The Volatility Paradox: Tranquil Markets May Harbor … has declined across major asset classes and markets. Note: Realized volatility is the standard deviation of daily returns over

Second Quarter 2017 (data as of June 30)

A review of financial market themes and developments

This monitor reflects the best interpretation of financial market developments and views of the staff of the Office of Financial Research (OFR). It does not

`n ecessarily reflect a consensus of market participants or official positions or policy of the OFR or the U.S. Department of the Treasury. Contributors: Meraj Allahrakha, Viktoria Baklanova, Danny Barth, Ted Berg, Jill Cetina, Dagmar Chiella, Arthur Fliegelman, Dasol Kim, Francis Martinez, John McDonough, Philip Monin, Mark Paddrik, Eric Parolin, and Daniel Stemp.

The Volatility Paradox: Tranquil Markets May Harbor Hidden Risks Financial markets were exceptionally calm in the second quarter by most measures. Only three times in the past 90 years has volatility been so low: twice during bull markets in the 1960s and 1990s, and once in the lead-up to the financial crisis of 2007-09 (see Figure 1). Is today’s low volatility a sign of calm or a threat to financial stability — or both? This edition of the OFR’s Financial Markets Monitor investigates the volatility paradox: the possibility that low volatility leads investors to behave in ways that make the financial system more fragile and prone to crisis. We analyze three channels through which a prolonged period of low market volatility may introduce financial stability risks: increased leverage, reduced hedging, and institutional investors’ use of risk-management models. We find some supportive evidence of these channels at work, but better data are needed to make definitive conclusions. Volatility alone is not a good indicator of impending financial stress.

Figure 1: S&P 500 Index 90-day Realized Volatility (percent)Volatility in U.S. equity markets is the lowest in decades

Note: Volatility is the standard deviation of daily returns over 90 days expressed as annualized percent change.Source: Bloomberg Finance L.P.

Key findings • Volatility for most asset classes across the world fell below historical averages during the second quarter. In some

cases, volatility is near all-time lows. Drivers of low volatility may include expectations that the long U.S. economic expansion and still-easy funding conditions will persist.

• Some institutional investors have adapted by increasing leverage and the use of yield-enhancing strategies. • Shocks could produce procyclical responses if market participants use measures of realized volatility to manage

the risk of their portfolios.

0

25

50

75

1928 1938 1948 1958 1968 1978 1988 1998 2008 2018

4.01 on April 23, 1964

6.45 on May 2, 1995

6.47 on May 14, 2017

7.12 on Jan. 2, 2007

https://www.financialresearch.gov/financial-markets-monitor/

OFR MARKETS MONITOR Second Quarter 2017 | 2

Volatility alone is a weak risk indicator. Volatility measures for most asset classes across global financial markets fell below their historical averages during the second quarter (see Figure 2). Some measures approached all-time lows (for example, see Figures 1 and 3), which may have been driven by expectations that the long U.S. economic expansion and still-easy funding conditions will persist. There are two types of volatility: realized and implied. Realized volatility reflects the historical price fluctuations of an asset. Implied volatility is forward-looking. It captures the market’s expectation offuture price fluctuations of an asset, derived from the options markets. When implied volatility exceeds realized volatility, the difference reflects the extra return investors demand to hold a security solely because it is volatile. This difference is known as the volatility risk premium. One of the most widely cited measures of implied volatility is the Chicago Board Options Exchange Volatility Index (VIX). The VIX is the 30-day implied volatility of options on the benchmark S&P 500 equity index. A low VIX doesn’t necessarily signal that severe financial stress is unlikely. For instance, the VIX provided no advance warning of extreme volatility in the months leading up to the financial crisis. Realized volatility of the S&P 500 index was often substantially higher than the VIX had predicted 30 days earlier (represented by the blue dots over the 45-degree line in Figure 4). The relationship between realized and implied volatility for other asset classes followed a similar pattern during the crisis. Market risks may seem low when volatility is low. However, low volatility may also serve as a catalyst for market participants to take more risk, thereby making the financial system more fragile. Thisphenomenon is known as the volatility paradox.

Low volatility directly incentivizes risk-taking. Lower volatility may contribute to greater leveragingand risk-taking through at least three channels. Thefirst channel is through changing asset-returncorrelations, which tend to increase when markets are volatile. Low correlations could entice investors to

-3

0

3

6

9

2007 2010 2013 2016

Figure 2: Realized Volatility by Asset Class (z-score)Volatility has declined across major asset classes and markets

Note: Realized volatility is the standard deviation of daily returns over 30 days, expressed as annualized percent change. U.S. equities are represented by the S&P 500 index. U.S. interest rates are the weighted average of the Treasury yield curve. Global currencies are based on weights from JPMVXY index. Global equities are MSCI All Countries World Excluding U.S. Index. Standardization uses data since Jan. 1, 1993.Sources: Bloomberg Finance L.P., OFR analysis

U.S. interest ratesU.S. equities

Global currenciesGlobal equities (ex-U.S.)

0

20

40

60

80

100

1990 1995 2000 2005 2010 2015

Figure 3: Chicago Board Options Exchange Volatility Index (VIX) (percent) Implied volatility on equities has fallen to near all-time lows

Note: Implied volatility is derived from options markets and is the expected standard deviation of daily returns over the next 30 days, expressed as annualized percent change.Source: Bloomberg Finance L.P.

9.31 on Dec. 22, 1993 9.75 on June 2, 2017

0

25

50

75

100

0 25 50 75 100

Figure 4: VIX and Realized Volatility of S&P 500 Index (percent)The VIX did not predict the global financial crisis

Note: Realized volatility is the standard deviation of daily returns over 30 days, expressed as annualized percent change. Sources: Bloomberg Finance L.P., OFR analysis

Global financial crisis(8/1/2008 to 11/1/2008)

VIX

Other dates(3/1/1993 to 5/31/2017)

Realizedvolatility


accumulate risky exposures, believing they arediversified. Prolonged periods of low volatility may further decrease correlations, encouraging further risk-taking. This procyclical behavior increasesinvestors’ risk of loss from a systematic shock, when volatility spikes and asset-return correlations revert to historical levels. Some evidence exists that this channel may be at work in equity markets. Sector correlations have declined significantly during the past two years, while volatility has remained low (see Figure 5). Second, low volatility could encourage the use of other yield-enhancing strategies, such as selling deep out-of-the-money put options (those with a strike price substantially below current prices). Investors collect a premium from selling these options, but can be obligated to purchase the underlying assets if the price drops below the strike price. Investors who accumulate these risky exposures could be more likely to experience financial stress if prices sharply decline. Available data on investor portfolios are not sufficient to assess this channel adequately. Third, low volatility can directly incentivizeleveraging by lulling investors into underestimating the odds of a volatility spike. One measure of marketwide leverage is the ratio of margin debt to market capitalization. This measure is imperfect because it doesn’t account for other positions on investor balance sheets, including derivatives positions. Figure 6 uses margin debt balances and market capitalization data from the New York Stock Exchange. The ratio increased from 2002 to 2007 amid low volatility, declined after the crisis, and has been climbing since as volatility again reached long-term lows. Evidence also exists that some large investors are highly leveraged and, for that reason, may be susceptible to volatility events. For example, the top decile of macro and relative-value hedge funds has been leveraged about 15 times in recent quarters. These funds combined account for more than $800 billion in gross assets, about one-sixth of all hedge fund assets.

Low volatility could also disincentivize investor hedging.

0

25

50

75

Jan2005

Jan2007

Jan2009

Jan2011

Jan2013

Jan2015

Jan2017

0.00

0.25

0.50

0.75

1.00

Figure 5: 3-Month Moving Average of S&P 500 Sector Correlations, VIX Index (correlation, percent) Correlations between sectors have fallen amid low volatility

Note: S&P 500 sector pairwise monthly correlation; 3-month moving average.Sources: Bloomberg L.P., OFR analysis

Sector correlation (left)VIX index (right)

0

25

50

75

100

Jan2007

Jan2009

Jan2011

Jan2013

Jan2015

Jan2017

1

2

3

Jan2005

Figure 6: Margin Debt Balance over Market Capitalization and S&P 500 Index 30-day Realized Volatility (percent) Realized volatility has fallen as investors increased margin debt

Note: Values are the New York Stock Exchange (NYSE) market capitalization and margin debt balances of its members. Dealer margin debt balances may reflect positions on securities not listed on the NYSE. Realized volatility is the standard deviation of daily returns over 30 days expressed as annualized percent change.Sources: Haver Analytics, OFR analysis

Margin debt / market capitalization (left)S&P 500 index realized volatility (right)


Another way investors may adapt to low volatility is by reducing their hedging of risky positions. This behavior was particularly relevant in recent years, when historically low interest rates pressured investors to reach for yield by holding more lower-rated fixed-income securities and more equities (see the OFR’s 2016 Financial Stability Report). OFR analysis of options trading suggests that investors have reduced their hedging of market exposure. Investor hedging activity is difficult to measure, although it can be captured to some extent using contracts outstanding in current-month SPY options. SPY is an exchange-traded fund that mirrors the benchmark S&P 500 equity index. Traders commonly sell SPY options to hedge equity market exposure. Options give investors the right, but not the obligation, to buy or sell a specific security at a specific strike price and time. A call option is a right to buy; a put is a right to sell. Options with a strike price near the current price of SPY are said to be “at the money.” Contracts with a strike price far from the current price are “away from the money.” These options are less likely to be held for hedging purposes and instead may represent yield-enhancing strategies. Investor hedging activity is captured through a hedging rate, calculated as the proportion of contracts on SPY options that is “at the money” versus “away from the money.” Hedging rates are currently lower on average than in the years immediately preceding the financial crisis (see Figure 7), suggesting a structural change in hedging activities after the crisis. However, the evidence is somewhat mixed. Considerable variation has occurred since 2010, and current levels appear to be higher relative to 2014 for both call and put hedging ratios. The absence of sharper measures of aggregate hedging activities makes drawing definitive conclusions difficult, though these hedging ratios at least suggest significant differences before and after the crisis. The Commodity Futures Trading Commission (CFTC) collects data on an alternative measure of hedging activity using positions of futures traders. CFTC data categorize hedge funds and other investors as “non-commercial,” or speculative, traders. As of May 2017, the net short position on VIX futures of non-commercial traders sat at levels larger than even before the crisis (see Figure 8). Common volatility strategies involve taking short positions in longer-dated contracts and long positions in shorter-dated contracts. Reduced

60

70

80

90

100

2005 2007 2009 2011 2013 2015 2017

Figure 7: SPY Options Held for Hedging Purposes (percent)Investors are less hedged compared to the pre-crisis period

Call hedging ratePut hedging rate

Sources: OptionsMetrics, OFR analysis

-150,000

-100,000

-50,000

0

50,000

2004 2006 2008 2010 2012 2014 2016

Figure 8: VIX Futures Noncommercial Net Total (contracts) Speculators increased short bets on VIX to the most since 2004

Sources: Bloomberg Finance L.P., Commodity Futures Trading Commission

Net position

-143,845 on June 20, 2017


hedging in these strategies would imply shorting in the aggregate, consistent with Figure 8. However, establishing a direct link without more granular data is difficult. Together, these data suggest that some investors may have adapted to the low-volatility environment by reducing risk hedges and increasing speculative bets. Data limitations temper the findings to some extent, and leave opportunities for further analysis. With less hedging, these investors’ balance sheets may be less resilient to large volatility shocks when volatility returns to financial markets. Value-at-Risk models may give faulty signals in low-volatility markets. Low realized volatility can affect the behavior of banks, hedge funds, and other asset managers that use a risk management framework based on realized volatility, including some Value-at-Risk (VaR) measures. About 40 percent of large hedge funds, representing about 62 percent of gross hedge fund assets, regularly calculate VaR statistics for their funds, according to Form PF data collected by the Securities and Exchange Commission (SEC). VaR measures the risk of investments. It captures how much value investments might lose over a set time. Although VaR can be a valuable risk-management tool, overreliance on VaR when volatility is low could result in procyclical behavior that makes investors more vulnerable to volatility shocks if market conditions change abruptly. A decline in realized volatility can reduce a portfolio’s VaR, allowing market participants to increase position sizes without exceeding predefined VaR risk limits. The reverse is true when volatility rises. In that case, VaR-sensitive investors may be forced to simultaneously sell assets to get their portfolios below risk limits. A selloff induced by a VaR shock can become self-reinforcing as liquidity dries up and as deleveraging occurs. Some market observers believe VaR shocks contributed to selloffs in the Japanese government bond market in 2003 and in the U.S. Treasury market during the 2013 taper tantrum (see Figure 9). Long-term investors that are not sensitive to VaR, such as pension funds and insurance companies, may not step in and provide liquidity unless prices fall sharply.

0

25

50

75

100

125Day 0 Day

10Day20

Day30

Day40

Day50

Day60

Day70

Day80

Day90

Figure 9: Cumulative Yield Change in 10-year Government Bonds (basis points)VaR shocks may have deepened past selloffs in bond markets

Note: The vertical axis is inverted to reflect lower bond prices as yields increase. Horizontal axis is the number of days since the beginning of the sell-off period.Sources: Bloomberg Finance L.P., OFR analysis

Japanese Government Bond market (6/12/2003 to 9/10/2003)

U.S. Treasuries taper tantrum (5/2/2013 to 7/31/2013)


Most large U.S. banks report data on the VaR of their trading books in quarterly 10-Q filings to the SEC. These data show a dramatic decline since 2010 in the VaR of banks’ trading books, without acommensurate decrease in the fair value of those trading books (see Figure 10). All else being equal, this change suggests that the reduction in VaR may reflect falling realized volatility rather than a decline in the size of banks’ trading books during the period. If volatility rises and banks aim to keep their VaR stable, the banks would need to shrink their trading books. Another possibility is that the declining VaR is evidence that banks have reduced the overallmarket risk in their portfolios, in part responding to additional regulatory oversight. A definitiveconclusion is difficult without detailed data on dealepositions. Targeting a specific level of volatility has recentlybecome an investment strategy. Many institutionainvestors now are holding so-called “volatility control funds” in their portfolios. Assets under management in variable annuity volatility control funds rose to $325 billion at the end of 2016 (see Figure 11). These funds make asset allocation decisions aimed atmaintaining a stable level of volatility for their whole portfolios. If volatility were to rise suddenly in a previously stable asset class, these funds may beforced to rebalance and sell assets. These investors’ activities could have a procyclical effect on asset prices and exaggerate volatility.

Note: G-SIB = Global systemically important bank

r Sources: Bank 10-Q forms filed with Securities and Exchange Commission, OFR analysis Figure 11: Variable Annuity Volatility Control Funds ($ billions,

count)

l Variable annuity volatility control funds have more than doubled

Conclusion Prolonged low market volatility may introduce financial stability risks through at least three channels. First, investors could respond by directly taking on more leverage and risk. Second, investors could reduce hedging activities. Third, institutional investors’ use of VaR or other risk-management models that have realized volatility as a key input could lead them to take more risk. A spike in volatility can result in outsized investor losses from sharp asset price changes. Data limitations hinder the ability to make definitive conclusions regarding the extent to which these channels are at work. However, the evidence is consistent with these channels operating and suggests the need for further analysis.

0.00

0.25

0.50

0.75

1.00

Mar2010

Mar2011

Mar2012

Mar2013

Mar2014

Mar2015

Mar2016

Mar2017

0

500

1,000

1,500

2,000

Figure 10: U.S. G-SIBs' Combined Trading Books ($ billions) Big banks' VaR has collapsed but portfolio size is little changed

Fair value of trading book (left)Trading book VaR (right)

0

100

200

300

400

Dec2010

Dec2011

Dec2012

Dec2013

Dec2014

Dec2015

Dec2016

0

100

200

300

in size

Source: Milliman Financial Risk Management LLC

Number of funds (left)Assets under management (right)


Selected Global Asset Price Developments

1

Select U.S. Interest Rates

U.S. T reasury yields and yield curve

S ource: Bloomberg Finance L.P .

percent basis po ints2-year (Left Axis)10-year (Left Axis)10-year - 2-year spread (Right Axis)

Jul2016

O ct2016

Ja n2017

Ap r2017

Jul2017

0

1

2

3

75

100

125

150

175

200

225

250

U.S. T reasury t erm premium (basis point s)

Note : Adrian, Crump, & Moe nch mode lS ource: Bloomberg Finance L.P .

10-year 2-year

2010 2011 2012 2013 2014 2015 2016 2017-100

-50

0

50

100

150

200

250

300

Professional vs. market - implied U.S. inflat ion expect at ions(percent )


Survey o f Pro fessional Forecasters 10-year annual average5y5y forward breakeven rate CPI current

Sep2015

No v2015

Ja n2016

Ma r2016

Ma y2016

Jul2016

Sep2016

No v2016

Ja n2017

Ma r2017

Ma y2017

Jul2017

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

Short -t erm market rat es (percent )


1-month Treasury bill 3-month LIBORGCF Treasury Repo

Jul2016

Sep2016

No v2016

Ja n2017

Ma r2017

Ma y2017

Jul2017

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

T hree-mont h Eurodollar fut ures (percent )

Note s: T he high and low points of the De c FOMC proje ctions are the maximumand minimum fore casts. T he re ctangle re pre se nts the me dian.S ource: Bloomberg Finance L.P .

Jun-30 May-17 Apr-17FOMC pro jections from June 2017

Jul2017

Ja n2018

Jul2018

Ja n2019

Jul2019

Ja n2020

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

Money market and policy int erest rat es (percent )


GCF Treasury repoFed funds effectiveInterest on excess reservesReverse repo facility

2011 2012 2013 2014 2015 2016 20170.0

0.25

0.5

0.75

1.0

1.25

1.5

2

U.S. Corporate Debt Markets

U.S. corporat e bond opt ion-adjust ed spreads(basis point s)

S ource: Haver Analytics

Investment grade (Left Axis) High yield (Right Axis)

Jul2016

O ct2016

Ja n2017

Ap r2017

Jul2017

100

120

140

160

180

200

220

240

200

400

600

800

1000

U.S. corporat e CDS indexes (basis point s)

Note : Five -ye ar maturity CDS Inde xS ource: Bloomberg Finance L.P .

Investment grade (Left Axis) High yield (Right Axis)

Jul2016

O ct2016

Ja n2017

Ap r2017

Jul2017

0

30

60

90

120

150

200

300

400

500

600

U.S. non-financial credit gross issuance ($ billions)

S ources: Dealogic, S tandard & Poor's Leveraged Commentary & Data

Investment Grade High Yield Leveraged Loans

Aug2015

O ct2015

D ec2015

Feb2016

Ap r2016

Jun2016

Aug2016

O ct2016

D ec2016

Feb2017

Ap r2017

Jun2017

0

50

100

150

U.S. corporat e credit fund flows ($ billions)

Note : Flows data are re le ase d with one -month lag.S ource: Haver Analytics

High yield Leveraged loans

Jul2016

O ct2016

Ja n2017

Ap r2017

Jul2017

-15

-10

-5

0

5

10

Leveraged loan issuance by use of proceeds (percent )

Note : Data for 2017 are ye ar-to-date as of January.S ources: S tandard & Poor's Leveraged Commentary & Data, OFR analysis

M&A/LBO Dividend/Buyback RefinancingOther

2000 2002 2004 2006 2008 2010 2012 2014 20160

10

20

30

40

50

60

70

80

90

100

Leveraged loan price act ivit y

Note s: S&P Le ve rage d Loan Inde x. Inde x 100=January 01, 2012.S ource: Bloomberg Finance L.P .

Ja n2012

Jul2012

Ja n2013

Jul2013

Ja n2014

Jul2014

Ja n2015

Jul2015

Ja n2016

Jul2016

Ja n2017

Jul2017

95

100

105

110

3

Primary and Secondary Mortgage Markets

Primary mort gage rat es (percent )


5-year/1-year adjustable rate 30-year fixed

Jul2016

O ct2016

Ja n2017

Ap r2017

Jul2017

2

3

4

5

MBS yield and opt ion-adjust ed spread t o U.S. T reasurysecurit ies


percent basis po ints

Current coupon (Left Axis) Spread (Right Axis)

Jul2016

O ct2016

Ja n2017

Ap r2017

Jul2017

1

2

3

4

40

60

80

100

120

30-year home mort gage fixed and jumbo rat es and spread


percent basis po ints30-year fixed (Left Axis)30-year jumbo (Left Axis)30-year jumbo-conforming spread (Right Axis)

Jul2016

O ct2016

Ja n2017

Ap r2017

Jul2017

3.0

3.5

4.0

4.5

5.0

5.5

0

20

40

60

80

100

120

Refinance and purchase loan applicat ions

Note : Inde x 100 = July 01, 2016.S ource: Bloomberg Finance L.P .

percent

Purchase Index (Left Axis) Gov Refi Index (Left Axis)Conv Refi Index (Left Axis)Refi % of to tal apps (Right Axis)

Jul2016

O ct2016

Ja n2017

Ap r2017

Jul2017

40

90

140

190

240

40

45

50

55

60

65

70

Convent ional mort gage severe delinquencies(percent , 90+ days lat e, seasonally adjust ed)

S ource Haver Analytics

Prime Subprime

2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 20170

4

8

12

16

4

Equity Markets

Global equit y indices

Note : Inde x = July 01, 2016.S ource: Bloomberg Finance L.P .

S&P 500 MSCI EM Nikkei 225Shanghai Euro Stoxx 50

Jul2016

O ct2016

Ja n2017

Ap r2017

Jul2017

85

95

105

115

125

135

U.S. equit y indexes

Note : Inde x 100 = Jan 01, 2000.S ource: Bloomberg Finance L.P .

S&P 500 NASDAQ Russell 3000

2000 2003 2006 2009 2012 201520

60

100

140

180

S&P 500 sect or performance

Note : Inde x 100 = July 01, 2016.S ource: Bloomberg Finance L.P .

S&P 500 Financials Consumer staplesEnergy

Jul2016

O ct2016

Ja n2017

Ap r2017

Jul2017

80

90

100

110

120

130

140

S&P 500 price-t o-earnings and price-t o-book rat ios(mult iple)


Price-to -earnings (Left Axis) Price-to -book (Right Axis)

2012 2014 20169

13

17

21

1

2

3

4

U.S. equit y valuat ions: Shiller CAPE (rat io)

Note s: CAPE is the ratio of the monthly S&P 500 price le ve l to trailing te n-ye arave rage e arnings (inflation adjuste d).S ources: Haver Analytics, Robert S hiller

1920 1930 1940 1950 1960 1970 1980 1990 2000 20100

10

20

30

40

50

0

10

20

30

40

50

S&P 500 implied volat ilit y and opt ion skew(percent )

Note s: Option ske w is the diffe re nce be twe e n thre e -month implie d volatility ofout of the mone y puts and calls with strike s e qual distance from the spot price(+/- 20 pe rce nt). Highe r value s re fle ct gre ate r de mand for downside riskprote ction.S ource: Bloomberg Finance L.P .

VIX 80% - 120% Skew

Jul2016

O ct2016

Ja n2017

Ap r2017

Jul2017

0

10

20

30

5

Volatility

Implied volat ilit y by asset class (Z-score)

Note s: Z-score re pre se nts the distance from the ave rage , e xpre sse d instandard de viations. Standardization use s data going back to January 01, 1993.S ources: Bloomberg Finance L.P ., OFR analysis

U.S. equities (VIX) Global currencies (JPMVXYGL)U.S. Treasuries (MOVE) Average

2013 2014 2015 2016 2017-3

-2

-1

0

1

2

3

Realized volat ilit y by asset class (Z-score)

Note s: T hirty-day re alize d volatility. Equitie s base d on S&P 500 inde x, inte re strate s base d on we ighte d ave rage of T re asury yie ld curve , FX base d on we ightsfrom JPMVXY inde x. Standardization use s data going back to January 01, 1993.S ources: Bloomberg Finance L.P ., OFR analysis

Global FX U.S. interest rates U.S. equitiesAverage

2013 2014 2015 2016 2017-3

-2

-1

0

1

2

3

Global equit y indexes 1-mont h implied volat ilit y (percent )


Eurostoxx 50 SP500 German DaxUK FTSE 300 Japan Nikkei MSCI EM

Jul2015

Sep2015

No v2015

Ja n2016

Ma r2016

Ma y2016

Jul2016

Sep2016

No v2016

Ja n2017

Ma r2017

Ma y2017

Jul2017

0

10

20

30

40

50

60

Slopes of implied volat ilit y curves(basis point s)

Note s: Se ve n-day moving ave rage . Slope re pre se nts diffe re nce be twe e n one -ye ar and one -month maturitie s. G10 FX base d on we ights from De utsche Bank'sCVIX inde x.S ources: Bloomberg Finance L.P ., OFR analysis

G10 FX (Left Axis) S&P 500 (Left Axis)2-year USD Swaption Rate (Right Axis)10-year USD Swaption Rate (Right Axis)

Jul2016

Sep2016

No v2016

Ja n2017

Ma r2017

Ma y2017

Jul2017

-600

-400

-200

0

200

400

600

800

-30

-20

-10

0

10

20

30

40

Opt ion skew by asset class (z-score)

Note s: Option ske w is the diffe re nce be twe e n thre e -month implie d volatility ofout of the mone y puts and calls with strike s e qual distance from the spot price(+/- 10 pe rce nt). Highe r value s re fle ct gre ate r de mand for downside riskprote ction. Equitie s re pre se nts S&P500 inde x. Inte re st rate s re pre se ntwe ighte d ave rage ske w of T re asury future s curve . Curre ncie s re pre se nt dollarske w against major curre ncie s base d on JPMVXY inde x we ights. Z-scorestandardization use s data going back to January 01, 2006.S ources: Bloomberg Finance L.P ., OFR analysis

U.S. equities U.S. interest ratesGlobal currencies Average

Jul2016

O ct2016

Ja n2017

Ap r2017

Jul2017

-3

-2

-1

0

1

2

3

Volat ilit y of equit y volat ilit y

Note : VVIX Inde x me asure s the e xpe cte d volatility of the 30-day forward priceof the CBOE VIX Inde x.S ource: Bloomberg Finance L.P .

Jul2016

O ct2016

Ja n2017

Ap r2017

Jul2017

60

80

100

120

140

6

Advanced Economies

2-year sovereign bond yields (percent )


U.S. Germany U.K. Japan

Jul2016

O ct2016

Ja n2017

Ap r2017

Jul2017

-1.5

-1

-0.5

0

0.5

1

1.5

2

10-year sovereign bond yields (percent )


U.S. Germany U.K. Japan France

Jul2016

O ct2016

Ja n2017

Ap r2017

Jul2017

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

Breakeven inflat ion (percent )


U.S. ten-year Germany ten-year U.K. ten-yearJapan ten-year

Jul2016

O ct2016

Ja n2017

Ap r2017

Jul2017

0

1

2

3

4

10-year euro area periphery government bond spreadsover German bunds (basis point s)


Italian govt (Left Axis) Spanish govt (Left Axis)Portuguese govt (Left Axis) Greek govt (Right Axis)

Jul2016

O ct2016

Ja n2017

Ap r2017

Jul2017

80

120

160

200

240

280

320

360

400

0

400

800

1200

1600

2000

Major currency indexes

Note s: Fore ign curre ncy incre ase s re pre se nt gre ate r stre ngth ve rsus the U.S.dollar. DXY incre ase s re pre se nt gre ate r stre ngth of the U.S. dollar ve rsus abaske t of major world curre ncie s. Inde x 100 = July 01, 2016.S ource: Bloomberg Finance L.P .

DXY (U.S. do llar) euro British poundJapanese yen Swiss franc

Jul2016

O ct2016

Ja n2017

Ap r2017

Jul2017

80

90

100

110

U.S. dollar long posit ioning vs. major currencies(net speculat ive posit ions, t housands of cont ract s)

Note s: Positive value s re pre se nt ne t U.S. dollar long positions. T he DollarInde x (DXY) is a future s contract base d on the U.S. dollar's value against abaske t of major world curre ncie s. T o e xpre ss a U.S. dollar long position in non-U.S. dollar contract, the contract must be shorte d.S ource: Bloomberg Finance L.P .

a

DXY (U.S. do llar) euro British poundJapanese yen Total

Jul2016

O ct2016

Ja n2017

Ap r2017

Jul2017

-100

0

100

200

300

400

7

Emerging Markets

Emerging market currencies(U.S. dollars per foreign currency unit )

Note s: Incre asing value s indicate stre ngthe ning ve rsus the U.S. dollar. Inde x100=July 01, 2016.S ource: Bloomberg Finance L.P .

RussiaChina o ffshoreBrazilEM currency Index

Jul2016

O ct2016

Ja n2017

Ap r2017

Jul2017

90

100

110

120

Emerging market sovereign debt yield


EM local currency

Jul2016

O ct2016

Ja n2017

Ap r2017

Jul2017

3.5

4.0

4.5

5.0

5.5

Equit y price indexes

Note s: T he US e quity inde x is the S&P 500 Inde x. T he Chine se e quity inde x isthe Shanghai Composite Inde x. T he De ve lope d Economie s inde x is the MSCIWorld Inde x and the Eme rging Marke ts inde x is the MSCI EM Inde x (both are inlocal te rms). Inde x 100 = July 01, 2016.S ource: Bloomberg L.P .

U.S. China Developed economiesEmerging markets

Jul2016

O ct2016

Ja n2017

Ap r2017

Jul2017

80

90

100

110

120

130

1-mont h realized emerging market s volat ilit y (percent )

Note s: Re alize d volatility is the annualize d standard de viation. Hard curre ncysove re ign de bt base d on the J.P. Morgan Eme rging Bonds - Global Price Inde xand curre ncie s base d on a we ighte d ave rage of EM curre ncy re turns against thedollar using we ights from J.P. Morgan VXY-EM curre ncy volatility inde x.S ources: Bloomberg L.P ., OFR analysis

EM sovereign hard currency debt EM currencies

2010 2011 2012 2013 2014 2015 2016 20170

10

20

IIF port folio f lows t o emerging market s ($ billion)

Note s: Data re pre se nt the Institute of Inte rnational Finance 's monthly e stimate sof non-re side nt flows into thirty EM countrie s. Data for late st obse rvations arede rive d from IIF's e mpirical e stimate s using data from a smalle r subse t ofcountrie s, ne t issuance , and othe r financial marke t indicators.S ource: Bloomberg

Debt flows Equity flows

2010 2011 2012 2013 2014 2015 2016 2017-30

-20

-10

0

10

20

30

40

50

China's Foreign Exchange Reserves ($ t rillion)

S ource: Bloomberg

FX reserves

Ja n2006

Ja n2007

Ja n2008

Ja n2009

Ja n2010

Ja n2011

Ja n2012

Ja n2013

Ja n2014

Ja n2015

Ja n2016

Ja n2017

$0.5

$1

$1.5

$2

$2.5

$3

$3.5

$4

$4.5

8

Commodit ies

Major commodit ies prices

Note s: Inde x 100 = January 01, 2010S ource: Bloomberg Finance L.P .

Bloomberg commodities indexCrude o il front month (Brent) Gold front month

2014 2015 2016 20170

25

50

75

100

125

Crude oil

Note : WT I and Bre nt are front-month contracts.S ource: Bloomberg Finance L.P .

$/Barrel Million BarrelsWTI (Left Axis)Brent (Left Axis)U.S. inventories (Right Axis)

2014 2015 2016 201720

50

80

110

140

300

350

400

450

500

550

Oil and nat ural gas fut ures curves

Note : Data as of July 05, 2017.S ources: Bloomberg Finance L.P ., OFR analysis

$/barrel $/mmbtu

Brent (Left Axis) Natural gas (Right Axis)

2M 6M 12M 24M 60M45

50

55

60

2.0

2.5

3.0

3.5

4.0

Oil supply and demand fact ors

Note : Global production and consumption are e stimate s by the Inte rnationalEne rgy Age ncy.S ource: Bloomberg Finance L.P .

Million barrels per dayGlobal production (Left Axis)Global consumption (Left Axis)U.S. rig count (Right Axis)

2014 2015 2016 201785

90

95

100

105

110

115

120

200

400

600

800

1000

1200

1400

1600

1800

Speculat ive fut ures posit ioning(t housands of cont ract s)

Note s: Positive value s re pre se nt ne t long positions. Ne gative value s re pre se ntne t short positions.S ource: Bloomberg Finance L.P .

Brent (Left Axis) Gold (Left Axis)Copper (Left Axis) Steel (Right Axis)

Jul2015

Ja n2016

Jul2016

Ja n2017

Jul2017

-140

-70

0

70

140

210

280

350

-20

-10

0

10

20

30

40

50

Met als spot price indexes

Note : Inde x 100 = January 01, 2010.S ource: Bloomberg Finance L.P .

Copper Steel Precious metals

Jul2015

Ja n2016

Jul2016

Ja n2017

Jul2017

0

50

100

150

200

Documents

The Volatility Paradox: Tranquil Markets May Harbor … has declined across major asset classes and markets. Note: Realized volatility is the standard deviation of daily returns over