PASSENGER FORECASTING

Authors:

Burcu Söğüt , Gizem Koz , Hilal Sergili , Büşra Kayhan

This paper will analyse and examine the forecasted passenger numbers of Emirates.  Forecasting plays a pivotal role in the airline operation management as a time perishable service industry. Forecasting passenger numbers is important for planning and controlling to the future. Forecasted passenger numbers can be used for some decisions such as pricing, new routes, and also aircraft purchase. There are many different ways to forecast passenger numbers of an airline; to exemplify, using past data as a quantitative method and surveys as a qualitative method are some ways for forecasting. In this report, it is used quantitative method with linear and exponential models, in order to calculate passenger numbers. There are also independent variables such as population, GDP, propensity to travel, educational levels, supply factors and so on.

For the first chart, linear model is used to calculate 2015 passenger number. After drawing the chart with the help of trend line, y=3366,9x+4867,5 is found as a formula. When it is replaced the necessary number with x, 45.270 is found as passenger number of 2015. The process until adding a trendline is same with the linear model for exponential model. This time it is found that y = 9582,7e0,1438x and then 51.416 is found. The real passanger number in 2015 is 49.292 due to all numbers calculated by the date of March of every year. Therefore, the result of exponential model is closer to the real data in comparison to linear model.


 In order to compare linear and exponential model, also R Squared formula can be used. This formula represents how real points are far from the trend line. Due to some distances are positive and some of them are negative, it make them all squared. R Square is always between 0-1 and bigger number is good for appropriateness of chart type. 
When it is calculated for linear function chart, R Square is 0,9823; by contrast with, R Square is 0,9923 for exponential function chart. In the light of this interpretation, it can be said that exponential function chart’s forecast result is closer to real result than linear one. Making assumptions for future, linear and exponential models can be used for different goals. For example, if there is regular increase over years, linear one might be more appropriate. However; like passenger numbers, if there is accelerating increase on numbers, exponential growth might be suitable for interpreting. On the other hand, exponential growth might not be useful in the long run due to it increases incrementally. Both exponential and linear models have advantages and disadvantages.

LINEAR

EXPONENTIAL

 Investing in the aviation is very costly. There are many questions come up with the customer. To answer these questions the company experts will make such Forecasts.
Critical questions were:
·         Are orders at another peak?
·         Should we be adding more capacity?
·         When will be the best time to introduce the next generation of aircraft?

         As illustrated in following table, the aviation industry have some charecteristic subjects which are always important to be able to forecast.
External factors, denoted by the ovals, it can influence the Dynamics such as:
·         GDP
·         Population
·         fuel and other prices
·         Interest rates
·         Target fleet utilization of the airlines. 

The decisions that the aviation industry made are based on forecasts. This should allow managers to make better decisions and can be make further enhance business performance. The 2 forecasts that we have made in this report  let us understanding the effects of forecasting and has taught us to structured work.

SOURCES

System Dynamics