arthur yang - a1 poster
TRANSCRIPT
Econometric Analysis of Qantas Domestic and Jetstar AirwaysArthur Yang
School of Aerospace, Mechanical and Manufacturing Engineering, RMIT University, Melbourne, VIC 3001, Australia
Abstract
The aim of this project is to develop an econometric model that demonstrates the
relationship between endogenous and exogenous factors, and the demand for air
travel in the Australian domestic market. Based on the literature, eight
appropriate independent variables were identified: Gross Domestic Product
(GDP), GDP per capita, air fare prices, population size, unemployment rate,
interest rates, bed spaces, and jet fuel prices. Through multiple regression
analysis, these variables were used to build models to predict Qantas Domestic
revenue passenger kilometres (RPKs) and Jetstar Airways RPKs. For Qantas, the
most important explanatory variables were IATA global airline yield which
measures airline profitability, GDP and the unemployment rate. For Jetstar, only
price (Real Restricted Economy air fare) and income (GDP per capita) were
statistically significant. The forecasts developed from the models face several
limitations, but show growth due to rising GDP and GDP per capita. Overall the
research affirms findings from the literature that price and income play a primary
role in determining air travel demand whilst other macroeconomic and social
factors play a smaller role.
1. IntroductionEconometric models have been used for decades to explain or forecast air traffic
demand (Wang & Song 2010). These models make use of historical data through
statistical regression analysis, which allows measurements of the relationship
between demand and its determinants (Wang & Song 2010). This project uses
quarterly traffic, air fare, and macroeconomic data from 2006-2015 to develop an
appropriate model to be used in forecasting.
Econometric models are useful because they can be used as a forecasting tool.
Accurate forecasts are essential to an airline’s strategic plans and its ultimate
financial performance. Forecasts assist airlines in scheduling, developing fleet
requirements, route development, product planning, ascertaining station staffing
and facility requirements, pricing, marketing and advertising (Grosche, Routhlauf
& Heinzl 2007; Doganis 2010; Radnoti 2002).
Qantas Group has experienced significant highs and lows throughout the decade
of 2006-2015. The company has faced slowing economic growth, rising oil prices
for much of the period, weakened consumer confidence, and a strong Australian
dollar which encouraged more Australians to travel overseas, and less foreign
travellers to visit and fly domestically within Australia. Faced with such a
challenging and capricious environment, Qantas’ ability to develop a sound
strategic plan and deliver a superior product will depend partly on the quality of
their demand forecasts. This project aims to develop those forecasts through
econometric modelling.
2. Research Questions1. How much do macroeconomic factors and air fare prices impact passenger
demand for Qantas Domestic’s services?
2. How much do macroeconomic factors and air fare prices impact passenger
demand for Jetstar Airways’ services?
3. MethodologyThe econometric model was developed using the ICAO Manual of Air Traffic
Forecasting (2006) model. The main steps are:
Final model equations
Qantas Domestic
log (Qantas Domestic RPK) = log(-11.521) + -0.655log (Yield) + 1.559log (GDP) + -
0.471log (Unemployment Rate)
Jetstar Airways
log (Jetstar Airways RPK) = log(-52.604) + -0.37log (Real Restricted Economy Air
Fare) + 6.405log (GDP per capita)
5. DiscussionThe final models produced for both Qantas Domestic and Jetstar Airways are relatively
simplistic. Of the eight explanatory variables identified from the literature review, the
Qantas model used three: IATA Global Yield, GDP and Unemployment; whilst the
Jetstar model used two: Real Restricted Economy Air Fare and GDP per capita.
Whilst attempting to develop the Qantas model, no price variable except IATA Global
Yield proved to be statistically significant. It would have been more preferable to use
the yield for all domestic Australian airlines instead for relevance, but data for this was
limited.
The forecasts produced from the models have some limitations. For the Qantas model,
yield was held constant, and forecast GDP and unemployment figures were retrieved
from the International Monetary Fund’s World Economic Outlook. Real GDP is
expected to grow at 2.5% in 2016, 3% in 2017, and then 2.8% until 2021. The Outlook
forecasts unemployment only until 2017 and expects it to remain steady at around
5.8%. Therefore the majority of the growth depicted by the forecast is driven by GDP
changes only.
The forecast for the Jetstar model faces similar issues. Holding air fares constant, the
only variable influencing the forecast is GDP per capita. As GDP is expected to grow
faster than the population (2.8% to 1.7%), the forecast growth in Jetstar’s RPKs is
derived solely from the expected growth of Australia’s GDP per capita.
Dr. Graham WildLecturer and Project Supervisor
RMIT Aviation
& Aerospace
0
1000
2000
3000
4000
5000
6000
7000
8000
20
06
Q1
20
06
Q3
20
07
Q1
20
07
Q3
20
08
Q1
20
08
Q3
20
09
Q1
20
09
Q3
20
10
Q1
20
10
Q3
20
11
Q1
20
11
Q3
20
12
Q1
20
12
Q3
20
13
Q1
20
13
Q3
20
14
Q1
20
14
Q3
20
15
Q1
20
15
Q3
RP
K (
MIl
lion
s)
Quarter
Qantas Domestic Revenue Passenger Kilometres
0
500
1000
1500
2000
2500
3000
3500
4000
4500
20
06
Q1
20
06
Q3
20
07
Q1
20
07
Q3
20
08
Q1
20
08
Q3
20
09
Q1
20
09
Q3
20
10
Q1
20
10
Q3
20
11
Q1
20
11
Q3
20
12
Q1
20
12
Q3
20
13
Q1
20
13
Q3
20
14
Q1
20
14
Q3
20
15
Q1
20
15
Q3
RP
Ks
(Mill
ion
s)
Quarter
Jetstar Revenue Passenger Kilometres
0
50000
100000
150000
200000
250000
300000
350000
400000
450000
20
06
Q1
20
06
Q3
20
07
Q1
20
07
Q3
20
08
Q1
20
08
Q3
20
09
Q1
20
09
Q3
20
10
Q1
20
10
Q3
20
11
Q1
20
11
Q3
20
12
Q1
20
12
Q3
20
13
Q1
20
13
Q3
20
14
Q1
20
14
Q3
20
15
Q1
20
15
Q3
GD
P (
MIL
LIO
NS)
QUARTER
Australian Real Gross Domestic Product
14500
15000
15500
16000
16500
17000
17500
20
06
Q1
20
06
Q3
20
07
Q1
20
07
Q3
20
08
Q1
20
08
Q3
20
09
Q1
20
09
Q3
20
10
Q1
20
10
Q3
20
11
Q1
20
11
Q3
20
12
Q1
20
12
Q3
20
13
Q1
20
13
Q3
20
14
Q1
20
14
Q3
20
15
Q1
20
15
Q3
GD
P P
ER C
AP
ITA
QUARTER
GDP Per Capita
0.000000
0.020000
0.040000
0.060000
0.080000
0.100000
0.120000
0.140000
0.160000
0.180000
0.200000
2006 2007 2008 2009 2010 2011 2012 2013 2014
Glo
bal
Yie
ld
Year
IATA Yield
0.00
20.00
40.00
60.00
80.00
100.00
120.00
20
06
Q1
20
06
Q3
20
07
Q1
20
07
Q3
20
08
Q1
20
08
Q3
20
09
Q1
20
09
Q3
20
10
Q1
20
10
Q3
20
11
Q1
20
11
Q3
20
12
Q1
20
12
Q3
20
13
Q1
20
13
Q3
20
14
Q1
20
14
Q3
20
15
Q1
20
15
Q3A
ir F
are
(Bas
e:
July
20
03
)
Quarter
Real Restricted Economy Air Fare Price
0.00
1000.00
2000.00
3000.00
4000.00
5000.00
6000.00
7000.00
8000.00
9000.00
10000.00
2006Q1
2006Q2
2006Q3
2006Q4
2007Q1
2007Q2
2007Q3
2007Q4
2008Q1
2008Q2
2008Q3
2008Q4
2009Q1
2009Q2
2009Q3
2009Q4
2010Q1
2010Q2
2010Q3
2010Q4
2011Q1
2011Q2
2011Q3
2011Q4
2012Q1
2012Q2
2012Q3
2012Q4
2013Q1
2013Q2
2013Q3
2013Q4
2014Q1
2014Q2
2014Q3
2014Q4
2015Q1
2015Q2
2015Q3
2015Q4
2016Q1
2016Q2
2016Q3
2016Q4
2017Q1
2017Q2
2017Q3
2017Q4
2018Q1
2018Q2
2018Q3
2018Q4
2019Q1
2019Q2
2019Q3
2019Q4
RP
K (
Mill
ion
s)
Quarter
Qantas Domestic RPK Forecast
Actual RPKs Forecast RPKs
1. Define the
problem
The desired forecast variable are RPKs for Qantas Domestic
and Jetstar Airways. The time horizon for the forecast will
be five years.
2. Select the
relevant causal or
explanatory
variables
The explanatory variables selected are: Gross Domestic
Product (GDP), GDP per capita, air fare prices, population
size, unemployment rate, interest rates, bed spaces, and jet
fuel prices
3. Collect data
for those
variables
Data were collected from various online sources, including
the: Qantas Investor Relations Website, IMF, BITRE, ABS,
RBA, and the US Energy Information Administration. Check
for multicollinearity by producing a correlation matrix and
exclude those explanatory variables that have high
multicollinearity.
4. Formulate the
model
The type of functional relationship between the dependent variable
and the selected explanatory variables need to be specified. The
ICAO Manual of Air Traffic Forecasting provides several
alternative mathematical forms that can be used. The main form
used for the project was multiplicative or log-log.
log 𝑌 = log 𝑎 + 𝑏 log𝑋1 + 𝑐 log𝑋2 +⋯+ 𝑧 log𝑋𝑛
5. Carry out an
analysis to test the
relationship being
hypothesised
Regression analysis was carried out in Microsoft Excel using the
Data Analysis ToolPak, which provided all the information. The
program provided summary statistics on the model coefficients,
their magnitudes and signs and statistical measures.
6. Establish the
model in the final
form
The best model is chosen after considering goodness of fit and the
statistical significance of the coefficients of the explanatory
variables.
7. Develop
forecasts of future
scenarios
In order to develop forecasts for the dependent variable, it is
necessary to obtain forecasts for the explanatory variables.
Forecasts of the explanatory variables were obtained from the
IMF’s World Economic Outlook publication.
4. Results
Dependent variable scatter plots
Explanatory variable bar charts/scatter plots
0
1000
2000
3000
4000
5000
6000
2006Q1
2006Q2
2006Q3
2006Q4
2007Q1
2007Q2
2007Q3
2007Q4
2008Q1
2008Q2
2008Q3
2008Q4
2009Q1
2009Q2
2009Q3
2009Q4
2010Q1
2010Q2
2010Q3
2010Q4
2011Q1
2011Q2
2011Q3
2011Q4
2012Q1
2012Q2
2012Q3
2012Q4
2013Q1
2013Q2
2013Q3
2013Q4
2014Q1
2014Q2
2014Q3
2014Q4
2015Q1
2015Q2
2015Q3
2015Q4
2016Q1
2016Q2
2016Q3
2016Q4
2017Q1
2017Q2
2017Q3
2017Q4
2018Q1
2018Q2
2018Q3
2018Q4
2019Q1
2019Q2
2019Q3
2019Q4
RP
Ks
(Mill
ion
s)
Quarter
Jetstar Airways RPK Forecast
Actual RPKs Forecast RPKs
6. ConclusionBased on the research, macroeconomic variables such as GDP and unemployment
play a much larger role in the demand for Qantas Domestic’s services than air
fares, which is somewhat expected from the literature.
For Jetstar, it also appears that macroeconomic variables such as income (GDP per
capita) play a greater role in driving demand than air fares. This seems to
contradict the literature, which states that in developed countries in Australia,
demand for leisure travel is driven by changes in price more than changes in
income.
ReferencesDoganis, R 2010, Flying Off Course: Airline economics and marketing, 4th edn, Routledge, Oxon.
Grosche, T, Routhlauf, F & Heinzl, A 2007, ‘Gravity models for airline passenger volume estimation’, Journal of Air Transport
Management, vol. 13, no. 4, pp. 175-183.
International Civil Aviation Organisation 2006, Manual of Air Traffic Forecasting, International Civil Aviation Organisation,
viewed 8 June 2016, available at
<http://www.icao.int/MID/Documents/2014/Aviation%20Data%20Analyses%20Seminar/8991_Forecasting_en.pdf>.
Radnoti, G 2002, Profit Strategies for Air Transportation, McGraw-Hill, United States.
Wang, M & Song, H 2010, ‘Air Travel Demand Studies: A Review’, Journal of China Tourism Research, vol. 6, no. 1, pp. 29-49.