let the figures talk
DESCRIPTION
airBaltic forecasting 2013, based on period of 2009-2012. many scenarios are addressed, as best , trend, worse, and Final, the accuracy of these models are measure by developing Forecasting Accuracy Matrix.TRANSCRIPT
Let The
Figures Talk !!!
Date of Issue: 22 July 2013
Let The
Figures Talk !!!
Setting Goals and Targets (2013)
airBaltic On Spot By :
Mohammed Salem Awad
Introduction:
airBaltic is the Latvian national carrier (99.8% state ownership) and a hybrid-LCC. It is the first national flag carrier airline to market itself as an LCC. The carrier is based in the Latvian capital Riga and also operates bases in the other Baltic capitals of Vilnius (Lithuania) and Tallinn (Estonia). The airline has also moved away from offering mostly point-to-point services in the Baltic region, and now pursues a network strategy, with Riga International the main hub. Until Jan-2009, airBaltic was 47.2% owned by Scandinavian flag SAS, which airBaltic Corporation purchased. airBaltic maintains close links with SAS, operating frequent services to the latter's hubs in Copenhagen, Oslo and Stockholm thereby operating a ‘dual hub’ system with Riga. The airline operates scheduled services to destinations in Europe, the Middle East and the CIS. (Source: Airline Profile – CAPA)
Objective :
To forecast 2013 and define the seasonality model of airBaltic in spite of the negative results of
2012 financial year.
We will use the concept of Forecasting by Objectives to develop a fair decision, so forecasting
by objective ; can be either by
- Maximizing R
- Setting Signal Tracking R. T. (36 ) to Zero
- Defining the Max/Min R. T. in the control band.
- Targeting the final results of the annual long term forecast.
- Reflecting the impact of the most recent monthly data.
Basic Data
Two set of data are collected as shown in the figures
- 4 years on monthly bases ( 2009-2012)
- 8 years annul data ( 2005-2012)
Current Situation :
The situation is hard as the airline lose a large share in
the market for 2012 compare with others last years of
2011, and 2010.
Scenario Solutions:
Two scenarios we will practice,
1- Challenge Target Scenario
2- Fair Target Scenario
1- Challenge Target Scenario
Based on 8 years annul data, we forecast and get for
2013 the expected passengers = 3,420,629 , and by using
the concept of forecasting by objective “Targeting the
final result of the annual long term forecast” we can
move further to develop the corresponding seasonality
model that reflects the seasonality index of airBaltic that
deliver in final total sum of 2013 to = 3,420,629 pax. As
shown in the figure.
The Signal Tracking (S. T.) analysis shows the
maximum value (S. T.)27 = - 13.17 While the final one
of (S. T.)36 = - 5.96
2- Fair Target Scenario
It is the outcome of two another scenarios, which act
into in two directions
a- Based on 2009 – 2011 period
b- Based on 2010-2012 period
a- Scenario based on 2009 – 2011 period:
By referring to the actual data of this period, we note
that 2010-2011 are more likely to be stable and have a
seasonality pattern so by applying the concept of
forecasting by objective - Reflecting the impact of the
most recent monthly data (Visual Adjusting) and
the grey area shows the gap between the actual and
the forecasted
b- Scenario Based on 2010-2012 period:
While in this part we try to evaluate the negative
impact of 2012 on 2010 and 2011, which a has negative
trend and that can be explore by forecasting by
objective - Defining the Max/Min R. T. in the control
band.
Best Vs Worse Forecast:
Two scenarios placed on graph to get comparison and to
see the gap difference which represent by Best Vs Worse
graphs, the idea behind that is to get a midpoint of these
scenarios in term of Average Resulting Model.
Average Resulting Model
The outcome of the two scenarios represents by the
Average Model, the graph display the actual data Vs the
forecasting one ( Resulting ), and the brown line
represents 2013 forecasting figures.
The benefit of the average model that merge the impact
of years 2010, 2011, and 2012. And can be represent
further by a mathematical model to utilize its parameters
in the analysis.
Final Words
Mathematical Forecasting Model:
By developing Average forecasting model, we get a series numbers that can be reloaded again
in the program to define its Mathematical Model and Signal Tracking Analysis, so that we get
adjusted final forecasted figure = 3,289,398 pax , R2 = 99 , S. T. = ± 7.84
Resulting Table:
Forecasting Accuracy:
Two main parameters can define the accuracy of the
above mathematical models, - Coefficient of
Determination R2 , and Signal Tracking, and the bond
region are as follows:
R2 ≥ 80 % While
Signal Tracking ( S. T.) 4 ≥ ( S. T.) ≥ −4
the forecasting accuracy matrix shows most of reading
lay on “Misleading” area, which mean these analysis
may subjected to displacement issue, but the visual look
deny this statement which qualify them to be a fair
analysis, while only worse scenario lay in fair area.
The average number cannot be display in these graphs
as it not a mathematical model.
Summary:
A four years (2009-2012) basic data of airBaltic passengers are analysis, in two direction, to
develop a Fair Target and Challenge Target. So based on that many scenarios are addressed to
reflect best case and worse case with the corresponding time span.
The mathematical models are develop and the results are tablet, exploring the final forecasting
figure that may represents the actual traffic data. While forecasting matrix accuracy is create to
define most of forecasting models impacts in analysis.
The final forecasted traffic figures for 2013 of airBaltic may listed by the following table
Appendix:
Data Base of airBaltic:
Contact:
Mohammed Salem Awad
Consultant
Tel: 00967736255814
Email: [email protected]