Modeling Airline Fares

Transcription

1 Modeling Airline Fares Evidence from the U.S. Domestic Airline Sector Domingo Acedo Gomez Arturs Lukjanovics Joris van den Berg 31 January 2014

2 Motivation and Main Findings Which Factors Influence Fares? Distance Competition Seasonality Carrier Ticket class Economic situation Total passengers Hubs Our Model Results 22 explanatory factors included Overall accuracy of 50% of fare variation Adding Southwest increases accuracy to 55% ENAC Modeling Airline Fares 31 January / 16

3 DB1B Origin & Destination Database Itinerary ID Main Information Coupons Carrier Breaks Itinerary $ fare Fare class ENAC Modeling Airline Fares 31 January / 16

4 A Look Inside the Database ENAC Modeling Airline Fares 31 January / 16

5 Reducing the Database Which tickets do we keep for our study? 1 Round trip 2 Two or four coupons 3 Single carrier 4 No extreme fares 5 Economy class 6 Main majors & low cost carriers 7 Regular routes 8 Lower 48 states ENAC Modeling Airline Fares 31 January / 16

6 Data Flow ENAC Modeling Airline Fares 31 January / 16

7 Final Dataset 128,192 Observations 65% Direct Flights 35% Indirect Flights ENAC Modeling Airline Fares 31 January / 16

8 Descriptive Statistics ENAC Modeling Airline Fares 31 January / 16

9 Descriptive Statistics ENAC Modeling Airline Fares 31 January / 16

10 Model Results Dependent Variable: ln(avgweightedfare) Method: Least Squares Sample (adjusted): Included observations: after adjustments Variable Coefficient Std. Error t-statistic Prob. C AVERAGEROUTEPOPULATION/ TOTALPAX/ (TOTALPAX/1000) DISTANCE/ (DISTANCE/1000) H-INDEX/ CLASSRATIO CLASSRATIO FREQAIRPORT= DIRECT= CARRIER= AS CARRIER= FL CARRIER= AA CARRIER= DL CARRIER= NW CARRIER= UA CARRIER= US CARRIER= 9E CARRIER= B CARRIER= WN CARRIER= EV YEAR R-squared Mean dependent var Adjusted R-squared S.D. dependent var S.E. of regression Akaike info criterion Sum squared resid Schwarz criterion Log likelihood Hannan-Quinn criter F-statistic Durbin-Watson stat Prob(F-statistic) ENAC Modeling Airline Fares 31 January / 16

11 Model Performance Predicted fare ENAC Modeling Airline Fares 31 January / 16

12 Including Southwest Dependent Variable: ln(avgweightedfare) Method: Least Squares Sample (adjusted): Included observations: after adjustments Variable Coefficient Std. Error t-statistic Prob. C AVERAGEROUTEPOPULATION/ TOTALPAX/ (TOTALPAX/1000)ˆ DISTANCE/ (DISTANCE/1000)ˆ H-INDEX/ CLASSRATIO CLASSRATIOˆ FREQAIRPORT= DIRECT CARRIER= AS CARRIER= FL CARRIER= AA CARRIER= DL CARRIER= NW CARRIER= UA CARRIER= US CARRIER= 9E CARRIER= B CARRIER= WN CARRIER= EV YEAR R-squared Mean dependent var Adjusted R-squared S.D. dependent var S.E. of regression Akaike info criterion Sum squared resid Schwarz criterion Log likelihood Hannan-Quinn criter F-statistic Durbin-Watson stat Prob(F-statistic) ENAC Modeling Airline Fares 31 January / 16

13 Model Performance with Southwest Predicted fare ENAC Modeling Airline Fares 31 January / 16

14 Forecasting Southwest enters a new route! LAS - ORD Las Vegas - Chicago Population: 4,296,645 B : 140 PAX, 2 week, 13 weeks, 3,640 (364) Currently 4 carriers Distance: 1600NM 90% restricted class tickets Direct flight LAS is a frequent airport ENAC Modeling Airline Fares 31 January / 16

15 A Reality Check (Booking LAS - ORD) ENAC Modeling Airline Fares 31 January / 16

16 Conclusion We... Processed 122 GB of DB1B data with Python Constructed an econometric model with 22 variables Were able to capture 55% of the observed fare variation ENAC Modeling Airline Fares 31 January / 16