BUSINESS MODEL MANAGEMENT OF LOW-COST CARRIERS: IN A SEARCH FOR THE IMPACT- FACTORS OF PERFORMANCE (CASE OF AIRASIA GROUP AIRLINES)

The study is aimed at analysing how social and economic development indicators, global and regional economic indices are influencing low-cost carriers (LCC), AirAsia Group Berhad (AAGB) in particular. It is crucial not only to define the impact-factors, but to embed them in management framework for further decision-making. Passenger traffic is the main indicator of LCC performance, unlike the Full-Service Network Carriers (FNSC) that taking advantages on both passengers and freights. However, both categories depend on the macroenvironment and business environment dynamics, and KPIs should be reconsidered to face the current global challenges. The global GDP, GDP per capita are commonly used to access the economic and social development trends, the passenger numbers per annum, unemployment rate and else are used to understand the status of operations in LLC performance management. This study deals with several overlapped categories of research, such as low-cost carriers business model, impact-factors of air transport development, global trends in several industries. The research methodology is combination of comparative analysis, correlation analysis, regression analysis and forecasting, using secondary data from annual reports and quaternary financial reports. The comparative analysis gave us the understanding of the general performance trend of the group and subsidiaries. One of the study components is the correlation analysis that revealed the most correlated factors for the economic development of AAGB, such as global GDP, regional GDP, regional GDP per capita, population growth. The global and regional dimensions were presented in the research to reveal what affects the airline performance the most. Global GDP is the most correlated indicator for the global and regional development within AAGB, and the regional GDP per capita comes the second by its significance. The population size has a great influence on performance indicators (globally and regionally), and if this indicator is taken into account for forecasting the potential growth is expected in the next five years. These findings enable to design the business-model of LLC more accurate in accordance to the forecast analysis towards innovative costs decisions.

average route distance has been extended, in addition of frequencies reduction. Rely on this, its fare to state that LCCs are no-longer conducting the original concept by carry-out pure Short-Haul operation. Furthermore, the constraint from traffic density, result in LCCs adopting various method to ensure an organic growth for the future. Barrett (2004), LCCs eliminated these hobbling costs by serving a point-to-point system, often operated from secondary airports at low fares. The increasing use of low-cost carrier services reflects customer preference for low fares and point-to-point travel, and a dislike of the congestion, confusion and long walking and waiting times at hub airports.
Dewitt & Zuidberg (2012), LCCs have been able to cope with route density problems by generating new demand, as well as by attracting. Moreover, due to the colossal growth of LCC, some companies attempt to solve with the upcoming route density problem by adjusting route frequency, by selecting city pair markets with less price-sensitive demand, and by generating ancillary revenue. Scholar stated the LCC business model analyses usually focus on cost reductions and cost structure. Methods to generating additional incomes are; unbundling the traditional; all-inclusive airline product and offering unbundled low air fares, airlines can attract price-sensitive passengers to their secondary airport from even greater distances away. And they can also more effectively compete against the network carriers for higher yield passengers at the major airports. Francis et al. (2007) deduce that the relative cost advantage of a long-haul no-frills operation is around 20%, on the other hand, the advantage of short haul is above 50% or more.
Impact-Factors of Air Transport Development. Many studies have been conducted to analysis the correlation between the economic indicators and performance indices of air transportation. Previous studies have typically depended on static Ordinary Least Squares model, (Ba-Fail et al. 2000, Chin et al. 2001, Peterson 2007, Karlaftis 2008, Ba-Fail et al. (2000) analysed and forecasted the air travel market is a part of cooperate planning process in the UK and Saudi Arabia. The development model for domestic air traffic demand in the UK and Saudi Arabia was made with different combinations of explanatory variables using stepwise regression analysis. Chin, Tay (2001) stated implication on investment decisions to profitability of Asian airlines. Likewise, scholars analysed their correlation between Air traffic growth with GDP growth in Asia Pacific region, investigated the region growth and profitability of the same region and identified how marketing strategy can be adopted to improve profitability in Asian airlines.
The correlation analysis was used to examine the investment patterns, regression metrices was analysis level of impact on passenger traffic growth to airline's profitability. This study has also shown that the Asian airlines generally obtain higher profit yet to have higher survival profitability, but the airlines shouldn't be aiming for over-necessary expansion of fleet in a hurry.
Peterson (2007) sought to provide insights into the relationship between economic growth and air freight industry.
Karlaftis (2008) investigated the general issue of forecasting demand in high season in regional airports in Greece. Scholar proposed a modelling combination, dynamic Tobit models using monthly passengers traffic data for a 20-year period.
Chang & Chang (2009), investigated the relationship between air cargo transportation and economic growth in Taiwan, from 1974 to 2006. The outcome of this article is the evidence of direct correlation between economic and air cargo expansion. Marazzo, Scherre & Fernandes (2010) investigated the nexus between Gross Domestic Product (GDP) and air passenger traffic in Brazil between 1966Brazil between -2006 In this article, the econometric method 357 http://mmi.fem.sumdu.edu.ua/en proposed by Johansen and Juselius (1990) has been adopted for analysis of long-term equilibrium between variables, such as GDP and air transport traffic. Scholars investigated co-integration, reaction to shocks and causality relationship between variables. They proposed an econometric approach which can be implemented as an input for air passenger traffic forecasting model comprising time series analysis in the air transportation industry. Likewise, they also presented the empirical evidence of air transport multiplier effect in the economy using Granger test and variance decomposition on cycle component. Chi & Baek (2013) has conducted a serial test to examine the effects in both long-and short-run of the economic growth on trafficking passengers and cargos in air within the US. By utilizing the autoregressive distributed lag (ARDL) to approach to cointegration what was invented by Pesaran et al. (2001). Before ARDL numerous studies adopted two types of methods, the Granger causality and impulse response analysis. However, both instruments were mainly focusing on the dynamics in short-term instead of figuring the long-term equilibrium. According to Hu et al. (2015), the economic growth and domestic air passenger traffic in China has shown a long-term equilibrium correlation. Scholars proven that each percentage raised in air passenger traffic has direct portion to have an increasing 0.934% of GDP in China. The impact of GDP, airfare, service level, travel time and population growth to passenger traffic demand in Singapore has been stated by Phyoe et al. (2016) Impact of the Social or Economic Development to Aviation. This part of the study regards to the impact on either social or economic development in aviation, to the belonging country, to the region or the world. Whitelegg (2014) supported the concept by stating aviation industry is a fast-growing sector of the economic what associate with numbers of social and economic benefits. However, with certain growth of economy will result in direct proportional environmental impact, due to the greenhouse gases deposition and else. Nonetheless, the aggressive expansion on either air traffic or airports. Scholar mentioned with such inclining requirement have frequently supported by demand of employees to fulfil the growth. According to the study, the correlation between jobs and economic, has proved that raising employment rate can be used as one of the social development's indication, in accordantly, with the unemployment rate can also be an indication on same category. After considering the important factors, researcher believed will have reduction in air travel demand will have positive economic benefits. Dhingra & Yadav (2018) attempted to summarise the development in LCC affecting the economic, also mentioned the future. Moreover, with the revenue that influence the economic development has showed to generate better competition with the open skies policy, subsidy, and ownership issues are explicitly addressed (Forsyth, King, & Rodolfo, 2006). Major inference from the current studies shows that there is significant correlation among economic development & LCC operations (Donzelli, 2010).
Methodology and research methods. Research methods combine several dimensions: comparative analysis, correlation analysis, regression analysis and forecasting of the outcomes. There are some data limitations that are discussed below. The data processing was done via SPSS Statistics. Oum & Yu (1995) have demonstrate their method of conducting comparative analysis on 23 major airlines around the world, utilising data from 1986 to 1993, to acquire their productivity and unit cost index (UCI). Moreover, with the effect of input price changes on unit cost, gross total factor productivity (TFP) are also examined. Their result has revealed multiple discovers, what includes: (a) major European and some Asian carrier achieved tremendous productivity over North American counterparts; (b) the productivity gap between North American and other carriers have significant alleviate, etc. Furthermore, scholars have provided figures of Asia-Pacific, Europe and North American's performance of index over time on each major airline. It's clear that airlines like ANA and SAS have outstanding performance that receive near 2 UCI, however, all North American airlines has yet to achieve 1.5 UCI. In regional average, European airlines has the most stable outcome, on the other hand Asia-Pacific region has dramatically veered to have the lease performance on Singapore Airline overall in UCI.
Ahmed & Khan (2011) studied the effectiveness of productivity on economic growth of Pakistan Airlines (PIA), Singapore Airlines (SIA) and Sri Lanka Airlines (ALK) over 1995-2009 by conducting comparative analysis. Scholar acquired the result to prove that the best performed in average employee productivity and stage length. ALK managed to have best unit cost and PIA required additional improvement on all three fields compare to others. This study utilised the descriptive methods to state various concepts related to airline productivity. Calculations of dividends on total operating revenue and aggregate traffic (RPK) resulted in seats sold times average stage length. Moreover, by dividing total operating expenses and aggregate productivity (ASK) will equal to seats produced multiple average stage length. Also, study has supported the calculation by partition productivity (ASK/ATK) on numbers of employees to acquired average employee productivity. As a result, specific recommendations are suggested for each airline in this study, yet to provide finding to support scholar's opinion. Khakpour (2013) stated the numbers of comparative research approaches, such as historical, social and methodological analysis, etc. providing strengths and weaknesses on each method. SWOT analysis, PEST analysis and Porter's Five Forces has been conducted in Rahman, Azad & Mostari (2015). According to the findings, Scholars stated that Bimen Bangladesh Airlines with the situation of lack of resources, aircrafts and qualified personnel, it has successfully partitioning a respectable portion of market share in UK's highly compatible air travel industry in comparison to other major airliners. Rhman, Azd & Mostari (2015) used quantitative research and qualitative research to provide more flexibility and versatility for this case study. In addition, by performing surveys to Bimen's passengers to acquired primary data, also support by secondary data from books, newspapers, internet and magazines.
According to Mukaka (2012), correlation is a statistical method to assess a possible linear association between two or more continuous variables. Correlation coefficient is main matrix which represents the strength of the putative linear association between the variables. Regularly, values will remain within the range between -1 to +1, moreover, scenario in result with 0 reveals to nil linear relationship between two variables existed. Two main correlation analysis are introduced in Mukaka (2012), Spearsman's and Pearson's. However, Kendall's tau-b correlation coefficient hasn't be mentioned in the study. Pearson correlation coefficient uses when both variables being studied are normally distributed.
According to Taylor (1990), the correlation coefficient is referred to Pearson's r coefficient. The positive correlation coefficient means an increase in the first variable would corresponding to an increase in the second variable. On the other hand, the negative correlation coefficient indicates inverse relationship between two variables. Xiao et al (2016) presented three well-known correlation coefficient such as Pearson's, Spearman's and Kendall's, are compared from definition to application domain. Bishara & Hittner (2012) used twelve tests to identify the level of the significance of a correlation. Which includes Pearson t-test, z-test of Fisher r-to-z transformation, Spearman rank order correlation ttest, exact test, transformation box-cox, Yeo-Johnson, Arcsine and RIN, Resampling Permutation test, Univariate bootstrap test, Bivariate bootstrap test and Bivariate bootstrap with bias correlation and acceleration (BCa) test. Asuero, Sayago &Gonzalez (2006) has presented the relationship between the coefficient of determination and multiple correlation coefficient, the covariance, the correlation coefficient and the coefficient of alienation, fir the case of two related variables x and y. In addition, they explained the scatter plots and scatter diagrams. Explanation on coefficient of determination R 2 . R 2 is equal to the ratio of the sum of squares accounted for by the regression (S SReg) to the total sum of quares of deviation about the mean (SY Y) for a model with constant term (homoscedastic case, wi = 1). Goodwin & Leech (2006) discussed six factors that affect the size of a correlation, including the amount of variability in the data, differences in the shapes of the 2 distributions, lack of linearity, the presence of 1 or more "outliners", characteristics of the sample and measurement error, also mentioned the ways to determine possibility of these factors affecting the correlation. 359 http://mmi.fem.sumdu.edu.ua/en Puth, Neuhausser & Ruxton (2014) introduced the effectively using Pearson's coefficient correlation. And for this purpose, they presented several different methods. According to their results, all investigated methods performed similarly and there is no preferred or most recommended method overall. Ghobbar & Friend (2001) explained the method called Analysis of Variance (ANOVA) in regard to defining multiple attributable experimental factor and its interaction to the variation. As a reference to the result, scholars presented a summary of the General Linear Model (GLM) that used to calculate the pvalue for the main factors and both side's interaction. GLM is also demonstrated in the study, description point out the GLM is a candid extension of diverse regression allowing for multiple sole variables; that is, exploiting the relationship among dependent variable and independent variable to estimate or construe dependent variable's manifestation. However, due to multiple factors have to be researched and analyse in this case study, therefore, multiple regressions will be conducted to examine the figures to support the result. Thirteen different forecast methods regularly used by airlines were illustrated in Ghobbar & Friend (2001) Jifri et al. (2017), dependent variable is sampled as load demand, etc. For independent variable are labelled with cases such as temperature, rainfall, population and GDPs.
Data Limitation. This study is performed by gathering the secondary data from AirAsia Group Airline's annual reports and quaternary financial report. However, some of the subsidiaries like AirAsia X and AirAsia Indonesia only have data recorded since 2012. Also, AirAsia India as the latest established organization that opened in 2014, numbers has obtained ever since then. Furthermore, the rest subsidiaries have not released the financial status before 2009. World Bank has listed multiple indicators from various sector. However, this study was based on the social and economic development that affect/influence on AAGB or conversely. Which includes adolescent fertility rate, children in employment, vulnerable employment rate and unemployment rate, etc. for social development. Furthermore, economy indicators include GDP, GDP growth, GDP per capita, inflation rate, import and export of goods, etc. However, research will be required to compare the result of forecasting between official released and own calculation. World bank as the main data provider is to offer actual volume for both types of indicators. However, to complete the following study, forecasted economy data from International Monetary Fund (IMF) has been applied. Consider to different organization's definition on geographical boundary in Asia and Asia Pacific region may varies. Therefore, the results may stand on dissimilar stage, but in order to achieve our research goal, minor inaccuracy with the digits and result must be considered.
Moreover, prices on fuel play a crucial role in global economy. The aviation industry is fully dependent on the petroleum. Nonetheless, the unstable fluctuation in fuel prices is uncontrollable to public, therefore exclusion in the fuel prices and the following exchange rate to USD$ is adopted. Due to official trade between Organization of Petroleum Exporting Countries (OPEC) and else were united to exploit USD$ as commerce currency. Hence, both significant elements will be omitted.

Results
. The results will be divided into four parts as it was presented at methodology part: comparative analysis, correlation analysis, regression analysis and forecast analysis. Results on Comparative Analysis. In this part or research, we will be conducting the comparative analysis on ASK (productivity), RPK (  Due to lack of data as mentioned in the data limitation regards to insufficient sources for the early days, hence, we can clearly identify the pie chart illustrate below has yet to have any obvious changes, however, most of the subsidiaries has a gradual reduction, Indonesia AirAsia especially. Furthermore, while others have had reduced their ASK in percentage, AirAsia X has evidentially increased. However, with the reduction in percentage, each subsidiary provides higher ASK than before, except Indonesia. This situation is in addition to support the smaller company such as India and Philippian have been expanding with an unforeseeable speed to acquire higher share. Secondly, comparison to ASK and RPK is showing there has yet to have difference level of over two percent. In order to satisfy these figures without causing significant difference can briefly estimate the load factor in these air carriers are relatively high.   As estimated the relatively high load factor is supported by the ASK/RPK pie chart. Without such load factor at each company will have been documented with great variation in ASK/RPK. The highest load factor has been recorded from 2017 with achievement or near 90% of load factor done by AirAsia Malaysia (see the Fig. 4). On the other hand, AirAsia Indonesia's performance is relatively lower than others with rough average of 78%. However, it's fair to say AirAsia Group Berhad has excellent performance over last 10 years. Results on Correlation Analysis. In this part of study, we investigated how social-economic development indicators affect the passenger traffic in AirAsia Group Berhad by conducting the Pearson's correlation analysis. Global GDP, Global GDP per capita, Global Inflation Rate, Import and Export of Goods were selected to conduct our analysis as the economic development factor. Furthermore, Global Population, International Tourism and Global Unemployment Rate were categories as social development factor. According to table 1, result for over 0.9 shows very high correlation between the variables. For global population has a correlation index of 0.919, which represent the highest correlation overall. In addition, Global GDP and International tourism also have fairly high correlation of near 0.9.  On the other hand, Global inflation and Global unemployment have negative effect to passenger traffic in AAGB between each.

Table 2. Pearson's correlation analysis between passenger traffic in AAGB and Asian Pacific
Region's (APR) socio-economic indicators. Following the interpretation mentioned in the last chapter, for AirAsia's region result GDP, GDP per capita and population have a very high correlation with passenger traffic in AAGB. International tourism is also having high correlation to group's passenger traffic. Again, inflation and unemployment rate in the region shows moderate and low negative effect to AAGB.
After stating the correlation between passenger traffic to the world's socio-economic indicators then shrinking down to regional. This section will be revealing how will passenger traffic perform with airline's performance indicator.
Numbers of passenger traffic directly affect airline performance. However, in reference to table 3, none of the performance indicator fulfilled very high correlation between each. Nonetheless, four out of six of the variables selected has scored high positive correlation that between 0.7-0.9. Interpretation shows the passenger traffic has great correlation with most of the indicators, that proposed the numbers in passenger traffic will affect direct proportion to these indicators. Results on Regression Analysis. By following the correlation analysis' result, we have received a great outcome, therefore, designation on the dependent factor (Passenger) and independent factors (Global Population, Regional GDP, Regional GDP per capita and Regional Population) through a regression test to seek for how it falls within the plot (see table 4).  Results on Forecast analysis. By following the regression analysis's result, we would be able to utilize the same model to forecast the upcoming outcome in the close future. However, to obtain the answers for this research we would only be using global population as sole factor. Due to the research were set to mainly focus on the passenger traffic in AAGB. Also, in addition to acquire the forecasted data of regional GDP and GDP per capita, sources from International Monetary Fund (IMF) are used. Due to inconsistent data sources (World Bank & IMF) and regional coverage in countries varies, consequences of inaccuracy According to the trend, AAGB has a constant growing trend of around 10-20% since 2007. However, year 2007 to 2008 and 2009 to 2010, AAGB achieved the most significant result in the last decade, which have doubled and nearly doubled the total passengers from the previous year.
Conclusions. The performance of LLC could be affected by various factors, starting from macroenvironment and micro-environment to internal causes. Every variable influences the rest causing a butterfly effect to elicit to different results. Firstly, we've discovered the most correlated indicator for both global and regional to economic development within AAGB. According to table 1, global GDP reveals the most signification correlation to group passenger with 0.881. Yet, regional GDP, regional GDP per capita have shown very high correlations to AAGB's passenger traffic.
Secondly, population have accounted the greatest influence on passenger traffic in AAGB for both globally and regionally. In global correlation analysis, population even performed higher outcome as a social economic indicator than another economic indicator.
Thirdly, as we stated in the forecast analysis' section, global and countries' population is the main factor to forecast the passenger traffic in AAGB. Lastly, it is expected to have a minimum of at least 5% growth in total passengers in the next five years till 2023.
However, both global and Asia's passenger forecasting study performed by IATA shows the potential growth of 3.9% each year. In comparison to IATA's result, we have acquired additional outcome and sanguinely anticipate AAGB not only fulfil but also carry out their operations and performed better than predicted. The findings of current research can be used in the performance management, for instance, if there is the growth of traffic is expected, the logistic should be reconsidered, maintenance costs and other significant parameters of business should be taken into account in further decision-making. The innovations can take place if the R&D costs will be increased due to resources allocation. All these decisions are manageable under the accurate performance evaluation as it was performed. The focus of low-costs business model is perceived only as cost-oriented decisions system, but at the same time the costs reduction tends to be innovative by their nature. Innovativeness of costs decisions in management of low-cost may be a further direction of research.
Author contributions: conceptualization, methodology, supervision, writing-original draft I. H.; formal analysis, investigation, software, writing-original draft W. C-Y.; writing-review and editing, writing-second draft H. S.
Funding: This research received no external funding