% [mean forecast 8.3%]

Evidence of Business Week Experts employing the anchoring andadjustment heuristic is demonstrated by the lack of variability in their fore-casts. Actual variability is approximately three to four times greater than thatcontained in the consensus forecasts.

4. How much better than random would expert forecasts have to be in order tosimply match the results from the statistically based Einhorn (1986) method?

This question, while couched in quantitative terms, is intended toaddress the broader issue of how realistic is it to expect that experts canpredict future market results where there is likely to be an unbalanced, albeituncertain, universe of outcomes. As discussed earlier, using the Einhorn(1986) example, if a universe contains 60% of outcome #1 and 40% of outcome#2, and an individual 60% of the time chooses outcome #1 and 40% of thetime chooses outcome #2, the choices would have to be correct at a rateapproximately 15% greater than random in order to match the results ofsomeone who choose outcome #1 in all cases.

Prospectively, no one knows, of course, exactly how often bondsand/or treasury bills will outperform stocks. Nonetheless, if one relies onhistorical returns as baselines, some determinations can be made as to howmuch better than random a prognosticator must perform in order to match thestatistically based Einhorn (1986) method.

Table 5 presents historical returns, showing the number of yearsintermediate-term government bonds and U.S. Treasury bills have outper-formed large company stocks (Ibbotson, 2005) and the expert performancethat would be required in excess of random to match the Einhorn (1986)standard, based on those time frames.

Over periods of both 79 years (1926 –2004) and 35 years (1970 –2004), stocks outperformed U.S. Treasury bills 63% of the time. Over thosesame periods of 79 and 35 years, stocks outperformed intermediate-termgovernment bonds 65% and 66% of the time, respectively. More recently,1985-2004, stocks outperformed both intermediate-term government bonds


Research &

Theory

and U.S. Treasury bills, 75% and 70% of the time, respectively. The requiredexpert performance in excess of random to match the Einhorn (1986) standardfor bonds and bills for the various time frames ranged from 18% to 21%.4

These findings do not consider the effect of taxes, which are addressed in thenext section.

The finding that a prognosticator’s accuracy may need to be 18% to21% greater than random simply to match the Einhorn (1986) method, com-bined with the lack of evidence that a consensus group of experts shows anyability to forecast beyond random, may suggest that investors should notreduce their stock holdings for any individual year based on consensus expertforecasts.

Table 5Historical Returns and Required Threshold of Accuracy Pre-Tax

Years Stocked Required ExpertOutperformed Performance inIntermediate-Term Excess of RandomGovernment Bonds/ To Match EinhornU.S. Treasury Bills Standard

1926 – 2004Intermediate Gov’t Bonds 51/79 65% 19%U.S. Treasury Bills 50/79 63% 18%

1970– 2004Intermediate Gov’t Bonds 23/35 66% 20%U.S. Treasury Bills 22/35 63% 18%

1985 – 2004Intermediate Gov’t Bonds 15/20 75% 20%U.S. Treasury Bills 14/20 70% 21%

4 Note that after base rates, in this case the number of years stocks outperform bonds/bills, reach approximately 71%, any further increase results in a reduction in therequired performance in excess of random by experts to match the statistically basedEinhorn method.



5. How do tax policies, including the reduction in taxes applicable to divi-dends and capital gains, enacted under the Jobs and Growth Tax ReliefReconciliation Act of 2003, affect the required threshold for expert perfor-mance?

The study found, perhaps surprisingly, that neither taxes norcategory of debt would be expected to have more than a negligible effect onan investor’s decision whether to invest in stock or debt holdings for anyindividual year. The loss of the benefits of compounding, however, as well astaxation of debt instruments at ordinary income rates were found to bepotentially significant deterrents to adopting a policy of switching betweenstock and debt investments. Such a policy would be expected to be especiallydetrimental to an investor who was planning to hold his or her portfolio untildeath.

Table 6 illustrates how taxes and class of debt investment appear tohave only a very negligible effect on the experts’ performance requirement formatching the statistically based Einhorn (1986) method. These findings arisedespite a) the substantial differences between taxation of stocks and debtinstruments following passage of the Jobs Growth and Tax Relief Reconcilia-tion Act of 2003, and b) the substantial difference in historical annual returnsbetween intermediate-term government bonds and U.S. Treasury bills

The study examined how the new tax provisions, had they been inforce in prior years, would have affected the determination of the percentageof years in which stocks, as opposed to bonds or U.S. Treasury bills, wouldhave provided the higher after-tax return.5 It was found that during theperiod, 1926 – 2004, in only one year, 1993, was the pre-tax/after-tax choicedifferent for stocks and intermediate-term government bonds and in only twoyears, 1978 and 1987, was the pre-tax/after-tax choices different for stocks andU.S. Treasury bills.

After factoring in the effect of taxes, it was found that over both the79-year period and 35-year period stocks outperformed intermediate-termgovernment bonds and U.S. Treasury bills an equal number of times. Thisfinding suggests that despite the fact that intermediate-term governmentbonds have provided a higher average rate of return than U.S. Treasury bills,category of debt has no effect on required expert performance. Similarly,taxes, as a factor for determining whether for any individual year stock willoutperform debt instruments, seem to have only a very negligible effect. Incomparing required expert performance pre-tax vs. after tax, the amountsranged from 18% to 21% across the three time frames for both groups.

5 Following passage of the Jobs Growth and Tax Relief Reconciliation Act of 2003,investment returns for stocks (i.e., dividends and capital gains) are subject to amaximum federal tax rate of 15%. Interest generated from intermediate-term govern-ment bonds and U.S. Treasury bills are subject to tax at a taxpayer’s ordinary incometax rate, which can be as high as 35%.


Research &

Theory

Table 6Historical Returns and Required Threshold of Accuracy After-Tax

Years Stocks Required RequiredOutperformed Expert ExpertIntermediate- Performance Performance Term to Match to MatchGovernment Einhorn EinhornBonds/ Standard StandardU.S. Bills After-tax Pre-tax

1926 - 2004Int. Gov’t Bonds 52/79 66% 20% 19%U.S. Treas. Bills 52/79 66% 20% 18%

1970 – 2004Int. Gov’t Bonds 24/35 69% 21% 20%U.S. Treas. Bills 24/35 69% 21% 18%

1985 –2004Int. Gov’t Bonds 16/20 80% 18% 20%U.S. Treas. Bills 15/20 75% 20% 21%

After-tax returns are based on Intermediate Government Bonds andU.S. Treasury Bills being subject to a marginal tax rate of 35%, while stockreturns (dividends and capital gains) are subject to a marginal tax rate of 15%.

The study, in a further attempt to evaluate the viability of usingconsensus expert forecasts as a tool for asset allocation, also examined how astrategy of shifting between stocks and bonds/bills would fare, on an after-taxbasis, compared with a strategy of buying and holding stocks. The findingssuggest that losses of compounding benefits and preferred tax rates (i.e.,maximum tax rates of 15% for stock transactions), resulting from the strategyof shifting between stocks and bonds could potentially be highly detrimental,especially over extended periods of time.

The method used to attempt to evaluate the effects of a strategy ofswitching between stocks and bonds for individual years, was to model thereturns of an investor who would hold stocks for two years, sell them, paycapital gains taxes, and use the net proceeds to purchase bonds and/or billsto hold for one year. At the conclusion of that third year, the bonds/billswould be sold, taxes at ordinary income rates would be paid, and anotherthree-year cycle would repeat itself (i.e., use net proceeds to purchase stock,hold for two years, sell, pay taxes, and use net proceeds to purchase bonds/



bills). The employment of such strategy where stocks are held for two out ofevery three years is based on the findings in Table 6 that found that over anextended period of time, 1926 – 2004, stocks outperformed bonds and bills66% of the time. The model also presents, as points of comparison, thereturns for investors who a) buy and hold stock until an extended sale dateand b) who hold the stock until death

Illustration

Three investors each begin with initial stock investments of $100,000that generate pre-tax income of 10% per year. Each investor is also subject toa 35% tax rate for interest and a 15% tax rate for capital gains and dividends.For simplicity, all stock income is deemed derived from capital gains.

Investor #1 invests the $100,000 in stocks, holds it for two years,sells it, pays capital gains taxes and with the net proceeds buys fixed invest-ments. At the end of Year #3, the bonds are sold, income taxes are paid, andnew stock investments are purchased. This pattern repeats itself every threeyears.

Investor #2 holds the investment until it is sold at the end of theperiod. This models the buy-and-hold stock investor.

Investor #3 holds the investment until death, at the end of theperiod. This models the buy-and-hold stock investor who holds appreciatedstock as an estate-planning strategy. No capital gains taxes are assessed onappreciated capital assets that are held at time of death.

Payouts:Term After-Tax Values

[$100,000 Initial Investment](Years) Investor #1 Investor #2 Investor #33 $125,510 $ 128,135 $ 133,1006 157,528 165,583 177,1569 197,713 215,426 235,79512 248,149 281,767 313,84315 311,452 370,066 417,72518 390,904 487,593 555,99221 490,623 644,021 740,02524 615,781 852,227 984,97327 772,867 1,129,349 1,310,99930 970,026 1,498,199 1,744,940

The illustration shows that even where the pre-tax return is the samein each year, the combined effects of 1) partial loss of compounding benefitsand 2) partial taxation at ordinary income rates can be substantial and becomeincreasingly detrimental as there are increases in periods of time.


Research &

Theory

Conclusion

The findings from this study suggest that potential benefits (e.g.,higher net investment return) of utilizing consensus expert forecasts as a toolfor adjusting one’s asset allocation may be outweighed by at least threeimportant risks. These risks would seem to apply not only to users of theBusiness Week forecasts, but also to users of any set of yearly consensusforecasts.

The first and potentially very serious risk is that the prognosticators’forecasts may not be able to capture the variability in investment returns,because the decision makers, despite their exalted “expert” status may havethe same proclivity as laymen to revert to a decision making technique ofanchoring and adjustment. This proclivity was evidenced by excessively lowstandard deviations for forecasts and the clustering of similar forecasts (e.g.,Table 3: forecasts for years of 1998 – 2000 were very similar as were theforecasts for 2001 – 2003).

The second risk, the findings indicate, is that the experts forecasts, inorder to be useful, not only have to be better than random, but must besubstantially better than random, in order to provide useful information toinvestors. The determination that a prognosticator, based on historical ratesof performance, would have to perform at a rate of 18% to 21% above randomsimply to match the statistically based Einhorn (1986) method, may present aninsurmountable challenge. While some investors and/or experts may, ofcourse, argue that the prior rates do not apply, it would seem likely that thereshould be at least some skewing towards stocks to reflect the additionalcompensation required for assuming additional risk (i.e., equity premium).

The third risk, for investors holding their investments in a taxableaccount, is that the experts’ forecasts will not be sufficiently accurate tocompensate for the loss of benefits of compounding and taxation of income atpreferred rates. This risk assumes increasing importance for long-terminvestment strategies.

Overall, the findings strongly suggest that investors should behighly cautious about shifting assets from stocks into bonds based onconsensus expert forecasts. Even higher caution may be advised where theinvestments are held in taxable accounts, especially where the portfolio is tobe held for an extended period of time. Finally, investors who may still behopeful that the forecasts can provide useful information should keep in mindthat such benefits, once determined to be real, will likely be short-lived. Thisphenomenon, referred to the Lucas critique,6 essentially states that if priorpatterns become detectable, the market will adapt immediately, effectivelynullifying any future benefit (Lucas, 1976).



References

Bauer, A., Eisenbeis, R.A., Waggoner, D.F. & Zha, T. (2003). Forecastevaluation with cross-sectional data: The blue chip surveys.Economic Review – Federal Reserve Bank of Atlanta, 88 (2), 17- 29.

Dawes, R.M. (1986). Proper and improper linear models. International Journalof Forecasting, 2, 5-14.

Edwards, W., & von Winterfeldt, D. (1986). On cognitive illusions and theirimplications. Southern California Law Review, 59 (2), 401-451.

Einhorn, H.E. (2000). Expert judgment: Some necessary conditions and anexample. Judgment and Decision Making, Connelly, T., Arkes, H.R.,& Hammond, K. R. (eds.). Cambridge, UK: Cambridge UniversityPress.

Einhorn, H.E. (1986). Accepting Error to Make Less Error. Journal ofPersonality Assessment, 50 (3), 387-395.

Faust, D. & Ziskin, J. (1988). The Expert Witness in Psychology andPsychiatry. Science, 241, 31-35.

Fisch, H.U., Hammond, K.R. & C.R. Joyce (1982). On Evaluating the Severityof Depression: An Experimental Study of Psychiatrists. BritishJournal of Psychiatry, 140, 378-383.

Gillis, J. & Moran, J. (1981). An Analysis of Drug Decisions in a StatePsychiatric Hospital. Journal of Clinical Psychology, 37 (1), 32-42.

Ibbotson Associates. (2005). Stocks, Bonds, Bills, and Inflation: 2005Yearbook. Chicago: Ibbotson Associates.

Ling, D.C. (2005). A Random Walk Down Main Street: Can Experts PredictReturns on Commercial Real Estate? The Journal of Real EstateResearch, 27 (2), 137-144.

Lucas, R.E. (1976). Econometric Policy Evaluation: A Critique, in R.E. Lucas(ed.), Studies in Business-Cycle Theory (pp. 104-130). Cambridge,Massachusetts: MIT Press.

Meehl, P.E. (1954). Clinical vs. Statistical Prediction. Minneapolis:University of Minnesota Press.

Taler, N.N. (2001). Fooled by Randomness. New York: Texere LLC.Tversky, A. & Kahnemann, D. (1974). Judgment Under Uncertainty:

Heuristics and Biases. Science, 185, 1124-1131.Vickers, M. (2002). The Way to Play this Market. Business Week, 3814, 66-69.

Contact Information: Seth Hammer, Ph.D., Associate Professor of Accounting,Towson University, Department of Accounting, 8000 York Road, Towson,Maryland 21252-0001. Phone: (410) 704-2227; E-mail: [email protected].


Research &

Theory

ABSTRACT

This study investigated, using Dow Jones Industrial Average (DJIA)companies, whether writing a covered call provides a significant benefitover outright stock investment. It was found that even a simplestrategy using out-of-the-money calls brought significant benefits toinvestors in terms of risk and reward. The results demonstrated that,during the period of 1990-2003, covered call writing reduced the volatil-ity of returns significantly, but did not significantly lower returns. Thisstudy also investigated two other covered call strategies. Writing acovered call with half of the position in out-of-the-money calls and theother half in in-the-money calls worked as well as writing covered callswith out-of-the-money calls only. In fact, the former worked even betterthan the latter during the less bullish period. However, writing coveredcalls with in-the-money calls only did not work well.

COVERED CALL WRITING AS A STRATEGY FOR

SMALL INVESTORS

Ki C. Han, Ph.D.Suffolk University

&Kamlesh Dadlani

Ernst & Young LLP

Introduction

The wide spread market perception is that options trading is toorisky for most investors to be involved in. Therefore, investors feel thatbuying and selling options is beyond the domain of their investment. Themain reason for this negative market perception is that options are derivatives.The value of options, as derivatives, is primarily determined by the value oftheir underlying assets, and, naturally, options are more complicated to valuethan the underlying assets. Even financial planners and advisors do not feelvery comfortable with options, and as a result, they seldom recommend theseinvestments for their clients to include options as part of their portfolios. Asinvestors are getting more educated and sophisticated, however, their view ofoptions is changing. (See Schlosser, 2002 and Bittman, 2005)



Now, more than three decades after the standardized options beganto be traded on the exchange market, many people question if options tradingis indeed riskier and more complicated than the trading of the underlyingassets.1 For example, is the trading of stock options necessarily morecomplicated than the trading of stocks? Is the trading of stock options onlyfor high risk-taking investors? Of course, some option trading strategies,such as butterfly call spreads and ratio backspreads, are more appropriate forsophisticated investors. However, this does not mean that ordinary investorsshould entirely shun away from options trading. In fact, some of optiontrading strategies are easy to use and do not expose their users to greater risk.One good example is covered call writing.2 This strategy involves owning astock and writing call options on that stock. As a straight-forward strategycovered call writing does not involve the complicated calculations required forsome option trading strategies, such as delta, vega, etc.

This study attempts to answer the above questions using the realdata of the Dow Jones Industrial Average (DJIA) stocks and their call options.This study adopts a simple, naïve approach to covered call writing. Sophisti-cated investors would be able to implement a variety of follow-up actionsdepending upon the movement of the stock. For example, if the stock pricegoes down, sophisticated investors can implement so-called rolling downstrategies, in which they buy back the original call and sell a call with a lowerstriking price, thus lowering the breakeven price of the stock.3 In this paper,however, the focus is on the simple covered call writing strategy that can beused by anyone, without in-depth analysis of the nature of options.

Covered Call Writing

Although options have been available as a valuable investmentvehicle for a long time, even simple option trading strategies like covered callwriting are rarely used by investors.4 As was mentioned above, covered callwriting involves owning a stock and writing a call option on that stock.Therefore, if someone wants to write a covered call, all the investor has to dois to buy a stock and sell its call option. As the price of the stock rises, theinvestor gains on the long position of the stock. However, the gains arelimited because the call may be called away as the stock price rises above theexercise price of the call option. Thus, by selling the call option, the investoris not able to fully participate in a strong upward movement of the price of thestock. On the other hand, if the price of the stock declines, the investor willget partially compensated by the premium received when the call is written. Aloss materializes if the stock price falls by a distance greater than the premium.

Some people believe that covered call writing is profitable when thestock market is in downturn. However, that is not true. Although investorswho own stocks can lower their breakeven point by writing calls, they still


Research &

Theory

lose money if stocks are in downturn. Covered call writing is most profitablewhen the market is bullish. The reason is as follows. If an investor writes acovered call, one of two things can happen. If the stock price does not moveup significantly and finishes lower than the exercise price, the call that theinvestor wrote will expire worthless, meaning that the option will not beexercised. On the other hand, if the stock price rises sufficiently and finisheshigher than the exercise price, the call will be exercised, meaning that the stockwill be called away. In which case would the investor make more money? Ofcourse, the answer is the second case. Some people would argue thatoutright stock investment is better in a bullish market condition. However,there are three problems with this argument. First, stocks hardly ever move injust one direction. Even in a true bull market, stocks, in general, go throughups and downs. Covered call writing has a potential to smooth out these upsand downs, and this can benefit investors. Second, even in a bull market,covered call writing can work better than outright stock investment as long asthe price of the stock does not rise rapidly. Third, it is hard to tell whether themarket will be bullish or bearish in the future. The fact that the market hasbeen bullish does not guarantee that the market will continue to be a bullmarket. Therefore, even if the above argument is true, it would not help us tomake an investment decision between outright stock investment and coveredcall writing.

Data and Methodology

This study employed data from the Dow Jones Call Option databasefor January 1990 through December 2003, which was provided by QuantitativeAnalytics, Inc., a Chicago-based company. Because of missing data for sixcompanies (Exxon Mobil, GM, JP Morgan Chase, SBC, United Technology,and Wal-Mart), the sample consists of 24 companies.

For covered call writing, investors typically buy a stock and sell acall option with a strike price right above the stock price. This approach iscalled an “Out-of-the-Money Covered Write” (OTM). In this approachinvestors receive a call premium and leave some room for the stock to risebefore it hits the exercise price of the call. In computing returns on thiscovered call write, it is assumed that, on the third Friday of each month,investors buy 1000 shares of the stock at the closing price and sell, at theclosing price, 10 contracts of the call option that has the strike price rightabove the price of the stock and approximately one month of time to expira-tion. On the third Friday of the following month the position is liquidated.This is repeated month after month. Therefore, this strategy takes advantageof the fact that options have a faster decay of time value during the last monththan any other month.5 This study adopts a simple investment strategy, notconsidering any type of follow-up. Therefore, this investment strategy can be



used by anyone even if the investor has no sophisticated skills in optionstrading. In fact, this strategy is not any more complicated than outright stockinvesting. In stock investment it is always challenging to pick the right timeto buy or sell the stock. Compared to that, the covered call writing strategyadopted in this paper is fairly simple. All that an investor needs to do is buythe stock and sell its one-month call option on the third Friday of each month.Then the investor keeps the positions until the expiration date of the call, notworrying about timing their investment.

OTM is not the only way to do covered call writing. Two otherapproaches are investigated: In-the-Money covered write (ITM) and Diversi-fied covered write (DIV). In the ITM approach, 10 contracts of the call optionwith a strike price right below the stock price are written. Investors receive ahigher call premium because the exercise price of the call is lower. However,there is no room for stock price appreciation. In this sense this is an ultra-conservative approach that offers lower potential rewards but more downsiderisk protection. In the DIV approach, 5 contracts are written with a strikeprice right above the stock price, and the other 5 contracts, with a strike priceright below the stock price. This is, therefore, a combination of OTM andITM. This approach can be attractive when investors feel that out-of-the-money calls do not provide enough downside protection and in-the-moneycalls do not provide enough returns. Through this DIV approach investorscan acquire both the return and protection.

In comparing the performance of the above strategies, Sharpe ratios(Sharpe (1966; 1994)) are used. The Sharpe ratio is a measure of risk-adjustedperformance that uses a benchmark based on the ex post capital market line. Itmeasures returns relative to the total risk of the portfolio where total risk is thestandard deviation of portfolio returns. Other measures widely used inmeasuring portfolio performance include the Treynor index and Jensen’salpha. Given the controversial validity of the beta coefficient (e.g., Fama andFrench, 1992; 1993; 1996), however, the Treynor index and Jensen’s alpha arenot appropriate for the purpose of this study.


Research &

Theory

The Sharpe ratios are calculated as follows: ___ ___ R

j - R

f

Sj = —————— (1)

ój

where,S

j= Sharpe ratio for an investment strategy j,

___R

j= mean return on a investment strategy j,

___R

f= mean return on U.S. Treasury bills, and

ój = standard deviation of return on strategy j.

For the Sharpe ratio calculation, this study utilized the average of thethree month T-Bill as the risk-free rate. The Sharpe ratio of each strategy wascalculated for the whole sample period of 1990-2003. Two subperiods werealso investigated: the more bullish period and the less bullish period. Themore bullish period covers the period of January 1993 through December 1999.The rest of the sample period is called the less bullish period: from January1990 to December 1992 and from January 2000 to December 2003.

Empirical Results

Table 1 presents the performance results of the OTM and theoutright stock investment for the sample period of 1990 - 2003. For eachsample firm, the table lists the average returns, standard deviations, and theSharpe ratios of the two strategies. It should be noted that all of these resultsare based on monthly returns. For the Alcoa stock, as an example, the averagemonthly return on the OTM is 1.08% (approximately 13.8% on the annualbasis), while the average monthly return on the stock is 1.31% (approximately16.9% on the annual basis). The standard deviations are 5.65% and 8.64%,respectively. Given these average returns and standard deviations, theSharpe ratios turn out to be 0.13 and 0.11, respectively. Several statistical testresults are reported at the bottom of the table. Since the distribution ofreturns and standard deviations do not follow a normal distribution, thisstudy did not use parametric tests.6 Instead, the study used non-parametric



tests, which are less subject to distribution assumptions than parametric tests(Conover, 1980). Throughout this paper non-parametric test results arepresented. Both the Wilcoxon signed rank test and van der Waerden testshow that the average returns are not significantly different between the twostrategies. However, the test statistics demonstrate that the covered callwriting has a lower standard deviation than the outright stock investment, andthat the difference is significant at the 1% level. In fact, the OTM has a lowerstandard deviation for all of the sample stocks except Honeywell. Thedifference in standard deviation is most pronounced in the case of Intel andMicrosoft stocks. For these two stocks, the OTM has about 40% lowerstandard deviation than the stock investment. This highlights the fact thatcovered call writing effectively smoothes out returns by limiting upsidepotentials and providing some downside risk protection. Finally, the teststatistics suggest that the Sharpe ratios under the OTM are significantlyhigher than those under the outright stock investment at the 1% level. It isclear that the lower Sharpe ratios in the covered call writing are due to itslower standard deviations, not to its higher returns. This indicates that theprimary benefit that a covered call provides for investors is a reduction of riskwith little change in returns.7

Table 2 reports the Sharpe ratios of the three different covered callwriting approaches and the outright stock investment for the entire sampleperiod.8 All of the three covered call writing approaches involve owning astock, but writing calls with different exercise prices. The first column lists theSharpe ratios for the OTM; the second column, for the ITM; and the thirdcolumn, for the DIV. It is noteworthy that the OTM has the highest upsidepotential, but the ITM has the most downside protection. The DIV balancesbetween the two. The test results at the bottom of the table show that theOTM and DIV strategies yield significantly higher Sharpe ratios than thestock investment at the 1% level, but that the ITM does not generate signifi-cantly higher Sharpe ratios at any meaningful statistical level. These findingssuggest that the ITM is questionable as a covered call writing strategy. Itseems as though the ITM offers too much protection. In other words, it is toomuch concerned with the protection of downside risk, but too little concernedwith the potential of returns. However, the other two covered call writingstrategies, OTM and DIV, provide significant benefits for investors. Theseresults seem to suggest that those investors, who need more downsideprotection than they receive from the typical covered call writing with out-of-the-money calls, should write half of the calls in-the-money and the other halfout-of-the-money, and not use exclusively in-the-money calls.


Research &

Theory

Table 1Covered Call Writing vs. Stock Investment: Monthly ReturnJanuary 1990 – December 2003Dow Jones Covered CallWriting a Stock InvestmentCompanies Average Standard Sharpe Average Standard Sharpe

Return Deviation Ratio b Return Deviation RatioAlcoa 1.08% 5.65% 0.13 1.31% 8.64% 0.11Altria Group 1.27% 6.20% 0.15 0.28% 8.94% 0.10AIG 1.65% 5.04% 0.26 1.26% 8.36% 0.11American Express 2.96% 6.42% 0.40 1.95% 8.85% 0.18Boeing 0.60% 7.11% 0.03 0.93% 9.78% 0.06Caterpillar 1.54% 6.41% 0.18 1.46% 8.62% 0.13Citigroup 1.39% 6.89% 0.15 2.31% 9.42% 0.21Coca Cola 1.23% 5.38% 0.16 1.29% 7.24% 0.13Disney 1.10% 6.23% 0.12 0.72% 7.22% 0.05Du Pont 1.02% 5.32% 0.12 0.72% 7.22% 0.05GE 1.24% 5.22% 0.17 1.32% 7.42% 0.13Hewlett Packard 1.32% 7.20% 0.13 1.41% 10.73% 0.10Home Depot 1.82% 6.42% 0.23 2.08% 9.66% 0.18Honeywell 1.86% 9.75% 0.15 1.20% 9.17% 0.09IBM 0.93% 6.12% 0.09 1.26% 9.89% 0.09Intel 1.76% 7.64% 0.18 2.77% 12.37% 0.19Johnson & Johnson 1.34% 4.64% 0.21 1.40% 6.86% 0.15McDonald’s 1.83% 6.02% 0.24 1.00% 7.42% 0.09Merck 2.28% 5.91% 0.33 0.79% 9.14% 0.05Microsoft 1.84% 5.95% 0.25 2.84% 9.76% 0.25M M M 1.18% 4.46% 0.18 0.75% 7.51% 0.05Pfizer 1.43% 5.49% 0.20 1.82% 8.39% 0.17Procter & Gamble 1.15% 5.33% 0.15 1.33% 7.43% 0.13Verizon 0.78% 5.14% 0.08 0.47% 6.71% 0.02

a This covered call writing involves owning a stock and writing a call optionwith a strike price right above the stock price (Out-of-the-Money approach).b The Sharpe ratio is computed by dividing the mean return on the covered callwriting less the mean return on Treasury bills by the standard deviation ofreturn on the covered call writing.

Hypothesis Testing:(1) Average Returns Wilcoxon signed rank test = 0.49 (p-value = 0.63)are equal. van der Waerden test = 0.59 (p-value = 0.55)(2) Standard Deviations Wilcoxon signed rank test = 4.24 (p-value = 0.00)are equal. van der Waerden test = -4.01 (p-value = 0.00)(3) Sharpe Ratios Wilcoxon signed rank test = 3.66 (p-value = 0.00)are equal. van der Waerden test = 3.57 (p-value = 0.00)



Table 2Three Types of Covered Call Writing vs. Stock Investment:Sharpe RatioJanuary 1990 – December 2003Dow Jones Covered Call: Covered Call: Covered Call: StockCompanies OTM a ITM b DIV c InvestmentAlcoa 0.13 0.14 0.14 0.11Altria Group 0.15 0.10 0.13 0.10AIG 0.26 0.31 0.29 0.11American Express 0.40 0.54 0.48 0.18Boeing 0.13 0.00 0.02 0.06Caterpillar 0.18 0.07 0.17 0.13Citigroup 0.15 0.06 0.15 0.21Coca Cola 0.16 0.10 0.14 0.13Disney 0.12 0.04 0.10 0.05Du Pont 0.12 0.14 0.13 0.05GE 0.17 0.12 0.18 0.13Hewlett Packard 0.13 0.08 0.11 0.10Home Depot 0.23 0.17 0.21 0.18Honeywell 0.15 0.03 0.13 0.09IBM 0.09 0.06 0.08 0.09Intel 0.18 0.10 0.16 0.19Johnson & Johnson 0.21 0.18 0.21 0.15McDonald’s 0.24 0.50 0.37 0.09Merck 0.33 0.59 0.45 0.05Microsoft 0.25 0.24 0.25 0.25MMM 0.18 0.15 0.17 0.05Pfizer 0.20 0.13 0.21 0.17Procter & Gamble 0.15 0.07 0.12 0.13Verizon 0.08 0.06 0.08 0.02

a This covered call writing involves owning a stock and writing a call optionwith a strike price right above the stock price (Out-of-the-Money approach).b This covered call writing involves owning a stock and writing a call optionwith a strike price right below the stock price (In-the-Money approach).c This covered call writing involves owning a stock and writing a call option50% with a strike price right above the stock price and 50% with a strike priceright below the stock price (Diversified approach).


Research &

Theory

Hypothesis Testing:(1) Sharpe ratios are equal Wilcoxon signed rank test = 3.66 (p-value = 0.00)between OTM and Stock. van der Waerden test = 3.57 (p-value = 0.00)(2) Sharpe ratios are equal Wilcoxon signed rank test = 0.26 (p-value = 0.80)between ITM and Stock. van der Waerden test = 0.56 (p-value = 0.57)(3) Sharpe ratios are equal Wilcoxon signed rank test = 3.56 (p-value = 0.00)between DIV and Stock. van der Waerden test = 3.04 (p-value = 0.00)

To determine if the benefits from the covered call writing strategieswere consistent throughout the sample period, the sample period was dividedinto two equal periods: one sub-sample period covering 1993-1999, and theother one covering 1990-1992 and 2000-2003, combined. The first sub-periodwitnessed a more bullish stock market than the second, and the latter includedboth the recession in the early 1990s and the recent recession known as the“dot com crash.” Table 3 presents the performance results for the morebullish sub-period. The results are qualitatively identical to those reported inTable 2. The OTM and DIV yield significantly higher Sharpe ratios than thestock investment at the 1% level; However, the Sharpe ratios are not signifi-cantly different between the ITM and the stock investment. These resultsconfirm that the OTM and DIV provide significant benefits for investors, butthe ITM does not. Once again, the covered call writing approach with in-the-money calls only is questionable.

Table 4 reports the results for the less bullish sub-period: 1990-1992and 2000-2003. The results are quite different from those in Table 2. None ofthe covered call strategies is significantly better than straight stock invest-ment in terms of Sharpe ratios. All of the test statistics are below the criticalvalues at the conventional significance level. Only the DIV generatesmarginally higher Sharpe ratios than the stock investment. The p-values forthe Wilcoxon signed rank test and the van der Waerden test are 0.11 and 0.08,respectively. Therefore, the results can be summarized as follows. During thisless bullish sub-period, covered call writing did not work particularly well.Among the three covered call strategies, however, the diversified approachworked the best. These results seem to suggest that covered call writingworks better in a more bullish condition.



Table 3Three Types of Covered Call Writing vs. Stock Investment:Sharpe RatioJanuary 1993 – December 1999Dow Jones Covered Call: Covered Call: Covered Call: StockCompanies OTM a ITM b DIV c InvestmentAlcoa 0.19 0.20 0.21 0.21Altria Group 0.06 0.02 0.05 -0.01AIG 0.47 0.67 0.60 0.19American Express 0.78 1.14 0.99 0.37Boeing 0.08 0.07 0.08 0.10Caterpillar 0.23 0.15 0.20 0.16Citigroup 0.33 0.24 0.39 0.30Coca Cola 0.15 0.12 0.15 0.16Disney 0.12 0.02 0.09 0.09Du Pont 0.17 0.13 0.16 0.17GE 0.48 0.38 0.53 0.38Hewlett Packard 0.37 0.37 0.38 0.24Home Depot 0.30 0.17 0.33 0.21Honeywell 0.28 0.31 0.32 0.25IBM 0.27 0.25 0.27 0.29Intel 0.31 0.21 0.29 0.30Johnson & Johnson 0.25 0.18 0.24 0.24McDonald’s 0.38 0.75 0.55 0.21Merck 0.77 1.27 1.02 0.20Microsoft 0.51 0.52 0.53 0.43MMM 0.12 0.03 0.09 0.09Pfizer 0.25 0.22 0.28 0.26Procter & Gamble 0.29 0.21 0.27 0.25Verizon 0.24 0.16 0.23 0.15

a This covered call writing involves owning a stock and writing a call optionwith a strike price right above the stock price (Out-of-the-Money approach).b This covered call writing involves owning a stock and writing a call optionwith a strike price right below the stock price (In-the-Money approach).c This covered call writing involves owning a stock and writing a call option50% with a strike price right above the stock price and 50% with a strike priceright below the stock price (Diversified approach).


Research &

Theory

Hypothesis Testing:(1) Sharpe ratios are equal Wilcoxon signed rank test = 3.49 (p-value = 0.00)between OTM and Stock. der Waerden test = 3.44 (p-value = 0.00)(2) Sharpe ratios are equal Wilcoxon signed rank test = 0.64 (p-value = 0.52)between ITM and Stock. van der Waerden test = 0.97 (p-value = 0.33)(3) Sharpe ratios are equal Wilcoxon signed rank test = 3.34 (p-value = 0.00)between DIV and Stock. van der Waerden test = 3.33 (p-value = 0.00)

Table 4Three Types of Covered Call Writing vs. Stock Investment:Sharpe RatioJanuary 1990 – December 1992 & January 2000 – December2003Dow Jones Covered Call: Covered Call: Covered Call: StockCompanies OTM a ITM b DIV c InvestmentAlcoa 0.08 0.11 0.10 0.03Altria Group 0.25 0.22 0.25 0.10AIG 0.08 0.08 0.08 -0.04American Express 0.12 0.14 0.13 0.14Boeing 0.00 -0.04 -0.02 0.03Caterpillar 0.15 0.00 0.14 0.07Citigroup 0.13 0.08 0.19 0.12Coca Cola 0.17 0.09 0.14 0.14Disney 0.12 0.06 0.10 -0.01Du Pont 0.09 0.15 0.11 0.01GE 0.00 0.02 0.06 0.06Hewlett Packard 0.00 -0.06 -0.03 0.04Home Depot 0.20 0.11 0.17 0.19Honeywell -0.01 -0.06 -0.03 0.02IBM -0.05 -0.06 -0.05 0.04Intel 0.11 0.04 0.08 0.09Johnson & Johnson 0.18 0.18 0.19 0.16McDonald’s 0.16 0.41 0.28 -0.02Merck -0.05 0.05 -0.01 0.03Microsoft 0.09 0.11 0.10 0.14MMM 0.24 0.25 0.25 0.02Pfizer 0.15 0.04 0.18 0.03Procter & Gamble 0.06 0.01 0.04 0.12Verizon -0.03 0.01 -0.02 -0.05

a This covered call writing involves owning a stock and writing a call optionwith a strike price right above the stock price (Out-of-the-Money approach).b This covered call writing involves owning a stock and writing a call option



with a strike price right below the stock price (In-the-Money approach).c This covered call writing involves owning a stock and writing a call option50% with a strike price right above the stock price and 50% with a strike priceright below the stock price (Diversified approach).

Hypothesis Testing:(1) Sharpe ratios are equal Wilcoxon signed rank test = 1.26 (p-value = 0.21)between OTM and Stock. van der Waerden test = 1.51 (p-value = 0.13)(2) Sharpe ratios are equal Wilcoxon signed rank test = 0.17 (p-value = 0.87)between ITM and Stock. van der Waerden test = 0.47 (p-value = 0.64)(3) Sharpe ratios are equal Wilcoxon signed rank test = 1.61 (p-value = 0.11)between DIV and Stock. van der Waerden test = 1.77 (p-value = 0.08)

Conclusion

The wide-spread market perception that options trading is too riskyfor most investors is not entirely justified. Some of the options tradingstrategies are not as risky or complicated as many believe. One good exampleis covered call writing, which involves owning a stock and writing the calloption on it. This study investigated, using DJIA companies, whether writingcovered calls provides a significant benefit over outright stock investment. Itwas found that even a simple strategy using out-of-the-money calls broughtsignificant benefits to investors in terms of risk and reward. The resultsdemonstrate that, during the period of 1990-2003, covered call writing reducedthe volatility of returns significantly, but did not significantly lower returns.

Two other covered call strategies were also investigated. Writingcovered calls with half of the position in out-of-the-money calls and the otherhalf in in-the-money calls worked as well as the covered call write with out-of-the-money calls only. In fact, the former worked even better than the latterduring the less bullish period. However, writing covered call with in-the-money calls only did not work well. It was found that the Sharpe ratios werevirtually the same between this approach and outright stock investment.

Encouraged by positive results from covered call writing, somefinancial planners and advisors may go as far as suggesting that this strategyshould be considered as a way to enhance the risk-return relationship. Theyargue that the approach is relatively safe and even brings some extra returns.Their argument, of course, is not true. If people claim that they can get ahigher return with lower risk, their motivation should be questioned. There isno such thing in the market. If there were a way to increase returns with lowerrisk, there would be a lot of people taking advantage of it, so that it wouldeventually disappear (e.g., Cox & Rubinstein, 1985). On the other hand, theresults from this study suggest that through covered call writing investors donot have to sacrifice return in order to reduce risk, and that this strategy canbe done without difficulties.


Research &

Theory

Endnotes

1 The first listed option contracts were offered by the Chicago Board OptionsExchange in 1973 and were traded on common stock.2 Covered call writing is the most popular options strategy (see Board,Sutcliffe, and Patrinos (2000)).3 For a variety of follow-up actions, see McMillan (2002).4 According to The Wall Street Journal (July 1, 2003), fewer than 10 percent ofbrokers use the covered call strategy.5 The decay of time value of call options (theta) is the greatest during the lastmonth. For theta, see McMillan (2002).6 Merton, Scholes, and Gladstein (1982) and Trennepohl, Booth, andTehranian (1988) show that the distribution returns do not follow a normaldistribution. In general, they are negatively skewed.7 One could raise a question as to whether the results would hold even aftercontrolling for transaction costs and taxes. First, major transaction costsinvolve bid-ask spreads and commissions. Bid-ask spreads were alreadyincorporated into the results of this paper. In other words, returns werecomputed based on the assumption that an investor pays a closing ask pricewhen buying a call, but receives a closing bid price when selling it. All theclosing bid and ask prices were obtained from the Dow Jones Call Optiondatabase provided by Quantitative Analytics, Inc. However, commissionswere not included in computing returns, for commissions vary dependingupon which brokers are used by investors, and more importantly, commis-sions have been infinitesimal since internet brokers were available in themarket (see Opdyke (2002)). For example, if the price of the stock is in the low$30’s, a covered call writing position involves approximately $30,000 ofinvestment, but commissions are only $25 or so. Second, taxes were notexplicitly considered in this paper. All the gains are assumed to be short-termand therefore subject to short-term gains tax. Normally, investors do not tendto hold a covered call position for a long term. They hold it for a relativelyshort term, typically several months, so a short-term gains tax is appropriate.Obviously, long-term gains tax would be appropriate for buy-and-holdinvestors. However, those investors, who insist on buy-and-holding a stock,should not write calls on the stock in the first place (see McMillan (2002)).Therefore, the results of the paper, given the short-term holding period, arenot qualitatively affected by taxes, even though the magnitude of all thereturns declines proportionally with taxes.8 Although we only report the Sharpe ratios, average returns and standarddeviation are available upon request.



References

Bittman, J. B. (2005). Options for the Stock Investor. New York: McGraw-Hill.Board, J., Sutcliffe, C., & Patrinos, E. (2000). The Performance of Covered

Calls. The European Journal of Finance, 6, 1-17.Conover, W. J. (1980). Practical Nonparametric Statistics. New York: John

Wiley & Sons.Cox, J. C., & Rubinstein, M. (1985). Options Markets. Upper Saddle River, NJ:

Prentice Hall.Fama, E. F., & French, K. R. (1992). The cross-section of expected stock

returns. Journal of Finance, 47, 427-465.Fama, E. F., & French, K. R. (1993). Common risk factors in the returns on

stocks and bonds. Journal of Financial Economics, 33, 3-56.Fama, E. F., & French, K. R. (1996). Multifactor explanations of asset pricing

anomalies. Journal of Finance, 51, 55-84.McMillan, L. G. (2002). Options As A Strategic Investment. New York: New

York Institute of Finance.Merton, R. C., Scholes, M. C., & Gladstein, M. L. (1982). The Returns and Risk

of Alternative Put Option Portfolio Investment Strategies. Journal ofBusiness, 55, 1-55.

Opdyke, J. D. (2002, October 8). Who Says You Can’t Make Money off StalledStocks? More Investors Use Options, Trading Upside for SteadyGains. The Wall Street Journal, C1.

Reighenstein, W. (1998). Calculating a family’s asset mix. Financial ServicesReview, 7, 195-206.

Schlosser, J. (2002, December 9) How to Tiptoe Back into the Market: TradingOptions Can Be a Great Way to Hedge Your Stock Bets,” Fortune,97-99.

Sharpe, W. F. (1966). Mutual fund performance. Journal of Business, 39, 119-138.

Sharpe, W. F. (1994). The Sharpe ratios. Journal of Portfolio Management, 21,49-58.

Trennepohl, G. L., Booth, L. R., & Tehranian, H. (1988). An Empirical Analysisof Insured Portfolio Strategies Using Listed Options. Journal of Fin

Contact Information: Ki C. Han, Professor of Finance, School of Management,Suffolk University, 8 Ashburton P1ace, Boston, MA 02108-2770; Phone: (617)573-8561; Fax: (617) 305-1755; E-mail: [email protected] Dadlani, Tax Consultant, Ernst & Young LLP., 200 Clarendon Street,Boston, MA 02116.


Research &

Theory

SUMMARY OF AUTHOR AND ARTICLE CITATIONS:JOURNAL OF PERSONAL FINANCE

VOLUMES 1, 2, 3, & 4

Theresa KasperEditorial Assistant

&John E. Grable, Ph.D., CFP®, RFC

Editor

Kasper and Grable (2005) published the first summary of author andarticle citations from the Journal of Personal Finance in Volume 4, Issue 2. Atthe time, Kasper and Grable argued that “The importance of documenting theimpact of one’s writing within the academic community continues to gainimportance as universities, colleges, and programs vie for resources andrecognition.” This statement may be truer today than in the past. Universityand college administrators are increasingly calling upon faculty to documentthe impact of their research and writing. The purpose of this article is toextend the citation analysis conducted by Kasper and Grable to include allpapers published in Volume 4. It is hoped that those who contribute to theJournal of Personal Finance will be able to use this information to documentthe important role the Journal is playing in the field, and to help facultymembers provide evidence of how their published work is being used toinfluence the body of knowledge related to personal finance, financialplanning, and financial counseling.

Table 1 lists the most cited articles used in papers published inVolume 4. Only articles that were cited at least two times are shown. This listof articles points to the diversity of sources used by those who publish in theJournal.



Table 1Most Cited articles in Volume 4 # of Citations

Helman, R. & Paladino, V. (2004, April). Will Americans ever become savers?The 14th retirement confidence survey, 2004. Issue Brief Number268.Washington DC: Employee Benefit Research Institute. RetrievedDec. 29, 2004 from www.ebri.org/rcs/2004/index.htm. 3

Yuh, Y., Montalto, C. P., & Hanna, S. (1998). Are Americans prepared forretirement? Financial Counseling and Planning, 1 (9), 1-13. 3

Bach, D. (2004). The automatic millionaire. New York: BroadwayBooks. 2

Charupat, N. & Deaves, R. (2004). How behavioral finance can assist financialprofessionals. Journal of Personal Finance, 3(3), 41-52. 2

Fischhoff, B., Slovic, P. and Lichtenstein, S. (1977), Knowing with CertaintyJournal of Experimental Psychology, 3, 552-564. 2

Garman, E.T, & Forgue, R. E. (2006). Personal finance, (8th ed.)..Boston:Houghton Mifflin Company. 2

Ibbotson Associates (2003). Stocks, Bonds, Bills, and Inflation 2003 Year-book. Chicago, Illinois: Author 2

Keown, A. (2001). Personal Finance, Turning Money into Wealth. InvestmentBasics (2nd ed., pp. 340-367). Upper Saddle River, New Jersey: PrenticeHall. 2

Malkiel, B. (2003). A Random Walk Down Wall Street, (8th ed.). New York, NY:W.W. Norton. 2

Mittra, S. (1995). Practicing financial planning: A complete guide forprofessionals. Michigan: Mittra & Associates. 2

Prochaska, J. O., DiClemente, C. C., & Norcross, J. C. (1992b). In Search of theStructure of Change. In Y. Klar, J. D. Fisher, J. M. Chinsky, & A. Nadler (Eds),Self Change: Social psychological and clinical perspectives (pp. 87-114).New York: Springer-Verlag. 2

Shiller, R. J. (2000), Irrational Exuberance. Princeton University Press Staëlvon Holstein, C.-A. S. (1972), Probabilistic Forecasting: An ExperimentRelated to theStock Market. Organization Behavior and Human 2


Research &

Theory

Table 2 lists the most cited first authors from Volume 4. Only firstauthors quoted at least twice are shown. In almost all cases, the authors withthe most first author citations had more than one article cited in multiplepapers published throughout the year. This list consists of familiar names forthose who follow research within the domains of personal finance, financialplanning, and financial counseling.

Table 2Most Cited First Authors, Volume 4

# of CitationsHanna, S. D. 6Prochaska, J.O. 6O’Neill, B. 5Bernheim, B.D. 4Bajtelsmit, V.L. 3Deaves, R. 3Everett, M.D. 3Fuhrmans, V. 3Garman, E.T. 3Helman, R. 3Lyons, A.C. 3National Endowment for Financial Education 3Yuh, Y. 3Bach, D. 2Barber, B. 2Bosworth, B. 2Charupat, N. 2Clark, R L. 2Dychtwald, K. 2Fischhof, B. 2Gist, J.R. 2Gutter, M. 2Hatcher, C.B. 2Hershey, D.A. 2Hogarth, J.M. 2Ibbotson Associates 2Jamieson, D. 2Joo, S. 2Kahneman, D. 2Keown, A.J. 2Kerrey, B. 2Kim, J. 2King, C.A. 2



Table 2 (continued)Most Cited First Authors, Volume 4

# of CitationsLi, J. P. 2Malkiel, B. 2Mittra, S. 2Rajput, M. 2Shiller, R.J. 2Vogel, N. 0. 2Xiao, J.J. 2

Table 3 extends the citation analysis to include the most cited author,regardless of position within the reference. In order to make the list an authormust have been quoted at least three times.

Table 3Most Cited Authors, Volume 4 (All Authors)

# of CitationsHanna, S.D. 12Deaves, R. 7Garman, E.T. 6Montalto, C.P. 6Prochaska, J.O. 6Bernheim, B.D. 5O’Neill, B. 5DiClemente, C.C. 4Hogarth, J. 4Joo, S. 4Odean, T. 4Anthony, M. 3Bajtelsmit, V.L. 3Bernasek, A. 3Fischhoff, B. 3Fox, J.J. 3Fuhrmans, V. 3Grable, J.E. 3Gutter, M. 3Helman, R. 3Kim, J. 3Kotlikoff, L. 3Lichtenstein, S. 3Lyons, A.C. 3National Endowment for Financial Education 3


Research &

Theory

Norcross, J.C. 3Paladino, V. 3Tversky, A. 3Xiao, J.J. 3Yuh, Y. 3Bach, D. 2Bagwell, D.C. 2Barber, B. 2Bassett, R.L. 2Beverly, S.G 2Bosworth, B. 2Charupat, N. 2Chen, P. 2Clark, R.L. 2d’Ambrosio, M.C. 2Dychtwald, K. 2Everett, M.D. 2Fan, J.X. 2Forgue, R.E. 2Garrett, D.M. 2Gist, J.R. 2Gokhale, J. 2Harris, J.C. 2Hatcher, C.B 2Hershey, D.A. 2Hilger, M.S. 2Ibbotson Associates 2Jamieson, D. 2Jianakoplos, N.A. 2Johnson, J.L. 2Kahneman, D. 2Keown, A.J. 2Kerrey, B. 2King, C.A. 2Lee, J. 2Leon, B. 2Leon, B. 2Li, J.P. 2Lüders, E. 2Malkiel, B.G. 2Mittra, S. 2Newman, B.M. 2Palmer, B.A. 2Prochaska, J.M. 2



Table 3 (continued)Most Cited Authors, Volume 4 (All Authors)

# of CitationsRajput, M. 2Sabelhaus, J. 2Scherpf, E. 2Shiller, R.J. 2Shovic, P. 2Sung, J. 2Thaler, R. 2Turner, J.A. 2Verma, S. 2Vogel, N.0. 2Wu, K. 2

Table 4 extends the work of Kasper and Grable (2005) to include allarticle citations from Volumes 1, 2, 3, and 4. A general theme emerges from thiscomprehensive list. Authors who publish in the Journal of Personal Financehave an interest in behavioral finance and the impact that financial riskattitudes have on personal finance topics. The list is limited to articles thatwere referenced at least three times.

Table 4Comprehensive Article and Author InformationMost Cited Articles, Volumes 1-4

# of Citations

Kahneman, D. & Tversky A. (1979). Prospect theory: An analysis of decisionunder risk. Econometrica, 47 (2), 263-292. 5

Malkiel, B. G. (multiple dates). A random walk down Wall Street. New York:Norton. 5

Garman, E. T., & Forgue, R.. (multiple dates). Personal finance Boston:Houghton-Mifflin. 4

Markowitz, H. M. (1952). Portfolio selection. Journal ofFinance, 1, 77-91. 4

Aizcorbe, A. M., Kennickell, A. B., & Moore, K. B. (2003). Recent changes inU.S. family finances: Evidence from the 1998 and 2001 survey of con-sumer finances. Federal Reserve Bulletin, 1-31. 3


Research &

Theory

Bajtelsmit, V. L., & Bernasek, A. (1996). Why do women invest differently thanmen? Financial Counseling and Planning, 7, 1-10. 3

Bajtelsmit, V. L., & VanDerhei, J. L. (1997). Risk aversion and pensioninvestment choices. In M. S. Gordon, O. S. Mitchell, & M. M. Twinney(Eds.), Positioning Pensions for the Twenty-First Century (pp. 45-66).Philadelphia, PA: University of Pennsylvania Press. 3

Cordell, D. (2001). RiskPACK: How to evaluate risk tolerance. Journal ofFinancial Planning, 14(6), (June), 36-40. 3

Drentea, P. & Lavrakas, P.J. (2000). Over the limit: The association amonghealth status, race, and debt. Social Science andMedicine, 50, 517-529. 3

Gutter, M., Fox, J., & Montalto, C. P. (1999). Racial differences in investordecision making. Financial Services Review, 8(3), 149-162. 3

Hariharan, G., Chapman, K. S., & Domian, D. L. (2000). Risk tolerance andasset allocations for investors nearing retirement. Financial ServicesReview, 9(2), 159-170. 3

Helman, R. & Paladino, V. (2004, April). Will Americans ever become savers?The 14th retirement confidence survey, 2004. Issue Brief Number 268.Washington DC: Employee Benefit Research Institute. Retrieved Dec. 29,2004 from www.ebri.org/rcs/2004/index.htm. 3

Horvath, P. & Zuckerman, M. (1993). Sensation seeking, risk appraisal, andrisky behavior. Personality & IndividualDifferences,14, 41-52. 3

Jianakoplos, N. A., & Bernasek, A. (1998). Are women more risk averse?Economic Inquiry, 36(4), 620-630. 3

Joo, S. (1998). Personal financial wellness and worker job productivity.Unpublished doctoral dissertation, Virginia Polytechnic Institute & StateUniversity, Blacksburg. 3

Kennickell, A. B. (2003). Codebook for 2001 Survey of Consumer Finances.Federal Reserve Board. Retrieved from http://www.federalreserve.gov/pubs/oss/oss2/2001/codebk2001.txt 3



Miller and Brian L. Shiker, Provision of 401(k) Plan Investment Advice: Trapfor the Unwary Financial Planner, 5 Journal of Financial Planning 30(June 2001) for a comprehensive treatment concerning compensationissues for ERISA fiduciaries. 3

Mittra, S. (1995). Practicing financial planning: A complete guide forprofessionals. Michigan: Mittra & Associates. 3

Powell, M. & Ansic, D. (1997). Gender differences in risk behavior in financialdecision-making: An experimental analysis. Journal of EconomicPsychology, 18, 605-628. 3

Sunden, A. E. &. Surette, B. J. (1998). Gender differences in the allocation ofassets in retirement savings plans. American Economic Review. 88, 207–211. 3

Sung, J. & Hanna, S. (1996). Factors related to risk-tolerance. FinancialCounseling and Planning, 7, 11-20. 3

Yuh, Y., Montalto, C. P., & Hanna, S. (1998). Are Americans prepared forretirement? Financial Counseling and Planning, 1 (9), 1-13. 3

Table 5 lists the most quoted first authors from Volumes 1, 2, 3, and 4.The list is very similar to the citation list published by Kasper and Grable(2005). Note that the list is limited to those authors with four or more firstauthor citations.

Table 5Most Cited 1st AuthorsVolumes 1-4

# of CitationsGrable, J.E. 17Hanna, S. 13O’Neill, B. 12Bajtelsmit, V. 9Kahneman, D. 9Prochaska, J.O. 9Garman, E.T. 8Joo, S-H. 8Kennickell, A.B. 8Social Security Administration 8Everett, M.D. 6Thaler, R. 6


Research &

Theory

Bernheim, B.D. 5Bodie, Z. 5Cordell, D.M. 5DeVaney, S.A. 5Gutter, M.S. 5Markowitz, H. 5Sung, J. 5TIAA-CREF 5Detweiler, G. 4Hershey, D.A. 4Hogarth, J.M. 4Jamieson, D. 4Kerry, B. 4Lee, J. 4Lee, J. 4Malkiel, B.G. 4Roszkowski, M.J. 4U.S. Census Bureau 4Xiao, J.J. 4Yuh, Y. 4

The final table (Table 6) shows the most cited authors from Volumes1, 2, 3, and 4. The list is very similar in composition to the one published byKasper and Grable (2005). This list of author citations points to the cross-referencing of citations of authors published in Financial Counseling andPlanning, Financial Services Review, Journal of Personal Finance, andother consumer oriented publications.

Table 6Most Cited AuthorVolumes 1-4

# of CitationsHanna, S. 28Grable, J.E. 24Joo, S. 20Garman, T.E. 16O’Neill, B.M. 15Montalto, C.P. 2Thaler, R. 12Tversky, A. 12Kahneman, D. 11Kennickell, A.B. 11



Hogarth, J.M. 10Prochaska, J.O. 10Bajtelsmit, V.L. 9DeVaney, S.A. 9Lytton, R.H. 9Bernasek, A. 8Deaves, R. 8Gutter, M. 8Social Security Administration 8Fox, J. 7Sung, J. 7Xiao, J.J. 7Anthony, M.S. 6Bernheim, B.D. 6Everett, M.D. 6Bodie, Z. 5Callan, V.J. 5Cordell, D.M. 5Fan, J.X. 5Fritz, R.L. 5Jianakoplos, N.A. 5Kim, J.E. 5Lee, J. 5Markowitz, H.M. 5Odean, T. 5Prochaska, J.M. 5Statman, M. 5Sundén, A.E. 5Surette, B.J. 5TIAA-CREF 5VanDerhei, J. 5Yao, R. 5Zuckerman, M. 5

Summary

The use of journal citation data by administrators in granting tenureand promotion on America’s college campuses is increasing. It is hoped thatthose who publish in and read the Journal of Personal Finance will findthese impact rankings useful in supporting their continued efforts to conductresearch in the areas of personal finance, financial planning, and financialcounseling.


Research &

Theory

Reference

Kasper, T., & Grable, J.E. (2005). Summary of author and article citations:Journal of Personal Finance, Volumes 1, 2, & 3. Journal of PersonalFinance, 4 (2), 85-91.

Contact Information: Theresa Kasper, Editorial Assistant, 318 Justin Hall,Kansas State University, Manhattan, KS 66506; E-mail:[email protected]



TRADITIONAL VERSUS ROTH 401(K): A BENEFITS

ANALYSIS

G. Eddy Birrer, PhD, CPAGonzaga University

Many individuals have made contributions to Roth IRA accountssince Congress passed the Economic Growth and Tax Relief ReconciliationAct (EGTRRA) of 2001. Several articles have been written citing advantagesof a Roth IRA in retirement and estate planning, most notably the absence ofminimum distribution requirements (except for beneficiaries), and tax-freedistributions from accounts. Although contributions to a Roth IRA are madewith after-tax dollars, individuals in current low marginal income tax bracketscan avoid paying taxes at potentially higher rates when future distributionsare received.

Beginning in 2006, provisions of the EGTRRA permit employers toadd a Roth 401(k) option to their retirement plan offerings. A Roth 401(k)offers many of the same advantages currently afforded by a Roth IRA, inaddition to other benefits. Employees will have to consider the relativeadvantages and disadvantages of a Roth 401(k) versus a traditional 401(k)when deciding how to designate future contributions in their employer-sponsored retirement savings programs. The decision is critical because anemployee’s designation of a Roth 401(k) contribution is irrevocable, subject-ing current wages to taxation, and affecting future rollover options.

This article provides a comparative analysis of income tax and relatedfeatures of a Roth 401(k) versus a traditional 401(k) and a Roth IRA. Financialplanners and retirement plan participants can use the information provided toguide them in deciding whether or not to recommend or select a Roth 401(k)option. Since a Roth option may also be made available by employerssponsoring 403(b) retirement plans, information contained in this article alsoapplies to those plans. State income tax provisions pertaining to retirementaccount distributions vary by state and are not addressed in this article, norare Roth 401(k) requirements applying to employers. Note, however, thatadministrative record-keeping and related issues may deter some employersfrom offering a Roth 401(k) plan option, at least in 2006. Employers, forexample, will have to maintain separate employee records for Roth 401(k)contributions and accumulations until the accounts are completely distrib-uted.

The accompanying illustration provides a summary of key income taxand related provisions of both traditional and Roth 401(k) contribution plans


Research &

Theory

and the Roth IRA. Comparisons are made from the viewpoint of employees.The illustration serves as a basis for the following discussion of the relativemerits of each type of retirement contribution plan, in particular a traditionalversus Roth 401(k).

Allowable Contributions

Perhaps the most notable benefit of a Roth 401(k) plan is the abilityto make contributions to a Roth account without regard to adjusted grossincome (AGI) limitations. Contributions to a Roth IRA are restricted to thosewith an AGI of $160,000 or less filing a joint federal tax return, and $110,000 orless if filing singly. Allowable contributions are phased out between AGIlevels of $150,000 to $160,000 joint, and $95,000 to $110,000 single, with nocontributions permitted for married taxpayers filing separate returns. Sincethere is no AGI limit for making contributions to a Roth 401(k), some highincome taxpayers now will have an opportunity to take advantage of thebenefits associated with accumulations in a Roth account. Unlike the case forRoth IRAs, married couples filing separate returns will be able to establishRoth 401(k) accounts.

Another highly favorable benefit of Roth 401(k) plans is that thedollar amount of allowable contributions is much greater than for a Roth IRA.In 2006, for example, contributions to a Roth IRA cannot exceed $4,000 perindividual plus a $1,000 catch-up contribution for those aged 50 and over.Contribution limits under a Roth 401(k) plan are the same as for a traditional401(k), that is, $15,000 plus a $5,000 catch-up contribution for those aged 50 orover. An individual can thus accumulate a significantly greater dollar amountin a Roth 401(k) account than in a Roth IRA account. To illustrate, assume anindividual contributes $15,000 per year for 30 years to a Roth 401(k) account,and that invested amounts earn an 8% annual yield. The accumulated total atthe end of year 30 would be $1,699,248. By comparison, the same individualcould contribute only $4,000 to a Roth IRA account, and then only if AGI waswithin allowable limits. Assuming the same 30-year period and 8% investmentyield, the individual would accumulate only $453,133 within the Roth IRA,more than $1,200,000 less than within the Roth 401(k)! Of course, an individualwho could contribute to both types of accounts would accumulate an evengreater amount of Roth savings.

As noted by Caudill (2005), the combined total of contributions anindividual makes to traditional and Roth 401(k) accounts may not exceed theallowable amount for either one of the accounts (p. 36)—again, in 2006,$15,000 plus a catch-up contribution. Individuals whose companies offer aRoth 401(k), and who also meet AGI limits for a Roth IRA, may nonethelesscontribute up to $19,000 (plus $6,000 in catch-up contributions) to Rothaccounts in 2006. A married couple aged 50 or over conceivably could



contribute a total of $50,000 ($15,000 each to Roth 401(k) accounts, $4,000each to Roth IRAs, plus catch-up contributions of $5,000 and $1,000 each tothe respective accounts). Contributions can be made to a Roth 401(k),however, only while an individual is employed by a company offering a Roth401(k) plan option. On the other hand, contributions can be made to a RothIRA in any year so long as an individual meets AGI limitations and has earnedincome at least equal to the contribution.

Taxation of Contributions and Distributions

The ability to make significant dollar contributions to a Roth 401(k)account is not without cost. Designated contributions are made with after-taxdollars; that is, Roth 401(k) contributions are included in wages subject tofederal income tax withholding, and are taxed currently. Contributions to atraditional 401(k) account, on the other hand, are made with pre-tax dollars;that is, the contributions are excluded from wages subject to withholding,thereby reducing current taxable income. Individuals seeking to maximizecontributions to a Roth 401(k) account, therefore, need sufficient discretion-ary income to pay the higher dollar amount of taxes due in the year contribu-tions are made to a Roth 401(k) instead of to a traditional 401(k) account.

Investments and earnings within both types of 401(k) accountsaccumulate tax-free. As for a Roth IRA, no income tax is due on qualifieddistributions (discussed later) from a Roth 401(k) account. Distributions froma traditional 401(k) account, however, are fully taxable as ordinary income.Given the same accumulated retirement savings under both types of accounts,the traditional 401(k) will result in a smaller distribution.

Mathematically, it can be demonstrated that the after-tax benefit ofboth a Roth 401(k) and a traditional 401(k) would be the same over time if anindividual is in the same tax bracket when making contributions as whenreceiving distributions. For example, assuming a constant 25% tax rate, aninitial $4,000 contribution to a traditional 401(k) account, invested for 25 yearsat 8%, will accumulate to $27,394. The after-tax distribution will then be$20,546 ($27,394 minus the 25% tax of $6,848). Assume the $1,000 current taxsavings arising from the initial $4,000 contribution to the traditional 401(k) canbe invested in either a Roth 401(k) or Roth IRA account, also for 25 years at8%. The $1,000 investment will accumulate to provide a tax-free distribution of$6,848, an amount equal to the tax paid on the traditional 401(k) distribution.The total after-tax benefit of the combined traditional 401(k) and Roth IRAaccumulations is thus $27,394 ($20,546 plus $6,848). By comparison, an initial$4,000 contributed instead to a Roth 401(k) account, also invested for 25 yearsat 8%, would result in an accumulation of $27,394 distributable tax free, thesame benefit as for the traditional 401(k) plus Roth IRA combination.


Research &

Theory

Referring to the preceding examples, assume an individual does nothave sufficient discretionary income to invest the $1,000 taxes saved initiallyunder the traditional 401(k). The end result in both cases will still equal$20,546, because the individual presumably could afford to invest only $3,000of after-tax dollars in a Roth 401(k) or Roth IRA. As observed by Caudill(2005), “the Roth election will be more advantageous to employees who candefer as much on an after-tax basis as they would have deferred pretax” (p.38).The overall results could differ, however, if an individual’s pre-tax contributionto a traditional 401(k), including any catch-up amount, resulted in current taxsavings in excess of allowable contributions to a Roth IRA or Roth 401(k).

The decision to contribute to a Roth 401(k) versus traditional 401(k),therefore, is in large part a function of expectations concerning future incometax rates relative to current income tax rates. Individuals whose currentmarginal income tax rate is lower than expected future tax rates, for example,could benefit by making Roth 401(k) and Roth IRA contributions. Thiscategory might include young professionals who anticipate considerablyhigher future earnings and retirement income. Given the current low level offederal tax rates plus the large federal deficit, it may be rational for others alsoto assume their future tax brackets may rise above current or pre-EGTRRAlevels.

Other examples concerning current versus expected tax rates shouldalso be noted. First, consider individuals with gross income greater than theAGI limits for a Roth IRA whose goal is to minimize current taxes, yet who alsowant to contribute to a Roth IRA. They might still achieve their goal byincreasing their pre-tax contributions to a traditional 401(k) to reduce their AGIbelow the limits set for a Roth IRA contribution.

A second situation involves expenditures such as medical costs thatcan be treated as itemized deductions on a tax return only if they exceed aspecified percentage of AGI. In a given year, individuals with high medical orother AGI-percentage-based itemized deductions might realize greater overalltax savings long term by maximizing contributions to a traditional 401(k) toreduce AGI instead of contributing to a Roth 401(k). Short-term tax savingsassociated with higher current itemized deductions may offset potentiallyhigher future taxes.

Two final cases concern current low income individuals. Thosewhose current marginal tax rate is 10% receive only nominal benefit from pre-tax contributions to a traditional 401(k). A Roth 401(k) thus may be a betteroption for those low income individuals because they can lock in a currentlow tax rate and be unconcerned about future earnings or tax rate changes. Onthe other hand, low income individuals may qualify for a tax credit rangingfrom 10% to 50% of their retirement savings contributions up to $2,000 if theirAGI is less than $50,000 in 2006 (the last year credits are available). For someof those individuals, therefore, it may be beneficial to contribute to a tradi-



tional 401(k) rather than to a Roth 401(k) if necessary to reduce their AGIbelow limits established for the savings tax credit. The combined tax creditplus current tax savings on pre-tax contributions to a traditional 401(k) maylikely exceed any tax on future distributions.

Designated contributions to a Roth 401(k) are irrevocable, soindividuals should exercise care when estimating current versus future taxsavings and other benefits. One alternative might be to hedge tax rateexpectations, making some contributions to a Roth 401(k) account (or to asupplementary Roth IRA) and some to a traditional 401(k). As noted by theTransamerica Center for Retirement Studies (TCRS, 2005), regardless of howan employee may designate contributions within a company retirement plan,an employer’s matching contribution is always treated as being made on a pre-tax basis (p. 2). A certain amount of tax hedging, therefore, is implicit evenwhen contributing to a Roth 401(k) since an employer’s matching contribu-tions and related earnings will be taxed upon distribution to the individualemployee.

Qualified Distributions, Minimum Distribution Requirements,and Rollovers

Limitations are placed on distributions from qualified retirementsavings plans to discourage their use for other than retirement purposes.Early distributions (those made before age 59.5) from a traditional 401(k) aresubject to a 10% penalty tax in addition to ordinary income taxes. As outlinedby the IRS (Topic 558, n.d.), the penalty does not apply to qualifying distribu-tions made because of death or disability, separation from service at age 55 orlater, and certain other cases, including rollovers to other qualified plans (p.1).A traditional 401(k) account balance cannot be rolled over to a Roth 401(k) orRoth IRA, however.

Less clear are rules governing taxation of early distributions from aRoth 401(k) not treated as qualifying distributions. Young (2005) reported thatas a consequence some “companies are waiting for the Internal RevenueService [IRS] to clear up . . . how early withdrawals will be taxed” beforeoffering a Roth 401(k) option (p. 116). Since both a 401(k) and Roth 401(k) arequalified retirements plans, the taxation of early distributions should beconsistent. Thus, only early distributions of Roth 401(k) earnings should besubject to both ordinary taxes and a 10% penalty tax. This view agrees with anIRS (FAQs, n.d.) response to a question about early distributions, specificallythat a “hardship distribution . . . (of) earnings will be included in gross incomeunless you have had the Roth 401(k) account for 5 years and are eitherdisabled or over age 59.5” (p. 3). Under Internal Revenue Code section 72(t)(1)the 10% penalty tax is applied only to amounts included in gross income.


Research &

Theory

There should be no tax on distributions of Roth 401(k) contributions, becausethe contributions were made on an after-tax basis.

In some cases individuals may be better off making contributions toa Roth IRA account instead of to a Roth 401(k), especially if they are planningtotal Roth contributions less than allowable Roth IRA limits. Distributionsfrom a Roth IRA that are a return of capital (not earnings) may be made tax-free for any purpose at any time. Qualifying distributions under a Roth IRA,furthermore, may include those made for first time home buyer expenses. Incontrast, only qualifying distributions may be made from Roth 401(k) ac-counts, and then only if permitted under the retirement plan.

Distribution rules for a Roth 401(k) after age 59.5 differ from those fora Roth IRA. Both types of plans are subject to a five-year aging requirement,described by the IRS (Tax Trails, n.d.) as “the 5-year taxable period beginningwith the first taxable year for which a contribution is made” (p. 1). Under aRoth IRA there is no minimum distribution requirement (MDR). An individualmay take a tax-free distribution of any and all accumulations in the account atany time for any purpose after age 59.5, provided the five-year aging require-ment is met. Alternatively, the individual could choose not to take anydistributions from a Roth IRA, leaving the account to continue growing tax-free for future use. Amounts not distributed prior to the account owner’sdeath would then be distributed tax-free to a named beneficiary followingMDR rules. Instead of receiving distributions over his or her lifetime, abeneficiary could, for example, elect to take a lump sum distribution to paytaxes on the estate of the deceased, thus preserving other estate assets.

Distributions under a Roth 401(k) plan after age 59.5 are morerestrictive than for a Roth IRA. Tax-free distributions may be made from aRoth 401(k) only if the individual has separated from service or if distributionsare allowed under terms of the plan. Furthermore, a Roth 401(k) has a MDRparalleling that for a traditional 401(k). The MDR rules require individuals tobegin taking distributions at age 70.5 (technically, by April 1 of the year afterthey turn 70.5, or upon retirement, whichever is later), and a 50% penalty taxlevied on any excess of required distributions over actual distributions. Anindividual can avoid the MDR rules and potential penalties by making a tax-free rollover of the accumulated Roth 401(k) account balance to a Roth IRAupon retirement (or earlier date if allowed under the plan).

A Roth 401(k) rollover compares favorably with rolling over anaccumulated balance in a traditional 401(k). While the latter may be rolled overtax-free to a traditional IRA, the account is subject to MDR rules. If anindividual then wants to roll over the traditional IRA balance to a Roth IRAaccount, the entire rollover amount is subject to taxation at ordinary incomerates—and certainly the rollover amount will push an individual into a highertax bracket. Finally, the rollover to a Roth IRA from a traditional IRA in thiscase is limited to taxpayers with an AGI of $100,000 or less (excluding the



amount rolled over), regardless of filing status, and is not permitted formarried couples filing separate returns.

Conclusion

Beginning in 2006, high income taxpayers not eligible to contribute toa Roth IRA may contribute instead to a Roth 401(k) account. The amount ofallowable contributions, furthermore, is much greater than for a Roth IRA.Individuals can thus enjoy Roth-related benefits, including significantlygreater accumulations of tax-free funds for estate planning and other pur-poses, and avoidance of potentially higher future tax rates. Certain low incomeindividuals may also benefit by contributing to a Roth 401(k).

Designation of contributions to a Roth 401(k) account is irrevocable,however, and should be made only after careful consideration of alternatives.Limitations of a Roth 401(k) include current taxation of contributions, anddistribution options more restrictive than under a Roth IRA. Contributionsmay be distributed tax-free from a Roth IRA at any time, with both earningsand contributions available tax-free at any time after age 59.5 provided thefive-year aging requirement is met. Distributions of Roth 401(k) accumula-tions, however, parallel those for a traditional 401(k) plan, including beingsubject to MDR rules (unless rolled over to a Roth IRA).

Provisions of EGTRRA are set to expire at the end of year 2010unless extended by Congress. While it is likely individuals will still be able tocontribute to Roth accounts after 2010, the ability to do so is not guaranteed.Individuals seeking the benefits of Roth accounts might therefore be well-advised to make related contributions during the next five years provisions ofthe EGTRRA continue in effect.


Research &

Theory

References

Caudill, A. K. (2005, September). Planning for the Roth 401(k) feature. Journalof Financial Services Professionals, 36-38.

Department of the Treasury, Internal Revenue Service. (2005, March 2).Designated Roth Contributions to Cash or Deferred ArrangementsUnder Section 401(k). Federal Register, 70 (40), 10062-10066.

FAQs Regarding Roth 401(k) Contributions. (Retrieved January 13, 2006).Available at: http//www.irs.gov/retirement/article/0,,id=152956,00.html

Tax Trails - Roth IRA Distributions. (Retrieved November 1, 2005). Availableat: http//www.irs.gov/individuals/article/0,,id=133720,00.html, 1.

TCRS 2005-02: Roth 401(k). Transamerica Center for Retirement Studies, 1-3.Topic 558 - Tax on Early Distributions from Retirement Plans. (Retrieved

November 1, 2005). Available at: http//www.irs.gov/taxtopics/tc558.html, 1-2.

Young, L. (2005, October 17). Why the Roth 401(k) is a ‘Dud’ so far. BusinessWeek, 116.



Comparison of Retirement Contribution Plans

Federal Income Traditional 401(k) Roth 401 (k) Roth IRATax Provision (2006)Employee funding Pre-tax contributions After-tax After-tax

contributions contributionsEmployer matching Pre-tax contributions Pre-tax Not applicable

contributionsAllowable $15,000 plus $5,000 $15,000 plus $5,000 $4,000 pluscontributions catch-up catch-up $1,000 catch-up

contribution* contribution* contributionAdjusted gross None None $110,000 singleincome limitations $160,000 jointAllowable Only during term Only during term Any years ancontribution of employment of employment individual hasperiods earned incomeQualified Yes, all amounts Only employer Nodistributions matchingafter age 59.5 contributions andsubject to tax? related earningsMinimum Yes, starting at later Yes, starting at later Nonedistribution of age 70.5 or of age 70.5 orrequirement? retirement retirementRollover options Tax-free rollover to Tax-free rollover to Not applicable

traditional IRA; then another Roth 401 (k)taxable rollover to plan or to Roth IRARoth IRA from IRAif AGI less than$100,000

Five year aging No Yes YesrequirementDistribution of No; time of No; time of Yesemployee distribution is distribution iscontributions at any restricted, and all restrictedtime without tax distributions are

taxable*Traditional and Roth 401 (k) contributions combined cannot exceed $15,000 + $5,000catch-up amount

Contact Information: G. Eddy Birrer, PhD, CPA, Professor of Accounting,School of Business Administration, Gonzaga University, Spokane, WA 99258;Phone: (509) 323-3435; E-mail: [email protected]

Volume 5, Issue 2

Book R

eview

121

BOOK REVIEW

Blink: The Power of Thinking without Thinking

Author: Malcom Gladwell

Reviewer: Tim Griesdorn, MBA, CFM, CMA

Is it possible to know when a client is lying to you? Can you tell ifsomeone is a good teacher after watching them for only a couple of minutes?Is it possible to predict if a couple will get divorced with 95% accuracy afterwatching them for only 60 minutes? In Malcom Gladwell’s new book “Blink”he asserts it is not only possible to know these things but much more!Gladwell sets out to accomplish three tasks in his book Blink. The first taskof Blink is to convince you of a simple fact: “Decisions made very quickly canbe every bit as good as decisions made cautiously and deliberately.” Thesecond task of Blink is to answer the question, “When should we trust ourinstincts, and when should we be wary of them?” The final and most impor-tant task of Blink is “To convince you that our snap judgments and firstimpressions can be controlled.”

To tackle the first task Gladwell uses a concept he calls “thin slicing.”Thin slicing is rapid cognition – what some would call intuition – based upona very limited amount of information and/or time. He reports the results ofseveral research studies that have honed this technique into a fine art withamazing results. The first research study is based upon psychologist JohnGottman’s work with young married couples. Gottman was able to predictwith 95% accuracy which couples would divorce within the next 15 years afterwatching them for just 60 minutes. Gottman found there are four negativeemotions to look for when couples interact. They are contempt, criticism,stonewalling, and defensiveness. Gottman noted that “once a couples’ levelof emotion starts going down towards negative emotion, ninety-four percent



will continue going down.” In addition, there is one emotion of the four thatis the best predictor of marriage trouble ahead. The emotion is contempt.Contempt tries to put the other person on a lower level than you. “Contemptis closely related to disgust, and what disgust and contempt are about iscompletely rejecting and excluding someone from the community.” If con-tempt or talking down to a spouse is so damaging to a marriage, think what itcould do to other relationships? If we talk down to our clients or generate anair of superiority and aloofness, we lose our opportunity to learn from thatperson. Contempt can prevent relationships of trust from forming.

Gladwell goes on to report the results of a study by psychologistSamuel Gosling. Gosling wanted to determine who would be more accurate indetermining a person’s personality traits of conscientiousness, emotionalstability, and openness to new experiences. He gave a questionnaire to theclose friends of college students, and he gave the same questionnaire to totalstrangers who were allowed to view the student’s dorm room for 15-minutes.The results, total strangers used thin slicing techniques and were moreaccurate in determining the student’s personality traits of conscientiousness,emotional stability and openness to new experiences than close friends.Why? Because thin slicing can eliminate the extraneous noise and allow ourminds to focus on the most important details. Who knows, maybe a 15 minuteinspection of a bedroom could replace the need for reference checks in thefuture!

Gladwell does a great job in his first task of convincing the reader“Decisions made very quickly can be every bit as good as decisions madecautiously and deliberately.” Unfortunately, the book is not as robust when itcomes to the other two tasks. His second task was to help the reader knowwhen he or she should trust their instincts and when they should be wary ofthem. To do this, Gladwell describes the Warren Harding error. Gladwellwarns first impressions rooting in stereotypes or prejudice should be dis-counted. The book could have been improved if Gladwell spent a little moretime helping the reader identify possible sources of bias other than theobvious prejudices. What he fails to identify are other scenarios where firstimpressions could lead to incorrect conclusions. For example, from behavioralfinance research it is known that overconfidence in one’s abilities can behazardous to your wealth. In addition, sunk cost fallacy, endowment effect,and bigness bias suggest that gut instinct can sometimes be wrong. In thisbook, Gladwell needed to take some more time to delve into the topic of whento be wary of thin slicing results. Instead, he ends up with a conclusion thatsounds a little too trite – “Truly successful decision making relies on abalance between deliberate and instinctive thinking.” And, “Take complexproblems and reduce it to its simplistic elements.” He provides only oneexample of a successful car salesman who is able to control his first impres-sion of customers. Adding additional examples like this one in other aspects

Volume 5, Issue 2

Book R

eview

123

of life would have added to the convincing power of his precepts. The authordoes offer a good suggestion for anyone who desires to overcome prejudice.To overcome prejudice you need to interact with minorities regularly andbecome more acquainted with the best that another culture has to offer. Forexample, if you want to reduce the effect of social prejudices developedagainst blacks, then study the lives of people like Martin Luther King Jr., ColinPowell, and Nelson Mandela.

Gladwell’s third task is, “to convince you that our snap judgmentsand first impressions can be controlled.” He reports results of a study thatshows novices in thin slicing when told what to look for in a couples’interactions were accurate in predicting which couples would divorce betterthan 80% of the time. He also sights examples of police officers who knewhow to gage a potential threat properly and army officers who knew when toreact to save the lives of their men. The example of the car salesman alsoshows that controlling first impressions can lead to a more successfulbusiness. Unfortunately, the book provides only these somewhat vaguegeneralities and does not take the difficult next step on how to apply andintegrate this knowledge into our daily lives. While the book providesexamples of people who have been able to control their snap judgments, hedoes not give the reader specific steps to follow to increase their skills in thisarea.

Another section of the book that could have benefited from moredetails was the discussion of priming. Gladwell spent eight pages on the topicbut more could have been written. Priming is the ability to impact a person’sperformance at a task without their conscious knowledge that they have beenaffected. In experiments conducted by John Bargh, people were asked tomake a grammatical four word sentence out of scrambled words. Bargh thenshowed that people could be primed with certain behavioral traits withouttheir conscious knowledge of what was being done. Bargh did studies ofNew Yorkers who were primed with politeness or rudeness to see how longthey would wait before interrupting another person’s conversation. Thepeople primed with rudeness only waited an average of five minutes prior tointerrupting. The vast majority of people primed with politeness did notinterrupt the conversation at all and waited a full 10 minutes before theexperimental time limit was reached. Another study done by Steele andAronson gave black college students 20 questions from the GRE exam.“When the students were asked to identify their race on a pretest question-naire, that simple act was sufficient to prime them with all the negativestereotypes associated with African Americans and academic achievement –and the number of items they got right was cut in half.” When asked after-wards if anything lowered their performance, the students had no consciousawareness of the priming effect. It is also possible to prime yourself forsuccess. Try repeating what two Dutch researchers did the next time you sit



Contact Information: Tim Griesdorn, MBA, Department of Applied andProfessional Studies, Box 41162 College of Human Sciences, Texas TechUniversity, Lubbock, TX 79409-1162; Phone: (806) 742-5050; E-mail:tim.griesdorn.edu.

down to play Trivial Pursuit. “Take five minutes beforehand to think aboutwhat it would mean to be a professor and write down everything that came tomind.” The students who did this got 55.6% of the questions asked correctcompared to 42.6% for those who thought about soccer hooligans prior to thequestions. As athletes can tell you, our frame of mind can have an impact onour performance.

In addition to helping your Trivial Pursuit game, the book evenprovides a logic puzzle that those interested in personal finance may findinteresting. “A giant inverted steel pyramid is perfectly balanced on its point.Any movement of the pyramid will cause it to topple over. Underneath thepyramid is a $100 bill. How do you remove the bill without disturbing thepyramid?” To find out, you’ll have to read the book; the answer is on page121.

Overall, Gladwell did an excellent job of taking rather esotericpsychological research studies, mining the main points out of them, andpresenting the results in a way that the average person can understand andlearn from. In addition, Gladwell does all this in an enjoyable, conversationaltone that allows the material to flow together to support his main points. Thewriting captivates the reader from the introduction. The book is a great choicefor anyone who desires to understand more about how the human mind worksand processes information on both a conscious and unconscious level. If aparticular subject piques your interest, Gladwell has a great notes section thatsites the sources of the various research studies that he used in the book.This section is a useful reference source for some of the latest in psychologi-cal research. For those who enjoy this book Gladwell’s other work may alsobe of interest (i.e., The Tipping Point: How Little Things Can Make a BigDifference).

Purchase Information: Gladwell, M. (2005). Blink: The power of thinkingwithout thinking. Boston: Little, Brown & Company. IBSN 780316172325; thebook may be purchased at BarnesandNoble.com for $25.95 in hardcover andon CD for $39.98.

Institutional Profile


INSTITUTIONAL PROFILE

Financial Planning

At

Western Carolina University

Western Carolina University, a regional comprehensive universitywithin the University of North Carolina system, is located on 589 acres in apicturesque valley between the Blue Ridge and Great Smoky mountains insouthwestern North Carolina. Western offers career-focused educationthrough nationally accredited programs, as well as abundant opportunities forresearch and recreation in the nearby mountains.

Western’s student enrollment, including approximately 8,400undergraduate, graduate and online students, is at an all-time high. Theuniversity attracts students from across North Carolina and the United States,and from around the world, with special emphasis on serving the people ofWestern North Carolina. As a major public resource, Western engages withthe region by assisting individuals and agencies through the expertise of itsstudents, faculty and staff.



With a student-faculty ratio of 16-to-1, Western continues its longtradition of offering small class sizes and personal attention from its outstand-ing faculty. The fully networked campus offers convenient access to computersupport, and students are required to have networkable computers. Theuniversity offers more than 120 majors and areas of concentration for under-graduates, and graduate students may choose from 51 areas of study at themaster’s level, as well as pursue doctoral studies in education.

Western’s residential Honors College, the first of its kind in the NorthCarolina system, now enrolls more than 900 students, providing a focusedliving and learning environment for those seeking advanced academicchallenges and more choices for independent study. Honors Scholars areexceptional students who take tutorial classes to prepare for regional andnational research projects and compete for national and international scholar-ships.

Freshmen admitted to Western are automatically considered foravailable scholarships, based on their academic achievements. The universityannually awards $1 million in merit-based scholarships and more than $30million in needs-based aid to students. About half of the university’s studentsreceive financial assistance through scholarships, grants, loans or work-study.

Students participate in a wide range of non-academic activities,including clubs and honor societies, community service and volunteerorganizations, concerts and performances at various venues on campus,Greek life, student-operated media and student government. The universityalso sponsors outdoor recreational programs such as hiking, rafting and snowskiing that take advantage of the university’s location in one of the nation’smost popular vacation areas. Western also offers club sports and intramuralsthat involve about one-quarter of the student body.

With the new millennium has come a new campus. Visitors toWestern’s campus today will find a radically different campus than the onethat existed five years ago. With voters’ approval of a statewide highereducation bond package in fall 2000, the university launched into a buildingboom of a magnitude never before seen on campus.

Western’s share of the bonds — $98 million – represents the largestsingle infusion of funds in the university’s history. Combined with federalfunds and self-supporting projects, the $130 million in new and renovatedbuildings and roads is redrawing the campus map as the university preparesfor projected enrollment growth in the decade ahead.

The new $30 million Fine and Performing Arts Center is amongprojects funded through state bonds. The 122,000-square-foot facility, justcompleted, includes classrooms and studios for students majoring in the artsand humanities, as well as public galleries and a 1,000-seat hall for Broadway-quality music and theatrical performances.



Other new facilities already completed at Western are three newresidence halls that offer modern conveniences and comfort for students; theCenter for Applied Technology, which houses audio and video productionstudios and engineering laboratories; an expanded student life center; newand improved athletics facilities that support Western’s intercollegiate sportsprograms; and a new One Stop Student Support Center that provides a quickand convenient way for students to receive help and conduct businessinvolving financial aid, registration, cashiers services and many otheradministrative services. Construction on a new student recreation center isscheduled to begin in the near future.

The College of Business’s four academic departments – Accounting,Finance and Entrepreneur-ship; Business Computer Information Systems andEconomics; Management, International Business and Hospitality andTourism; and Marketing and Business Law – offer nine, undergraduateacademic programs. The College’s sixty-plus faculty members place thehighest priority on creating a student-centered learning environment. Thatmeans working closely with each student, whether the task is advising,setting up an internship, or solving a homework dilemma. For faculty, student-centered learning also means interacting constantly with the region’sbusiness leaders to increase student opportunities – perhaps to arrange a tourof a manufacturing facility, find a speaker for one of the department’s clubs, orkeep up with the latest applications of a new workplace theory. Student-

2001 Second Place National Champions



centered learning means the flexibility to continually update the curriculum toensure that it offers the tools and skills for today’s global business world.

In the College’s programs, theory is blended with practical experiencethrough a variety of innovative teaching techniques and participation inprofessionally related clubs, internships, and research with faculty. Allundergraduate business students must complete the Foundation of Knowl-edge for Business. Students pursuing the 120 hour Bachelor of Science inBusiness Administration (B.S.B.A.) select from the following major programsof study: Accounting, Business Administration and Law, Computer Informa-tion Systems, Entrepreneurship, Finance, Hospitality and Tourism, Manage-ment, and Marketing. Several majors offer concentrations that allow addi-tional specialization. In addition to the undergraduate academic programsthere are four master level programs – masters of business administration,masters of accountancy, masters of project management and masters ofentrepreneurship. Both the masters of project management and entrepreneur-ship are on-line programs.

The finance major has had a concentration in financial planning andbank management. In 1996 the financial planning program was registered withthe Certified Financial Planning Board of Standards. Since becoming aregistered program, there have been many positive changes in the curriculumand increased opportunities available to the students. These factors, alongwith the exponential growth in the financial planning profession, have




resulted in the numbers of students choosing this concentration to more thandouble. WCU students majoring in finance with a concentration in financialplanning have been very successful in the job market upon graduation. Manygraduates begin their career in the banking industry as a financial specialist.Others receive job offers from financial planning firms, insurance companies,mutual fund companies and brokerage firms. Although most financialplanning majors seek jobs in the financial services industry, some choose jobsin corporate finance as well as other areas of business.

The BSBA in Finance with a concentration in financial planning is afour-year degree offered by the Department of Accounting, Finance andEntrepreneurship. Students may complement the curriculum by doublemajoring or by choosing recommended electives. Students often doublemajor in accounting, banking or marketing. Some of the most relevant andbeneficial electives are: professional selling, consultative selling, portfoliomanagement, financial institutions and markets, financial statement analysis,bank management and individual and group counseling. Students are highlyencouraged to complete an internship prior to graduation. Feedback, as wellas permanent job offers from companies where students intern, indicates thatour financial planning students are well qualified for a career in the financialservices market.




There are many beyond the classroom activities that students mayparticipate in to enhance their financial planning education. These activitiesinclude the Finance Club, the Financial Planning Association (FPA), theTennessee Valley Authority (TVA) Investment Challenge, the AmeriprisePlanning Invitational and a four-day educational field trip to New York City.

The TVA Investment Challenge is a student managed portfoliocompetition with twenty-five other universities. WCU finance studentsmange a portfolio in excess of $600,000. TVA requires that the studentsuphold their fiduciary responsibility by managing the portfolio ethically andprofessionally as if they were Wall Street money managers. Since 1998, thisportfolio has outperformed the S&P 500 seventy-five percent of the time. Thishands-on experience is invaluable for students entering the financial servicesprofession as well as any individual interested in learning how to analyze andmanage a personal portfolio. WCU students have eagerly participated in thethree-stage Ameriprise Planning Invitational each year. Appling knowledgefrom the classroom, the students write a financial plan for a fictitious family,present orally the plan and compete in a Jeopardy type contest. In three ofseven years, WCU students have finished second in this national competitiondemonstrating their knowledge of financial planning as well as their abilityand confidence to perform on a national level in a competitive setting.

Each fall Dr. Grace Allen takes approximately 15 students to New YorkCity. The highlight of this trip is the visit to the New York Stock Exchange. Inaddition the students find it fascinating to discuss portfolio management andpotential careers with financial professionals at some of the largest Wall Streetinvestment banks.

In addition to the undergraduate degree program in financialplanning, WCU also has a certificate program in financial planning. This nine-month program is taught in Asheville, the largest city in western NorthCarolina. Students can chose to attend all class meetings or do a combinationof in-class/on-line study. This allows professionals the flexibility to haveinstructor contact when desired and self-study when class attendance is notpossible due to work or personal obligations.

The faculty in the financial planning concentration are highlyqualified in their areas of interest, teaching and research. Grace Allen, theprogram director holds a Ph.D. in finance from the University of SouthCarolina and teaches the investment planning, estate planning and portfoliomanagement courses. Austin Spencer received his Ph.D. from IndianaUniversity and teaches bank management courses. Roger Lirely, the depart-ment chair, holds a D.B.A. from Southern Illinois University-Carbondale. Heteaches income tax planning and has been the faculty advisor for the VitaIncome Tax program. John Richardson, MPM, CFP® is an adjunct instructor.Employed full time with AIG Valic, John brings a vast amount of expertise andexperience to the areas of risk management and retirement planning.



For more information about the WCU’s program in finance or otherprograms in business please contact Dr. Grace C. Allen, Associate Professorof Finance and Director of the Financial Planning Degree and CertificatePrograms at 828-227-3189 or [email protected].

Financial Planning CurriculumThe following courses are required:FIN 400 Introduction to Financial Planning and Risk ManagementFIN 406 Investment PlanningFIN 410 Retirement PlanningFIN 496 Estate Planning and Seminar in Financial PlanningACCT 355 Income Tax Planning

In addition to the above courses, 6 hours of electives approved for financialplanning are required.

Contact Information: For more information about the financial planningcurriculum at Western Carolina University contact Dr. Grace Allen at Collegeof Business, Western Carolina University, Asheville, NC 28723; Phone: (828)227-3189; Email: [email protected].

WCU Students in the Board Room of the New York Stock Exchange



JOURNAL OF PERSONAL FINANCEGUIDELINES FOR AUTHORS

The Journal of Personal Finance publishes empirical research, casestudies, practice management tutorials, and literature reviews dealing withfinancial consultation and planning topics, programs, and issues.

Articles are written by practicing financial planners and counselors,personal finance academicians, behavioral scientists, psychologists, andspecialists employed in business, industry, and academia.

Articles deal with all aspects personal financial planning that applyto practice management, business settings, and planning techniques. Topicsof interest to readers include:

• Client Relationship Management

• Financial Planning Trends

• Technology Issues

• Planning For Special Needs

• Regulation Overview

• Ethics Of Financial Planning

• Practice Management Techniques

• Novel Planning Tools And Techniques

• Investment Decision Management

• Marketing Methods

• Personnel Selection And Training

• Risk Assessment And Management

• Marketing And Consumer Behavior Research

• Employee Benefits And Assistance Planning

• Employee Counseling

• Personal Finance Research Methodology And Statistics

• Computer Applications In Personal Finance

• Book Reviews

It is important that manuscripts be well organized and concise. Thedevelopment of ideas and concepts should be clear. In general, readers preferarticles that avoid dull, stereotyped writing and the use of scientific jargon.Authors should aim at achieving clear communication of ideas. Genericmasculine pronoun or other sexist terminology should be avoided.

Tables should be used only where appropriate and should includeonly essential data. Tables should be understandable as a stand-alone


product. All tables should be created using Word Table formats. Columnsmust have headings, and data should be doubled spaced.

Authors are responsible for supplying figures (e.g., graphs, illustra-tions, line drawings) as camera-ready art done by commercial artists. Art mustbe original and of a quality that can be photographed clearly. All lines andprinting must be in proportion to the text.

References should follow APA 5th edition style. Authors arerequested to check all references for completeness, including year, volumenumber, and pages for journal citations. It is imperative that authors check tobe sure that all references mentioned in the text are listed in the referencesection. Manuscripts should be written in third-person past tense style. Forother questions of style, consult the most recent APA Publication Manual.This manual is available at most bookstores and libraries. Finally, pleasedouble-space all references.

The use of footnotes should be minimized. Each manuscript title andarticle headings should be as concise – not to exceed 10 words if possible. Ona separate page, place the title of the article, the names of the authors, theirprofessional titles, and their institutional affiliations. Double-space all title-page material.

Articles should include an abstract of not more than 150 words. Theabstract should express the central idea of the article in non-technicallanguage and should appear on a page separate from the text. A list of 3 to 5key words should be provided directly below the abstract.

Send an original and three clean copies of all material (four (4) copiestotal), on regular clean 8-1/2" x 11" white paper. Double-space all material(including references), all lines of tables (including heads), and extensivequotes. Allow wide margins (at least one inch on all four sides).

Please never submit material for concurrent consideration by anotherpublication. The International Association of Registered Financial Consult-ants will copyright articles published in the Journal of Personal Finance.After a manuscript has been accepted for publication and after all revisionshave been incorporated, manuscripts should be submitted to the Editor’sOffice as hard copy accompanied by electronic files on disk. Authors shouldlabel disks with identifying information, including software used, first author’slast name, and manuscript title. The disk must be PC formatted. The editorcannot accept a disk without its accompanying, matching hard-copy manu-script.

Submit manuscripts to: Ruth H. Lytton, Ph.D., 101 Wallace Hall,Virginia Tech, Blacksburg, VA 24061, (540) 231-6678, (540) 231-3251 (fax),[email protected].

Note that processing fees may be charged for manuscripts requiringsignificant editing or formatting changes. The editor does not acceptresponsibility for damage or loss of papers submitted.



Advertisement

Financial Counselingand Planning

The Journal of the

Association for Financial

Counseling

and Planning Education

��

��

��

��

��

��

��

��

�� ! � ��

��"��!

�� !��

��

��

�� #�� $

��%��

&�� '

��!��

��

��

!��

��!��

��

��

welcomes

manuscript

submissions

for 2006 issues

��

��

��

��

��

��

By publishing the results of currentresearch on issues of personal and familyfinancial management, the Journal serves theAFCPE membership and, through librarysubscriptions, the broader public as well. Inproviding such a publication outlet, thejournal also serves authors, both membersand non-members, who value the opportu-nity to disseminate their work. For thediverse membership of AFCPE, the journal isa critical avenue for sharing knowledgeamong practitioners and educators. TheJournal is published semi-annually.

��

� � � � � �

��

��

!�� "�� # $%&%'

(��) **+,+$&,-$*.

�� /��


Advertisement

Financial Services Review

Now in its 13th year, Financial Services Review (FSR) is the officialpublication of the Academy of Financial Services. The purpose of thisrefereed academic journal is to encourage research that examinesthe impact of financial issues on individuals. Although it is written inacademic style, FSR also fit the needs of professionals looking formore rigorous treatment of financial issues than that provided bypractitioner publications. Please visit our website atwww.pfp.gsu.edu for more information. Or contact the editor at theaddress below:

Conrad S. Ciccotello, Ph.D., JDAssociate ProfessorDirector of Personal Financial Planning ProgramsDepartment of Risk Management and InsuranceRobinson College of BusinessGeorgia State UniversityP.O. Box 4036Atlanta, GA 30302-4036(404) [email protected]

At least one of the authors must be a member of the Academy ofFinancial Services (AFS) to submit a paper. Submission fee formembers is $50. If none of the authors belongs to AFS, pleasecomplete a membership application form, which is available throughour website, www.pfp.gsu.edu. Annual membership in AFS is $60.

Authors are encouraged to submit their papers electronically as an e-mail attachment (Word format preferred) to the journal atcciccotello.gsu.edu.

Documents

Journal of Personal Finance Tools, Techniques, Strategies, and Research to Aid Consumers and