InterVISTAS Airport Traffic Forecasting Workshop ... · Airport Forecasting Record. • Sometimes...

Preview:

Citation preview

Canadian Forecasting WorkshopSession 1: 

Introductory Remarks on the Science and Art Forecasting

Dr. Mike TrethewayInterVISTAS Consulting Chief Economist

Realizing the vision together1

Today’s Workshop

1. Introductory Remarks on Forecasting2. Transport Canada PODM/PTAM models3. Alternative Approaches-

Single Airport Forecasts4. Incorporating Uncertainty into

Air Traffic Forecasts5. Current Outlook

Realizing the vision together2

Today’s Workshop

• Dr. Mike Tretheway• Chief Economist, InterVISTAS Consulting Group

• Technical Director, Business Line Aviation

• Ian Kincaid• Vice President, Economic Analysis

• Head of Forecasting Practice

Realizing the vision together

Background Report ACRP 76

Addressing Uncertainty about Future Airport Activity Levels in Airport decision making

Undertaken by

• InterVISTAS Consulting Inc.

• Mike TrethewayIan Kincaid

• HDR Inc.

• David LewisStéphane Gros

3

Realizing the vision together

Airport Forecasting Record.

• Forecasting is an essential tool for airports

• Medium to long term master planning

• Financial forecasts

• Operational Forecasts

4

Realizing the vision together

Airport Forecasting Record.

• But the track record has not always been good

5

Realizing the vision together

Atlanta

6

35

37

39

41

43

45

47

49

51

53

55

2000 2002 2004 2006 2008 2010 2012 2014

Pas

seng

er E

npla

nem

ents

(Mill

ions

)

Actual TrafficTAF 2001TAF 2003TAF 2005TAF 2007TAF 2009

Actual and Forecast Total Passenger Enplanements at Hartsfield-Jackson Atlanta International Airport

Realizing the vision together

Washington Dulles

7

0

5

10

15

20

25

1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014

Pass

enge

r Enp

lane

men

ts (M

illio

ns)

Actual Traffic

TAF 2000

TAF 2001

TAF 2003

TAF 2004

TAF 2005

TAF 2009

Actual and Forecast Total Passenger Enplanements at Washington Dulles International

Realizing the vision together

Airport Forecasting Record.

• But the track record has not always been good

• Albeit with some learning

8

Realizing the vision together

St. Louis

9

0

5

10

15

20

25

30

1985 1990 1995 2000 2005 2010 2015

Pass

enge

r Enp

lane

men

ts (M

illio

ns)

Actual TrafficTAF 1998TAF 2001TAF 2002TAF 2003TAF 2009

TWA declaresbankruptcy

TWA declaresbankruptcy for

the second time

TWA declaresbankruptcy for

the third time andAA buys TWA.Construction of

new runway beginsAA reduces

services at STL

AA terminates itsfocus city at STL

Actual and Forecast Total Passenger Enplanements at Lambert-St. Louis International Airport

Realizing the vision together

St. Louis

10

0

5

10

15

20

25

30

1985 1990 1995 2000 2005 2010 2015

Pass

enge

r Enp

lane

men

ts (M

illio

ns)

Actual TrafficTAF 1998TAF 2001TAF 2002TAF 2003TAF 2009

TWA declaresbankruptcy

TWA declaresbankruptcy for

the second time

TWA declaresbankruptcy for

the third time andAA buys TWA.Construction of

new runway beginsAA reduces

services at STL

AA terminates itsfocus city at STL

Actual and Forecast Total Passenger Enplanements at Lambert-St. Louis International Airport

Realizing the vision together

Airport Forecasting Record.

• Sometimes the long run forecast has been good

• but with short term variance

• And different traffic mix than original forecast

11

Realizing the vision together

BWI

12

0

2

4

6

8

10

12

1972

1974

1976

1978

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

2004

2006

2008

2010

Pass

enge

r Enp

lane

men

ts (M

illio

ns)

Actual Traffic

1987 Master Plan Forecast (Baseline)

Piedmont announces hub

First Gulf Warand recession

Southwest Airlines launches services

9/11 andrecession

Recession

U.S. Airways "de-hubs"

Actual and Forecast Total Passenger Enplanements atBaltimore/Washington International Thurgood Marshall Airport

Realizing the vision together

Airport Forecasting Record.

• Sometimes unanticipated events dramatically change a market

13

Realizing the vision together

New Orleans

14

0

1

2

3

4

5

6

7

8

1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014

Pass

enge

r Enp

lane

men

ts (M

illio

ns)

Actual TrafficTAF 2001TAF 2004TAF 2005TAF 2009

Hurricane Katrina

Actual and Forecast Total Passenger Enplanements atLouis Armstrong New Orleans International Airport

Realizing the vision together

Airport Forecasting Record.

• The forecasting record can also be one of underforecasting

15

Realizing the vision together

Bellingham WA

16

0

50

100

150

200

250

300

350

400

450

1985 1990 1995 2000 2005 2010 2015 2020

Pass

enge

r Enp

lane

men

ts (T

hous

ands

)

Actual Traffic

TAF 2000

TAF 2003

Master Plan Forecast

United Express / SkyWest exits in 2001

Allegiant enters in August 2004 and rapidly develops service

Allegiant opens base at BLI in January 2008

Actual and Forecast Total Passenger Enplanements at Bellingham International Airport

Realizing the vision together

Guessing vs. Analysis

Midway (1976)

17

Realizing the vision together

• Setting:Station Hypo• Code breaking centre, Pearl Harbor Naval Base

• Early April 1942 (US fleet crippled after 7Dec1941)

• Issue: forecasting Japanese naval fleet intentions for coming 2 months

• MG is fictional character (Matt Garth)

• JR is historical character (Joe Roquefort, head of Honolulu code breaking unit)

18

Realizing the vision together

o MG: Now the boss is afraid Yamamoto's going to jump back at us. But where? JR: We got the latest intercepts here. Here's a list of Japanese ships we suspect will be assigned to amphib operations south of Rabaul. The Coral Sea! That's where we think they'll strike next. But something else is stirring, something out our way.

o MG: We need facts, not guesswork. o JR: Matt, we cracked Yamamoto's code, but we can't just reel it off. We get a flicker here

and a glimmer there. o MG: How much can you decipher? o JR: Hell, maybe... o MG: Really decipher. o JR: Ten percent. o MG: That's one word in . For Christ's sake, you're guessing! o JR: We like to call it analysis

19

Realizing the vision together

Types of Airport Forecasts.

• Passenger traffic• Total pax

• Enplaned/deplaned

• O/D vs Connecting

• Cargo tons• E/D vs on-board

20

Realizing the vision together

Types of Airport Forecasts.

• Aircraft Movements

• Commercial

• Heavy jet, turboprops, piston

• Typically driven by

• base pax/cargo-freighter forecast

• Pax/aircraft (cargo/aircraft)

• General Aviation

• See recent trend

• Other

• Military (Busan, new Beijing examples),

• rescue, government, tech stop21

Realizing the vision together22

Realizing the vision together

Total GA Movements for Canada

23

Source: Statistics Canada, Table 401-007, 401-0015 and 401-0021.

Realizing the vision together

Types of Airport Forecasts.

• Operational Forecasts• Day by day, Hour by hour forecasts

of a specific traffic type

• E.g., volumes through a security pointborder process

• Ex- Blackcomb Ski Corp.• Recognize effects of

• Annual volume drivers• weather, • Interaction between the sessions

• Increased local skiing today means less in 2 weeks

24

Realizing the vision together

Forecast Probability and Risk.

• Low-Central-High• What are the probabilities of each scenario?

• Threshold

• What is probability that traffic in each year will fall below 19mn pax?

• Risk

• What is the 20/80% range of the forecast 10 years from now?

25

Realizing the vision together26

Thank You

Subscribe to Monthly Aviation Intelligence Reportwww.InterVISTAS.com

Transport Canada Forecasts

InterVISTAS Consulting Group

10 April 2013

Realizing the vision together

Outline

28

•Overview of the Transport Canada Forecasts:• Methodology

• Data

• Pros and cons

•Alternative Approaches:• Single airport methodologies

• Methodology

• Addressing uncertainty

Transport Canada Forecasts

Realizing the vision together

Transport Canada Forecasts

30

•Each year, Transport Canada generated medium and long term forecasts of Canadian air traffic

• Up to 14 years in the future

• Forecasts of • E/D passengers and passenger-kms

• Air cargo tonnage

• Aircraft movements (commercial and GA)

• Breakdowns into domestic, transborder and international

• Published forecasts provided national and regional forecasts(Atlantic, Ontario, Quebec, Prairies/Northern, Pacific)

• Most recent document was 2007

• Forecasts for individual airports available for purchase

Realizing the vision together

Transport Canada Forecasts

31

•Transport Canada forecasts based on two inter-connected models•Originally developed in 1976•Generation of traffic and allocation of traffic:

• Generation: PODM-V2 (Passenger Origin-Destination Model)• Forecasted Origin-Destination traffic

• E.g., Vancouver-Montreal; Toronto – Los Angeles, etc.

• Allocation: PTAM (Passenger Traffic Allocation Model)• Allocates forecasted passenger traffic to air carrier operations

• E.g., Forecast Vancouver-Montreal traffic allocated to direct service and to connections via Calgary, Toronto, etc.

Realizing the vision together

Transport Canada Forecasts

32

•The Generation-Allocation approach is similarapproach to that used in urban transport modeling

• Road systems

• Toll roads

• Public transport

•It allows changes in the network to affect traffic flows

•However, the Transport Canada modeldoes not address congestion or other constraints

• This is often a major factor in urban models

Realizing the vision together

PODM (Passenger Origin-Destination Model)

33

•PODM:

Realizing the vision together

PODM (Passenger Origin-Destination Model)

34

•PODM (Translated):• Based on a zonal system

• PODM forecasts traffic between zones

• Domestic zones:• Approximately 36 zones based around major airports

• E.g., Toronto - Pearson, City Centre, Hamilton, Oshawa, Buttonville, Kitchener)Vancouver – Vancouver, Abbotsford, Vancouver Harbour

• Transborder zones:• Approximately 20 zones

• E.g., Los Angeles, New York, etc.

• International zones:• Country (e.g., UK) or continental (e.g., Africa)

Realizing the vision together

PODM (Passenger Origin-Destination Model)

35

•PODM (Translated):• Gravity Model:

• Traffic between two zones is a function of:

• Model was directional

Zone A

Zone B

Attractors:PopulationGDPLinguistic similaritiesOther factors

Impeders:Air fareLevel of direct serviceTravel time by car

Realizing the vision together

PODM (Passenger Origin-Destination Model)

36

•Calibration/estimation required a lot of data:• O/D passenger Data

• Directional Origin-Destination Database

• Based on 10% sample of all air tickets

• Especially developed for the model by Statistics Canada

• Also used U.S. data for transborder

• Air Fare Data• Airfare Basis Survey

• Quarterly survey of domestic, transborder and international air passengers

• OAG schedule data• For determining direct services

• Socio-Economic Data• Population, GDP, etc.

Realizing the vision together

PODM (Passenger Origin-Destination Model)

37

•Separate models for full economy and discount economy

• Proxies for trip purpose but also included cross-elasticities

(switching between full and discount)

•Based on data from several years – 1995 to 2001

• Panel data: based on variation over time and between routes

•Calibrated to ensure that the model reasonably

matches historical data – backcasting

Realizing the vision together

PODM (Passenger Origin-Destination Model)

38

•Forecasting Traffic• Requires forecasts of input variables

• GDP, population:• Based on Conference Board of Canada and other sources

• Air Fare:• Required separate model (Cost and Fare Model)

• Air fare based on input costs:

fuel, labour, aircraft equipment, other

plus productivity improvements

Realizing the vision together

PTAM (Passenger Traffic Allocation Model)

39

•Having established the O/D flows, these need to be allocated to airlines

• Moving from O/D to E/D

•Also attempts to address the impact of O/D traffic on airline services

• Development of direct services

• Incremental frequencies

•Incorporates assumptions about future airline fleets, aircraft technology and load factors •Could require iteration of the PDOM model:

• Introduction of direct service could stimulate O/D traffic

Realizing the vision together

PTAM (Passenger Traffic Allocation Model)

40

•Overall assumption of greater allocation of traffic to direct services:

Source: Transport Canada Assumptions Report (2006-20)

Realizing the vision together

PTAM (Passenger Traffic Allocation Model)

41

•Contained specific assumptions about the development of new direct services:

Source: Transport Canada Assumptions Report (2006-20)

Realizing the vision together

Pros and Cons

42

•Positives• Arguably the most sophisticated air traffic forecasting system

in the world• The FAA approach is much more basic:

National traffic model based on econometric analysisPlus separate forecasts for some individual airports(Terminal Area Forecasts)

• Could model in a complex way the implication of changes to the airline network

• E.g., O/D – Edmonton to Europe• Starts connecting through Toronto (and other hubs)

• Direct start-ups which impacts on traffic flows through Toronto

• Similarly, the gateway impacts on Vancouver:E.g., YOW-YVR-HK now moves YOW-YYZ-HKG

Realizing the vision together

Pros and Cons

43

•Negatives• Data!

• O/D Passenger and Air Fare Basis data generally not available due to confidentiality concerns

• Used for national accounts requirements and other purposes

• Raw data needs processing – Transport Canada had a bespoke data pulls to suit forecasting requirements

• More restrictive than U.S. equivalents (available to U.S. citizens)

• Alternative commercial sources are available – MIDT, DIIO, etc.

• But costs are high:• Tens of thousands of dollars per airport

• Prohibitive for individual airports but might still be useful for system-wide purposes (e.g., Nav Canada, CATSA)

Realizing the vision together

Pros and Cons

44

•Negatives

• Complexity

• Resources

• Required a small full-time team to maintain

• Costs of model maintenance, calibration and result production

development beyond the capabilities of individual airports and most

other organisations

Alternative Approaches

Realizing the vision together

Single Airport Forecasts

46

•Range of methodologies available:

• Time series / trend analysis

• Bottom-up / schedule based

• Econometric models

• Market share models

•A combination of these approaches can be used

Realizing the vision together

Single Airport Forecasts

47

•Time series / trend• Based on historical traffic growth rates

• Statistical techniques (ARIMA)

• “Historically has grown at 3.5% per annum so will grow at similar rates in the future”

• Can also reference global forecasts by Boeing, Airbus, FAA, IATA, etc.

Realizing the vision together

Single Airport Forecasts

48

•Bottom-up• Tend to be supply-side:

• Development of new routes

• Increase in frequencies and changes in gauge

• Route-by-route forecasts of air service and passenger volumes

• Can be based on announced schedules in the short term

• Guided by fleet acquisitions in the medium term

• Harder for long term forecasts

Realizing the vision together

Single Airport Forecasts

49

•Econometric models• Traffic as a function of:

• GDP (or GDP per capita)

• Personal income

• Population

• Air fare

• One-off factors: SARS, air failure, 9/11 (historical events)

• Separate models can sometime be developed for individual markets (Domestic, transborder, international)

• Can be seen as simplified version of the Transport Canada model:• One zone (airport) to small number of destination zones

• Dependent on the data available for airport

Realizing the vision together

Single Airport Forecasts

50

•Econometric models• Requires forecasts of explanatory variables

• Generally good sources available for GDP, population, etc.

• Inclusion of air fare can be problematic:• Hard to obtain historical data especially for a long time series

(need 10 years at least)

• Technical issues – fares are an endogenous variable, affected by demand and supply conditions

• Requires use of advanced statistical techniques(Two stage least squared regressions)

• Air fares also need to be forecast, e.g., using a airline cost model

• As a result, air fares are often not included in the analysis

Realizing the vision together

Single Airport Forecasts

51

•Yield Trend (U.S.)

0

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

1945

1948

1951

1954

1957

1960

1963

1966

1969

1972

1975

1978

1981

1984

1987

1990

1993

1996

1999

2002

2005

2008

Pass

enge

r Yie

ld (U

S$ P

er R

PKM

)

Inflation Adjusted 2008 Dollars

Nominal Dollars

1958Boeing 707

enters service

1970Boeing 747

enters service

1978U.S. Airline Deregulation Act

2001September 11th terrorist attacks

Realizing the vision together

Single Airport Forecasts

52

•Market Share Model• Generally involves forecasting the airport’s share of some aggregate

measure of air traffic (national traffic, regional traffic)

• Generally used where there are a number of airports realistically competing for the same traffic

• E.g., UK – five airports compete for the same traffic in London alone(Heathrow, Gatwick, Luton, Stansted, London City)

• New York market

• San Francisco / Oakland

• Few examples in Canada

Concluding Comments -Uncertainty

Realizing the vision together

Addressing Uncertainty

54

•Standard approach to uncertainty in both the Transport Canada forecasts and single airport forecasts

• Base case with low and high forecasts

Realizing the vision together

Addressing Uncertainty

55

Realizing the vision together

Addressing Uncertainty

56

•High-Base-Low:

• Everything is bad or everything is good all at once

• Variation tends to be arbitrary – why are these the outer bounds?

• No information or assessment of likelihood

• Often the range is not that large (+/- 25%) –history has shown us that bigger deviations are possible

• Has little input into the planning process• Low can be of interest for financing

Realizing the vision together

Addressing Uncertainty

57

•Other approaches:

• “What-ifs”

• Sensitivity tests•These provide some information on the impact of specific factors or the outcome of certain events•Again, can be arbitrary without reference to the likelihood of such an outcome•More on advanced approaches to uncertaintythis afternoon…

Thank You!www.intervistas.com

Addressing Uncertainty in Air Traffic forecasting

InterVISTAS Consulting Group

10 April 2013

Realizing the vision together

Outline

60

•Consequences of uncertainty•Causes of uncertainty•Identifying and evaluatingrisk•Incorporating risk intoforecasting

Consequences of Uncertainty

Realizing the vision together

Washington Dulles International Airport

62

0

5

10

15

20

25

1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014

Pass

enge

r Enp

lane

men

ts (M

illio

ns)

Actual Traffic

TAF 2000

TAF 2001

TAF 2003

TAF 2004

TAF 2005

TAF 2009

Realizing the vision together

Lambert-St. Louis International Airport

63

0

5

10

15

20

25

30

1985 1990 1995 2000 2005 2010 2015

Pass

enge

r Enp

lane

men

ts (M

illio

ns)

Actual TrafficTAF 1998TAF 2001TAF 2002TAF 2003TAF 2009

TWA declaresbankruptcy

TWA declaresbankruptcy for

the second time

TWA declaresbankruptcy for

the third time andAA buys TWA.Construction of

new runway beginsAA reduces

services at STL

AA terminates itsfocus city at STL

Realizing the vision together

Louis Armstrong New Orleans International Airport

64

0

1

2

3

4

5

6

7

8

1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014

Pass

enge

r Enp

lane

men

ts (M

illio

ns)

Actual TrafficTAF 2001TAF 2004TAF 2005TAF 2009

Hurricane Katrina

Realizing the vision together

Bellingham International Airport

65

0

50

100

150

200

250

300

350

400

450

1985 1990 1995 2000 2005 2010 2015 2020

Pass

enge

r Enp

lanem

ents

(Tho

usan

ds)

Actual Traffic

TAF 2000

TAF 2003

Master Plan Forecast

United Express / SkyWest exits in 2001

Allegiant enters in August 2004 and rapidly develops service

Allegiant opens base at BLI in January 2008

Realizing the vision together

Baltimore/Washington International Thurgood Marshall Airport

66

0

2

4

6

8

10

12

1972

1974

1976

1978

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

2004

2006

2008

2010

Pass

enge

r Enp

lane

men

ts (M

illio

ns)

Actual Traffic

1987 Master Plan Forecast (Baseline)

Piedmont announces hub

First Gulf Warand recession

Southwest Airlines launches services

9/11 andrecession

Recession

U.S. Airways "de-hubs"

Causes of Uncertainty

Realizing the vision together

Causes of Uncertainty

68

•Economics• Recessions and booms

• Regional conditions (closure of a local business)

• Fuel prices

•Airline Strategy• Start, expand, contract or shut service

• Transit traffic is fungible

•Airline failure/collapse• Canadian, TWA, Swiss, Sabena

•Regulatory / policy• Deregulation contributed to hubbing, changes in aircraft size, LCCs

• Also taxes, security, bilaterals

Realizing the vision together

Causes of Uncertainty

69

•Technology• More economical aircraft make new routes possible

• Implications for cargo – smaller aircraft reduce bellyhold; A380 has relatively small cargo space

•Airport Competition• New airports emerge as competitors, e.g., on the U.S. border

•Social / cultural • Concerns about environmental impacts

• Use of new communications media (+ve or –ve impact?)

•Shock events• 9/11

• SARS

Identifying and Evaluating Risk

Realizing the vision together

Identifying and Evaluating Risk

71

•Develop a risk register• Identify and evaluate various risks affecting the airport

• What is the particular risk?

• What is its likelihood?

• What is the impact if it occurs (short and long term)

•Information can be elicited from the airport team• Strategy

• Marketing

• Facilities

• Finance

•Can also examine historical examples• What was the impact of SAR on Toronto; how long was the recovery?

Realizing the vision together

Identifying and Evaluating Risk

72

•Presenting the results – Heat diagramLi

kelih

ood

Very High

High

Moderate

Low

Very Low

Very Low Low Moderate High Very

High

Impact on Activity

Realizing the vision together

Identifying and Evaluating Risk

73

Economic recession

Fuel price spikes

New FAA taxes

Terrorist attack

Loss/failure of Carrier X

Entry of new carrier(e.g., LCC)

Pandemic

Open Skies Liberalization

High Speed Rail Competition

Major tourism event

Increased security requirements

New aircraft technology

Economic boom

5%

10%

15%

20%

25%

30%

35%

-5 -4 -3 -2 -1 0 1 2 3 4 5

Prob

abilty

Impact Opportunity >< Threat

Macroeconomic

MarketRegulatory/PolicyTechnology

Social/Cultural

Key:

Shock Event

Use of internet for meetings

Incorporating Risk into Forecasting

Realizing the vision together

Incorporating Risk into Forecasting

75

•Risk analysis augments not replaces traditional forecasting

Time

Enpl

aned

Pas

seng

ers

Original Forecast

Traffic impact of carrier exit and partial recovery

Realizing the vision together

Incorporating Risk into Forecasting

76

•A simple approach:• Scenario analysis based on the risk register

• Development scenarios based on the high probability and high impact events (both positive and negative)

• Similar to high/low approach, but:• Based on comprehensive assessment of risk

• Scenario can be produced to examine extreme events – stress testing

• Should be a focus of planning decisions

• However, the approach still lacks information on likelihood or probability

Realizing the vision together

Incorporating Risk into Forecasting

77

68

242

280

329

401

110

123

152

114151

255

354

162

136150

195

25

2840

61

0

50

100

150

200

250

300

350

400

45019

85

1987

1989

1991

1993

1995

1997

1999

2001

2003

2005

2007

2009

2011

2013

2015

2017

2019

2021

Pass

enge

rs E

npla

nem

ents

(Tho

usan

ds)

Actual Traffic 1980-2000

Post-Masterplan Traffic (2000-2010)

Masterplan Forecast

Extreme Upside Scenario

Extreme Upside with New Carrier Exit

Extreme Downside Scenario

Realizing the vision together

Incorporating Risk into Forecasting

78

•A more advanced approach:• Monte Carlo simulations• Uses randomisation / probabilities to explore uncertainty

• Model inputs are probability distributions rather than fixed numbers

• Using computers, the models can be run multiple times, each time with the inputs randomly generated

• With enough iterations, the range of outcomes can be determined and probabilities applied to them

• Historical note: first used at Los Alamos in design of shiled for nuclear reactors

• Has also been used in finance, project planning, telecoms design, medicine,…

Realizing the vision together

Incorporating Risk into Forecasting

79

•Normal distribution

-8% -7% -6% -5% -4% -3% -2% -1% 0% 1% 2% 3% 4% 5% 6% 7% 8%

Deviation from Long-Term Economic Growth Rate

10% to 90% Range

Pro

babi

lity

Realizing the vision together

Incorporating Risk into Forecasting

80

•Pert Distribution

0 500 1,000 1,500 2,000 2,500

Loss of Enplaned Passengers (Thousands)

10% to 90% Range

Pro

babi

lity

Realizing the vision together

Incorporating Risk into Forecasting

81

Realizing the vision together

Incorporating Risk into Forecasting

82

•Determining the distribution (data based):

0

10

20

30

40

50

60

70

-1.5% -0.5% 0.5% 1.5% 2.5% 3.5% 4.5% 5.5% 6.5% 7.0%

Freq

uenc

y

GDP Growth Rate (Mid-Point)

Histogram of GDP Growth Data

Fitted Distribution

Realizing the vision together

Incorporating Risk into Forecasting

83

•Other events based on judgement (or combination of judgement and data)

•Can involve a complex set of connected inputs

•E.g., exit of a carrier:• Probability of exit

• Impact of exit – loss of traffic (can be randomised)

• Extent of recovery (also can be randomised)

• Time to recover (also can be randomised)

Realizing the vision together

Incorporating Risk into Forecasting

84

•Running the Monte Carlo (Sample)

Realizing the vision together

Incorporating Risk into Forecasting

85

0

5

10

15

20

25

30

35

Year 0 Year 5 Year 10 Year 15 Year 20

Annu

al E

npla

ned

Pass

enge

rs (M

illio

ns)

Most Likely Forecast

25th / 75th Percentile Range

10th / 90th Percentile Range

5th / 95th Percentile Range

Realizing the vision together

Incorporating Risk into Forecasting

86

0%

20%

40%

60%

80%

100%

0%

2%

4%

6%

8%

10%

12%

14%2,

500

2,75

0

3,00

0

3,25

0

3,50

0

3,75

0

4,00

0

4,25

0

4,50

0

4,75

0

5,00

0

5,25

0

5,50

0

5,75

0

6,00

0

6,25

0

6,50

0

6,75

0

7,00

0

7,25

0

7,50

0

7,75

0

Cum

ulat

ive

Prob

abili

ty

Prob

abili

ty

Passenger Enplanements (Thousands)

Probability Density (Left Hand Scale)

Cumulative Probability (Right Hand Scale)

Realizing the vision together

Incorporating Risk into Forecasting

87

•Can answer questions like:• What is the probability that passenger traffic growth will exceed

4% per annum over the next 20 years?

• What is the probability that passenger traffic will be greater than 20 million in five years time?

• What is the probability that passenger traffic in 2020 will be less than 25 million?

•Obvious applications for financial analysis

•But can also be incorporated into planning…

Realizing the vision together

Incorporating Risk into Forecasting

88

Thank You!www.intervistas.com

Canadian Forecasting WorkshopSession 5: 

Current Outlook

Dr. Mike TrethewayInterVISTAS Consulting Chief Economist

Economic Update

Realizing the vision together

Econo-geek VocabularyIf there is a recovery, will it be? • V-shaped

• A rapid recovery back to previous level• Most recessions are V-shaped

• U-shaped• A period of stagnation, with a slow recovery

• L-shaped• An extended period of stagnation

• Japan 1990s. Great Depression

• W-shaped• A V-shaped recovery, followed by another recession

• US, 1970s92

Realizing the vision together

US Real GDP Growth (Historical)

93

Sources: Historical – Bureau of Economic Analysis; Recessions as defined by the National Bureau of Economic Research

Realizing the vision together94

Sources: Historical – Bureau of Economic Analysis; Recessions as defined by the National Bureau of Economic Research

W WW

US Real GDP Growth (Historical)

Realizing the vision together95

Sources: Historical – Bureau of Economic Analysis; Recessions as defined by the National Bureau of Economic Research

W W W V VV

US Real GDP Growth (Historical)

Realizing the vision together

US Real GDP- recent

96

Source: NBER 27May2010, BEA 23Mar2013

Recession begins

Recession has ended

Realizing the vision together97

US Real GDP Growth - Forecast

Source: 2007-2012 U.S. Department of Commerce, Bureau of Economic Analysis2013-2017 International Monetary Fund, World Economic Outlook Database, April 2011

Forecast DataA

nnua

lized

Y-O

-Y G

row

th R

ate

Realizing the vision together

US Real GDP - Risk

• The recent recession had V-shaped recovery

• But there is still risk in the recovery • Managing contraction of Fed Assets

• Without 2nd recession• Without inflation

• will be a challenge• Risk: less than 50%, more than 25%

98

Realizing the vision together

US Real GDP (Historical)

99Sources: Historical (1946 to 2012) – Bureau of Economic Analysis;

Recessions are wiggles in a steadily growing economy

Realizing the vision together

Canada Real GDP (Historical)

100Sources: Historical Canada GDP (1961 to 2012) – Statistics Canada.

Recessions are wiggles in a steadily growing economy

Realizing the vision together

Canada Real GDP Growth

101

Sources: Historical and Forecast Data from International Monetary Fund, World Economic Database, October 2012.

Ann

ualiz

ed Y

-O-Y

Gro

wth

Rat

e

Historical Data

Forecast Data

Realizing the vision together102

Real GDP Growth - Mexico

Sources:

Historical Data: Mexico: International Monetary FundForecast Data: Mexico: International Monetary Fund

2 Recessions

Fuel Prices

Realizing the vision together

Fuel Cost per Litrefor Canadian Air Carriers

Source: 1980-1990- Statistics Canada, Aviation in Canada2006-2011- Statistics Canada, 51-004-X

104

Realizing the vision together

Source: 1980-1990- Statistics Canada, Aviation in Canada2006-2011- Statistics Canada, 51-004-X

Fuel Cost Percentage of Operating Expense Canadian Air Carriers

105

Realizing the vision together

Fuel Litres per RPK

Source: 1970-1990- Statistics Canada, Aviation in Canada 2006-2011- Statistics Canada, 51-004-X. Transport Canada

106

Realizing the vision together

Fuel Prices - historical

107

$-

$20

$40

$60

$80

$100

$120

$140

$160

Jan-

03A

pr-0

3Ju

l-03

Oct

-03

Jan-

04A

pr-0

4Ju

l-04

Oct

-04

Jan-

05A

pr-0

5Ju

l-05

Oct

-05

Jan-

06A

pr-0

6Ju

l-06

Oct

-06

Jan-

07A

pr-0

7Ju

l-07

Oct

-07

Jan-

08A

pr-0

8Ju

l-08

Oct

-08

Jan-

09A

pr-0

9Ju

l-09

Oct

-09

Jan-

10A

pr-1

0Ju

l-10

Oct

-10

Jan-

11A

pr-1

1Ju

l-11

Oct

-11

Jan-

12A

pr-1

2Ju

l-12

Oct

-12

Jan-

13A

pr-1

3

U.S

. $ p

er b

arre

l

Crude Oil Spot PricesJanuary 2003 to April 2013

Realizing the vision together

Fuel Prices

108

$-

$20

$40

$60

$80

$100

$120

$140

$160

Jan-

03A

pr-0

3Ju

l-03

Oct

-03

Jan-

04A

pr-0

4Ju

l-04

Oct

-04

Jan-

05A

pr-0

5Ju

l-05

Oct

-05

Jan-

06A

pr-0

6Ju

l-06

Oct

-06

Jan-

07A

pr-0

7Ju

l-07

Oct

-07

Jan-

08A

pr-0

8Ju

l-08

Oct

-08

Jan-

09A

pr-0

9Ju

l-09

Oct

-09

Jan-

10A

pr-1

0Ju

l-10

Oct

-10

Jan-

11A

pr-1

1Ju

l-11

Oct

-11

Jan-

12A

pr-1

2Ju

l-12

Oct

-12

Jan-

13A

pr-1

3

U.S

. $ p

er b

arre

l

Crude Oil Spot PricesJanuary 2003 to April 2013

2 year dramatic swingPrices rose by 250%Then crashed to 66%of original price

Realizing the vision together

Crude Oil Price Futures

109

$-

$20

$40

$60

$80

$100

$120

$140

$160

Jan-

03A

pr-0

3Ju

l-03

Oct

-03

Jan-

04A

pr-0

4Ju

l-04

Oct

-04

Jan-

05A

pr-0

5Ju

l-05

Oct

-05

Jan-

06A

pr-0

6Ju

l-06

Oct

-06

Jan-

07A

pr-0

7Ju

l-07

Oct

-07

Jan-

08A

pr-0

8Ju

l-08

Oct

-08

Jan-

09A

pr-0

9Ju

l-09

Oct

-09

Jan-

10A

pr-1

0Ju

l-10

Oct

-10

Jan-

11A

pr-1

1Ju

l-11

Oct

-11

Jan-

12A

pr-1

2Ju

l-12

Oct

-12

Jan-

13A

pr-1

3Ju

l-13

Oct

-13

Jan-

14A

pr-1

4Ju

l-14

Oct

-14

Jan-

15A

pr-1

5Ju

l-15

Oct

-15

Jan-

16A

pr-1

6Ju

l-16

Oct

-16

Jan-

17A

pr-1

7Ju

l-17

Oct

-17

Jan-

18A

pr-1

8Ju

l-18

Oct

-18

Jan-

19A

pr-1

9Ju

l-19

Oct

-19

U.S

. $ p

er b

arre

l

Crude Oil Spot Prices & Crude Oil Futures PricesJanuary 2003 to December 2019

Spot Prices

FuturesPrices

Source: Spot Prices from U.S. EnergyInformation Administration. Futures Prices from

Realizing the vision together

Oil Price Forecast

110

Realizing the vision together

Oil Price Consensus Forecast

111

•Its not down•Its not back to $145

Realizing the vision together

2 Sigma Range of Forecasts

Forecast 95% ranges

$-

$20

$40

$60

$80

$100

$120

2010 2011 2012 2013 2014 2015 2016

upper 2 sigma

average

low er 2 sigma

112

Realizing the vision together

2 Sigma Range of ForecastsForecast 95% ranges

$-

$20

$40

$60

$80

$100

$120

2010 2011 2012 2013 2014 2015 2016

upper 2 sigma

average

low er 2 sigma

113

•Everyone seems to agree

Realizing the vision together

Forecast 95% ranges

$-

$20

$40

$60

$80

$100

$120

2010 2011 2012 2013 2014 2015 2016

upper 2 sigma

average

low er 2 sigma

2 Sigma Range of Forecasts

114

•Everyone seems to agree

Note the scale

Realizing the vision together

If History Repeats the Swing….Oil Price with Full Historical Range

$-

$50

$100

$150

$200

$250

$300

2010 2011 2012 2013 2014 2015 2016

Upper range

Base price forecastLower range

115

Note the scale

Forecast Components

Realizing the vision together

Source: Statistics Canada Average Fare data, Cat. 51 -004-Xp = preliminaryMajor Air Carriers include Air Canada (mainline & AC Jazz), WestJet, Air Transat and Canada 3000

Average Fare: CanadaNominal: Not Adjusted for Inflation

117

Realizing the vision together

Sources: Statistics Canada Average Fare data, Cat. 51 -004-XStatistics Canada Consumer Price Index

p = preliminary air fare dataMajor Air Carriers include Air Canada (mainline & AC Jazz), WestJet, Air Transat and Canada 3000

Real Average Fare: Canada

118

Realizing the vision together

Source: Transport Canada Registered Commercial Aircraft database

Commercial Aircraft: Canada

119

Realizing the vision together

Passengers per Aircraft- Canada1980-2011

120

Source: InterVISTAS Calculations with data from: Aviation in Canada (1980-1990) and Table 401-0009, Statistics Canadaand Air Carrier Traffic at Canadian Airports. Statistics Canada

Realizing the vision together

Seats per Aircraft- Canada1980-2011

121

Source: InterVISTAS Calculations with data from: Aviation in Canada (1980-1990) and Table 401-0009, Statistics Canadaand Air Carrier Traffic at Canadian Airports. Statistics Canada and Transport Canada.

Realizing the vision together

Total GA Movements for Canada

122

Source: Statistics Canada, Table 401-007, 401-0015 and 401-0021.

Realizing the vision together

Cargo Tonnes Reporting: Statistics Canada vs. Actual Site Statistics

123

Airport

Statistics Canada (Tonnes)

Actual Airport Statistics (Tonnes)

Site to Stats Can Ratio

Calgary Intl, Alta. 83,524 116,000 1.39

Edmonton Intl, Alta. 22,955 36,112 1.57

Montréal/Mirabel Intl, Que. 66,899 95,518 1.43

Montréal/Pierre Elliott Trudeau Intl, Que. 76,623 105,113 1.37

Toronto/Lester B Pearson Intl, Ont. 339,065 492,171 1.45

Vancouver Intl, B.C. 186,385 223,878 1.20

Winnipeg/James Armstrong Richardson Intl, Man. 65,254 175,000 2.68

Source: Statistics Canada, 51-203-X. Individual Airport reports.

Realizing the vision together

Cargo Tons Gap: Site Stats to Stats Can

124

Source: InterVISTAS calculations with data from: Statistics Canada, 51-203-X. Individual Airport reports.

Realizing the vision together

US vs Canada Pax Traffic1990-2012

125

Source: InterVISTAS Calculations with data from: Canada-Air Carrier Traffic at Canadian Airports. Statistics CanadaUS- 1960-2006 ATA , 2007-2012 BTS .

Realizing the vision together

Load Factor- Canada

Source: Aviation in Canada, Statistics Canada. Transport Canada

126

Realizing the vision together

Commercial Aircraft Movements

Source: Aviation in Canada (1980-1990) and Table 401-0009, Statistics Canada.

127

Realizing the vision together

Annual Turboprop + Regional Jet Percentages

79% 72% 74% 75%

Scheduled Flight Frequency: Domestic Canada

128

Source: Official Airline Guide Schedule Data, full year data for 1998, 2002, 2007, and 2012.

Realizing the vision together

Annual Turboprop + Regional Jet Percentages

49% 37% 42% 50%

Scheduled Seat Capacity: Domestic Canada

129

Source: Official Airline Guide Schedule Data, full year data for 1998, 2002, 2007, and 2012.

Realizing the vision together

Domestic Canada Scheduled Flight Frequency by Aircraft Body Type

Widebody Regional Jets TurbopropsLegend:

Domestic Canada Scheduled Seat Capacity by Aircraft Body Type

Narrowbody

130Source: Official Airline Guide Schedule Data, full year data for 1998, 2002, 2007, and 2012.

Realizing the vision together

Canadian Air Carrier Total Revenue and Expenses

Source: Statistics Canada, 51-004-X

131

Realizing the vision together

Return on Assets for Canadian Air Carriers

Source: Statistics Canada, 1980-1985- Aviation in Canada. 2005-2011, 51-004-X.

132

Realizing the vision together

Accidents for Canadian Commercial Aircraft

Source: 2001-2011 Statistics Canada, 51-004-X. Transportation Safety Board of Canada1970-1990 Statistics Canada, Aviation in Canada

133

Realizing the vision together134

Thank You

Subscribe to Monthly Aviation Intelligence Reportwww.InterVISTAS.com

Recommended