S Air transport market analysis and forecasting

O :
H
introduction to air traffic forecasting, market trend analysis,
M time series analysis, forecasting techniques, evaluating
T forecasting results
 Forecasting
 The basic concept of forecasting is to estimate past trends and project them
S forward.
O  To estimate past trends, analysts have to make many assumptions.
 They have to select a span of data (over time or under different circumstances),
H some principal drivers of the phenomenon being forecast, the form of these
M factors and of the mathematical model, and the future values of their drivers.
T  Different analysts choose these factors differently and obtain quite different
results.
  Airport master plans are developed on the basis of forecasts. From forecasts, the
S relationships between demand and the capacity of an airport's various facilities can
be established and airport requirements can be determined. Short-, intermediate-,
O and long-range (approximately 5-, 10-, and 20-year) forecasts are made to enable
H the planner to establish a schedule of development for improvements proposed in
M the master plan.
T Two types of forecasting methods are available to assist planners in the
decision- making process:
 qualitative
 quantitative.
  Qualitative forecasting methods rely primarily on the
judgment of forecasters based on their expertise and experience with
the airport and surrounding environment.
 Judgmental predictions of future airport activity tend not to be based on
S historical data, but by the foresight that certain experts have, based on
their knowledge of the current and potential future environment.
O Qualitative forecasts may almost be thought of as "educated guesses,"
H opinions, or "hunches" of future activity, although they tend to be just as
accurate as quantitative methods.
M
 Despite this, qualitative forecasts tend to require the support of some
T quantitative analysis to justify the forecasts to the public.
 Four of the more popular qualitative methods include Jury of
Executive Opinion, Sales Force Composite, consumer market
survey, and the Delphi method
 Jury of Executive Opinion
 The Jury of Executive Opinion
S method seeks the predictions of
management and administration of
O the airport and the airport's
H tenants. Given that these persons
are the closest to the day-to-day
M operations of the airport, and
T typically have extensive experience
in airport activity at this airport
and perhaps others as well, the
Jury of Executive Opinion tends to
yield fairly accurate qualitative
forecasts.
 Sales Force Composite
 The Sales Force Composite method
S seeks the judgment of airport
employees, and the employees of those
O firms that do business at the airport for
H their predictions of future activity. The
M theory behind this method is that the
employees, or "sales force," of the
T
airport have direct interaction with the
users of the airport, and may provide
accurate judgments as to future activity
based on this interaction.
 Consumer market survey

S  A consumer market survey seeks the opinions

of the consumer base of the airport,
O specifically airport passengers, cargo
H shippers, and users of aeronautically and non
M aeronautically based businesses located in the
airport vicinity and the surrounding
T community. Because it's this population that
will actually partake in airport activity in the
future, soliciting this population's judgment
through a consumer market survey is a
reasonable qualitative forecasting method.
 Delphi method
 The Delphi method is a qualitative
S forecasting method originally developed
by marketing researchers in private
O sector businesses.
H  In the Delphi method, a group of
M experts in the field of interest is
T identified and each individual is sent a
questionnaire. The experts are kept
apart and are unknown to each other.

  The independent nature of the process ensures that the responses are truly
S independent and not influenced by others in the group.
O  This forecasting method involves an iterative process in which all the responses and
supporting arguments are shared with the other participants, who then respond by
H revising or giving further arguments in support of their answers. After the process
M has been repeated several times, a consensus develops.
T
  ………………….forecasts may almost be thought of as "educated
S guesses," opinions, or "hunches" of future activity.
O
 Qualitative
H
 Quantitative
M
T
  Which of the following forecast method includes sending
S questionnaire to experts and they are kept apart from each other?
O
H
 Jury of executive opinion method
M
 Delphi method
T
  Which of the following is not a qualitative forecasting method?
S
O
H  Delphi method
 Trend analysis
M
 Consumer market survey
T
 Sales Force Composite
  Quantitative forecasting methods are those that use numerical data
S and mathematical models to derive numerical forecasts. In contrast to qualitative
methods, quantitative methods are strictly objective.
O  Because only numerical data are used, quantitative methods do not directly
H consider any judgment on the part of the forecaster.
 Quantitative methods are either used as standalone forecasting methods or used to
M support forecasts made under qualitative methods.
T  Quantitative methods include time-series or trend analysis models, which forecast future
values strictly on the basis of historical data collected over time, and causal models,
which attempt to make accurate predictions of the future on the basis of how one area of
historical data affects another.
 Time-series
 Aviation analysts typically work with time
S series, as in Table, and have to decide on the
number of periods. However, they may also
O work with data on several different situations
H for a common period, using what is known as a
cross-sectional approach.
M  For example, in trying to model the number
T of passengers, researchers could look at the
data for many cities in some years and would
have had to choose the number of cities.
 Using either time-series or cross-sectional
data, analysts have to choose the span of the
data.
 Trend analysis models
 Trend analysis is a
S technique used in technical
analysis that attempts to
O predict future business
H based on recently observed
trend data.
M  Trend analysis uses
T historical data, such as
price movements and trade
volume, to forecast the
long-term direction of
market sentiment.
S
O
FORECASTING FOR
H
M
AVIATION PLANNING
T
 AIR NAVIGATION SYSTEMS PLANNING

S
O
 Estimates of aircraft Peak-period planning
H movements parameters
M
T
  Capacity Planning: Air navigation systems need to be able to handle the
S expected traffic volume. By forecasting the number of flights and the expected
traffic flow, air navigation service providers can plan for the necessary capacity
O upgrades, maintenance, and staffing levels.
H
M  Resource Allocation: Accurate forecasting helps air navigation service
T providers to allocate resources efficiently. For example, knowing which routes are
likely to be busy can help them deploy additional air traffic controllers, radar
systems, and other equipment in advance.
  Safety: Air navigation systems must ensure the safety of aircraft and passengers.
S By forecasting weather conditions, air traffic flow, and other factors that could
impact flight safety, air navigation service providers can make informed decisions
O and take proactive measures to mitigate risks.
H
M  Cost Optimization: Forecasting can also help air navigation service providers to
T optimize costs by avoiding overstaffing or underutilization of equipment. Accurate
forecasting helps them to plan their resources and budgets effectively.
S
O
H
M
T
 AIRPORT PLANNING

S Aircraft movements
 Identification of
O
airport planning
H parameters
M Airport Traffic in peak
passengers periods
T

Alternative development strategies

Other parameters
 1. Capacity Planning: Airports need to be able to handle the expected passenger
and aircraft traffic volume. By forecasting the number of passengers, flights,
S and expected traffic flow, airport operators can plan for the necessary capacity
O upgrades, maintenance, and staffing levels.
2. Resource Allocation: Accurate forecasting helps airport operators to allocate
H resources efficiently. For example, knowing which gates and runways are likely
M to be busy can help them deploy additional staff and equipment in advance.
3. Infrastructure Development: Airports need to continually upgrade their
T infrastructure to meet the changing needs of the aviation industry. Forecasting
helps airport operators to plan for the construction of new runways, terminal
buildings, and other facilities, ensuring that they are ready to meet future
demand.
 1. Service Level Planning: Airports need to provide an adequate level of
S service to their passengers, airlines, and other stakeholders. By
forecasting demand, airport operators can plan for the necessary service
O
levels, such as the number of check-in desks, security checkpoints, and
H baggage carousels.
M 2. Revenue Management: Accurate forecasting helps airport operators to
T optimize their revenue streams by planning for future business
opportunities, such as attracting new airlines, increasing commercial
space, and developing new revenue streams.
 AIRLINE PLANNING

S Planning of routes and

Impact of fare
services
O structures
H
M
T Fleet planning Estimation of an airline’s market share
  Capacity Planning: Airlines need to be able to handle the expected passenger demand
and aircraft traffic volume. By forecasting the number of passengers, flights, and expected
S traffic flow, airlines can plan for the necessary capacity upgrades, maintenance, and
staffing levels.
O
H  Resource Allocation: Accurate forecasting helps airlines to allocate their resources
efficiently. For example, knowing which routes are likely to be busy can help them deploy
M additional aircraft and crew in advance.
T
 Fleet Planning: Airlines need to plan for the size and composition of their fleets, taking
into account factors such as passenger demand, route network, and fuel efficiency.
Forecasting helps them to make informed decisions about purchasing, leasing, and
retiring aircraft.
  Revenue Management: Accurate forecasting is critical for revenue
S management, which involves optimizing the revenue streams of airlines by
managing ticket pricing, seat inventory, and route scheduling. Forecasting helps
O airlines to understand demand patterns and make pricing and scheduling decisions
H that maximize revenue.
M
T  Marketing and Sales: Airlines need to plan their marketing and sales activities
based on forecasted demand. By accurately forecasting demand, airlines can tailor
their marketing and sales campaigns to target the right customers at the right time.
 Forecasting Process
 The forecasting process generally involves the following steps:
S
O  Define the purpose and scope of the forecast: The first step in the
forecasting process is to determine why the forecast is needed, what information is
H required, and who the intended audience is.
M
T  Collect and analyze relevant data: The next step is to gather and analyze
historical data related to the phenomenon being forecasted. This may involve
looking at past trends, patterns, and relationships between variables that could
affect the outcome of the forecast.
S  Select the appropriate forecasting method: There are many different
O methods of forecasting, ranging from simple time-series analysis to complex
machine learning algorithms. The choice of method will depend on the nature of
H the data and the complexity of the forecast.
M
T
  Build the forecast model: Based on the selected method, the forecast model is
S built using the historical data. The model may need to be calibrated and adjusted to
fit the specific circumstances of the forecast.
O
H  Validate the model: Once the forecast model is built, it needs to be validated to
M ensure that it is accurate and reliable. This may involve testing the model against
T known outcomes or using statistical measures to assess its performance.
  Make the forecast: With the validated model in place, the forecast can be made.
S The forecast should be communicated clearly to the intended audience, along with
any assumptions or limitations of the model.
O
H  Monitor and update the forecast: Finally, the forecast should be monitored
M and updated as new data becomes available or circumstances change. This allows
T for ongoing refinement of the model and improved accuracy over time.
 Evaluating forecasting results
 There are several ways to evaluate forecasting results in aviation, depending on
the specific type of forecast being made. Here are some common evaluation
S methods:
O 1. Accuracy measures: One of the most basic evaluation methods is to use
accuracy measures, such as mean absolute error (MAE) or mean absolute
H percentage error (MAPE), to compare the forecasted values with the actual
M values. These measures can help determine how well the forecast performed
overall.
T 2. Bias analysis: It is important to examine whether the forecast has a bias
towards overestimating or underestimating the actual values. A simple way to
do this is to calculate the mean error, which is the difference between the
forecasted and actual values.
 3. Forecasting interval analysis: When making forecasts, it is often useful
S to provide a range of possible outcomes. One way to evaluate the quality of
these intervals is to calculate the coverage probability, which is the
O percentage of actual values that fall within the forecasted interval.
H 4. Model fit: Another way to evaluate a forecast is to examine how well the
model used to generate the forecast fits the historical data. This can be
M done by looking at measures such as R-squared or root mean squared error
T (RMSE).
5. Forecasting value-added: Finally, it is important to consider the value
that the forecast provides. This can be assessed by comparing the forecast
to a baseline model, such as a simple trend or seasonal model, and
calculating the forecasting value-added (FVA).