Click for next page ( 24

The National Academies of Sciences, Engineering, and Medicine
500 Fifth St. N.W. | Washington, D.C. 20001

Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement

Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 23
24 CHAPTER 4 Software Approach and Design The objective of this work effort was to provide a practical This approach is designed to produce useful information mechanism for airports to assess the risk of fuel price uncer- for airport users. If there have been significant changes in tainty and other economic factors to their future development expectations about the economy or jet fuel prices in the recent programs and operations. Early on, it was determined that the past, some airport sponsors may be asked questions or have software to be developed for this project should allow a user concerns about future air service, which in turn would have to analyze either their own customized forecast of future air- important implications for their operating budgets and for port activity, or a default baseline forecast. In either case, the their development programs. For example, the recession that goal is to assess how such forecasts may be affected by changes began in December 2007 and the fuel spike of 2008 were not in fuel prices and other sources of uncertainty. To make this well-anticipated by airlines or by airports. As information on assessment, key assumptions that underlie the forecast, includ- these events became apparent, many airports were forced to ing expectations about fuel prices, economic growth, and other alter development plans or cut operating budgets (examples factors, must be considered. Then, the program should allow of these impacts are discussed in the following subsection). the user to undertake sensitivity studies by varying assump- Airport sponsors would benefit if they could quickly assess tions about the key drivers, with the software generating a the impacts of these unanticipated events on their operations range of likely outcomes based on these assumptions. and development plans. Perhaps more important, the spon- An important feature of the software that was developed sors would be able to anticipate questions and concerns from is the creation of confidence bands for the forecast, which are business partners (e.g., airlines, financial intermediaries) and generated using an analysis based on the historic range of provide useful information in their continuing dialogues. errors in expectations of jet fuel prices and GDP growth. This This approach focuses on the impacts of unanticipated approach answers a fundamental question: How might an air- events on existing forecasts. This makes sense because no port forecast be affected given the historic errors in expected single, overarching model will be capable of considering the future jet fuel prices and economic growth? many details that determine air service at specific airports. The software uses information from the heating oil futures Airport sponsors themselves are better positioned to know market (which has a close correspondence to jet fuel prices) and their local markets and develop local forecasts, and are also in data on GDP forecast errors to create confidence bands that the business of interacting with their partners (including air- reflect the risk to an airport's forecast due to these very-difficult- lines) to anticipate changes in air services. to-forecast variables.9 The software also generates a one-page report that summarizes key inputs and the results of the risk 4.1 Embedding Uncertainty analysis. The overall process is illustrated in Exhibit I-26, into Forecasts showing how the inputs to the statistical model developed ear- lier tie into an airport's assessment of the uncertainty associated While the air service statistical models explain a high per- with its activity forecasts. centage of the variation in observed seat offers over the past 20 years, their primary purpose is to aid airport decision mak- ers in projecting future activity at their airport. The software 9 Again, it is important to emphasize that there may be other major factors driv- developed for this project allows users to employ these mod- ing any given forecast that are unknown to the software and are not accounted els to project activity five years out (through 2014) from the for in the confidence bands. end of the historical data in 2009, and then to apply the pre-

OCR for page 23
25 Exhibit I-26. Combining existing forecasts with the risk model. Existing Forecast ACRP 03-15 Risk Model How Far Off Could Range of Future Future Existing The Forecasts Be Enplanements and Enplanements Forecast Based on Past Operations Taking and (TAF; other) Experience w Key Risk into Operations Air Service Drivers? Account USER CUSTOMIZATION* Local Air Service Drivers MODEL RISK ANALYSIS Local Income Macro Air Service Drivers Competition at Airport Jet Fuel Prices Competition from Nearby Airports GDP Growth Average Aircraft Size at Airport Inflation * Model incorporates default values for each airport dicted changes in activity to a baseline TAF or user-supplied software provides default values for projections of the explana- forecast. tory variables out to 2014, but the user can override these val- The underlying motivation for such an approach is that all ues and has full control over what values to assign to future forecasts are inherently uncertain, and it can be useful to be variables. able to measure that uncertainty by placing confidence bands In the current context, it is important to focus on the jet around the baseline projection. To get a better feel for such fuel cost and income variables, both of which are difficult uncertainty, consider the annual TAF forecasts produced by to predict. Given that air carrier schedules are set well in the FAA. The latest 2009 forecasts make long-term projec- advance, the lagged representation for fuel price is consis- tions of operations and enplanements out to 2030. Like any tent with the notion that airlines use current fuel prices to forecast, inaccuracies in the TAF tend to increase with the help make decisions about future service offers. In practi- number of future years. But even over a much shorter time cal terms, however, it is important to note that airlines typ- frame, the TAF forecasts can be somewhat inaccurate. ically make scheduling decisions more often than once per An analysis of the TAF was conducted for each year from year. Most U.S. carriers set seasonal schedules approxi- 2003 through 2008 that measured the accuracy of the airport mately six months in advance. forecasts relative to actuals for domestic operations and total However, given the volatility in world oil prices, relying enplanements from one to five years out.10 The results, bro- only on current or recent historic fuel prices as guides to what ken out by airport hub type, are shown in Exhibit I-27. As expected, the projections become less accurate the fur- Exhibit I-27. TAF accuracy one to five years out. ther out the projection period and the smaller the airport. But for airports of any size, the results suggest that it is important Based on 2003-2008 Forecasts (Mean Absolute Percentage Error) to be able to assess the uncertainty associated with airport activity forecasts; that is the major motivation for the software Domestic Operations described here. Years Ahead Forecast Hub Type 1 2 3 4 5 To use the air service models to help address this issue, it is Large 3.4% 10.2% 13.9% 18.4% 25.5% necessary to provide expected future values of the models' Medium 5.3% 12.5% 17.3% 22.0% 25.7% explanatory variables. Looking back to Exhibit I-23, for some Small 8.0% 13.9% 17.9% 22.7% 26.0% variables such as average seat size and the HHI, a reasonable Non-Hub 14.0% 20.4% 25.3% 31.9% 38.7% All 10.4% 16.8% 21.4% 27.1% 32.8% default assumption may be that next period's value will be the same as the latest current period value. But others, in particu- lar the jet fuel cost and income variables, can be quite volatile Enplanements Years Ahead Forecast and/or difficult to predict even one or two years ahead. The Hub Type 1 2 3 4 5 Large 3.9% 9.3% 12.4% 15.7% 20.4% Medium 5.5% 11.3% 14.5% 17.9% 19.3% 10Thus, six years of data (from 2003 through 2008) were used for the one-year Small 8.7% 12.3% 14.4% 17.1% 18.6% ahead analysis, five years (from 2003 through 2007) for the two-year ahead Non-Hub 15.6% 20.2% 23.9% 26.3% 27.9% analysis, etc. All 11.5% 16.1% 19.3% 22.0% 23.9%

OCR for page 23
26 Exhibit I-28. Historical volatility in oil prices. 80% Change in World Oil Prices vs. Prior Year 60% 40% 20% 0% -20% -40% -60% 87 88 89 91 92 93 95 96 99 00 02 03 04 06 07 08 86 90 94 97 98 01 05 19 19 19 19 20 20 20 20 19 19 19 19 19 20 20 20 19 19 19 19 19 20 20 Source: EIA, Annual Energy Outlook Retrospective Review, 2009 Report Exhibit I-29. Historical volatility in GDP growth. 150% Change in Annual GDP Growth vs. Prior Year 100% 50% 0% -50% -100% 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 Source: EIA, Annual Energy Outlook Retrospective Review, 2009 Report they may be several months ahead can lead to large projected The described annual models would indicate that one errors.11 Exhibit I-28 shows how recent volatility could cause should use today's jet fuel price to help project next year's large misses in predicting future fuel prices. seat departures at a given airport, but for practical purposes One possible way to obtain more accurate predictions of it is suggested that users consider looking at current prices future fuel prices would be to utilize the financial futures mar- for heating oil futures contracts at least several months out ket for crude oil or related commodities. Many U.S. airlines in order to get a better understanding of where jet fuel prices engage in fuel hedging strategies using heating oil futures con- may be headed. tracts. Heating oil prices are closely correlated with jet fuel An assessment of average national income growth sug- prices, and the futures market for heating oil is large and very gests similar findings; as shown in Exhibit I-29, the historic liquid.12 data series is quite volatile. This volatility can become sig- nificantly more pronounced if one considers variations in 11 This is the "random walk" theory of prices, which states that this period's price local income, which is the metric actually used in the air is simply equal to last period's price plus a random error. service models. 12 Although there is a futures contract for kerosene (which is the primary com- One of the major objectives of the modeling effort is to ponent of jet fuel) that trades on the Tokyo Commodities Exchange, it is denom- inated in Japanese yen, which would introduce foreign exchange risk for U.S. obtain reasonable estimates of the uncertainty in airport-level companies. operations and enplanement forecasts by providing likely