Get Complete Project Material File(s) Now! »

**Methodology and Method**

*This chapter regards both the chosen method and approached of the methodology and method. In the first section the chosen research approach, design and philosophy are introduced. In the second section, the thesis data collection and analysis of the data collected are presented. Finally the credibility of the data will be discussed.*

**Methodology**

This paper is based on a quantitative design by gathering secondary data and analyzing the results by looking at statistical correlations between the chosen variables. Quantitative research is in a large sense based on measurement of amount (Kothari, 2004). This research is dealing with correlations between different variables that are readily available through secondary data gathering and then measured in SPSS, which makes the quantitative design most suitable given that *factual* empirical evidence is going to be found. The research is going to be objectively since the aim is to provide empirical evidence of how phenomena are ought to be, in line with the descriptive approach. This will be done using the statistics program SPSS which is the most used program when it comes to data analysis (Bryman, 2012).

When deciding upon what research design to use, it is important to be aware of and understand the philosophical commitments, due to its influence on the research process and outcome that is conducted by the authors. Different philosophies can be used depending on the research design chosen, which could be qualitative, quantitative or mixed research. In this case, positivism was the chosen research philosophy given the quantitative nature of the research. Since positivism is chosen to be reflected upon, the researchers will look from the standpoint of the traditional scientist. This means that the researchers will look at factual information gained through the research conducted. Furthermore, when choosing positivism the researchers decide to put human interests and feelings aside and mainly focus on the facts that are in front of them, to draw conclusions (Saunders, Lewis, Thornhill, 2009).

By using a larger sample sizes (in this case longer period of observation in regards to stock s and oil volatility) the more accurate will the SPSS correlation become. The research is based upon non-probability sampling where certain units of input are preferred (Bryman, 2012). The reason for the chosen approach is because of the subject chosen for the research. Since the thesis will examine the correlation between (1) oil price volatility and airline stock s and (2) the correlation between hedging percentages and stock s, a non-probability approach is suitable since the units examined need to be chosen beforehand to find the most appropriate data. It has been necessary to distinguish those airlines that are listed in the stock exchange and those who are not in order to collect the stock s. It is not possible to gather stock s for non-listed airlines so therefore these were not relevant for the research. Furthermore when choosing airlines to base the research on.

It can also be said that the research is purposive because it is conducted in a strategic way so that data gathering is relevant to the research questions (Bryman, 2012). This type of research approach distinguishes itself from convenience sample, which focus on the ease of gathering information (Bryman, 2012).

In line with the non-probability sampling is the self-selection of the samples used in the thesis. There are certain variables (stock s, oil prices and hedging percentages) that have to be gathered in order to accomplish the work. By using self-selection the sampling method determents which units of input that can be used in the research and which cannot (Saunders *et al*., 2009). From investigating annual financial reports of the sample as well as exchanging emails the data has been collected from the respondents as well as databases.

The research approach used is a deductive approach where the hypotheses are stated based on previous theories and where the observation made in this research can determine if hypotheses should be confirmed or rejected (Creswell, 2014). The theories that the research is based are established theories in the discipline of finance, which states that (1) commodity hedging does minimize cash flow volatility and (2) cash flow volatility has an implication for risk exposure. The aim of this research is therefore to fill a gap by hypothesize that there is a correlation between oil price volatility and stock s as well as hedging have an impact on stock s. By using scientific data this research aims to confirm or reject these hypotheses. From using pre-determined research questions and hypotheses, the whole study will gain a clear guideline of which direction to work in.

In a descriptive research an adequate method for measurements has to be clearly defined as well as a clear-cut definition of the population that will be studied (Kothari, 2014). This research is therefore further based on a descriptive approach because variables gathered (stock s, hedging percentages and oil price volatility) were predefined from the beginning and so was the method of measurements when using SPSS to correlate these variables as a mean to describe characteristics of the whole population, in this case the European aviation market. In this study, the aim was to gather data from the 50 largest airlines in Europe, which would make a good representative of the defined population (airlines in Europe). Furthermore, the study in line with a descriptive approach is the formation of close-ended questions and the predetermined analysis approach. From using such an approach, the collection of data will be fully structured so that no unnecessary data collection is done (Kothari, 2004).

**Method**

**Data Gathering**

The data gathered was secondary data and publicly accessible (both regarding Airline stock s and oil spot prices as well as airline hedging percentages). Before collecting stock values for the 50 largest European airlines, the authors checked that the airlines were listed on a stock exchange. The removed airlines are either (1) state owned (2) private owned (3) recently gone through IPO, which makes the time frame too short for analysis. There were also airlines that are a part of corporate conglomerates (such as Virgin Atlantic) as well as airlines that are part of a travel group (such as Jet2.com) that were excluded from the data gathering because we wanted to limit this research to airlines. By excluding companies where airline operation only constitute a fragment of the overall operations, the benefit is that the output provided will be more relevant. The remaining sample constituted of 16 European airlines as follows.

In the first step, we gathered quarterly stock values for these airlines over a five-year period (2010-2015) from DataStream and sectioned them by the main groups LCCs and traditional airlines. When determining LCCs we based the selection on a comprehensive list of LCCs from International Civil Aviation Organization (ICAO, 2003).

We also gathered information regarding if airlines did hedge between 2010 and 2015, and the amount hedged (as measured in percentage of the upcoming fiscal year jet fuel purchases). The data was gathered from publicly accessible financial statements from the airlines chosen as well as reaching out to airlines through personal contact to access more detailed information such as interim reports. The number of Airlines collected for this sample was further limited to seven airlines due to the fact that hedging information is not required to be included in financial statements, and was thus not always accessible. Airlines included in this sample are (1) Air France-KLM (2) Air Berlin (3) Lufthansa (4) Norwegian (5) Ryanair (6) SAS (7) Turkish Airlines.

In the second step, we gathered daily spot prices of the crude oil prices (dollar per barrel) from 2010 to 2015, the crude oil prices were gathered from Federal Reserve Economic Data^{1} from 2010-01-01 to 2015-12-31. The high frequency of daily spot prices enables us to compute quarterly volatility during this period.

**Implementation**

To be able to compute a correlation between crude oil price volatility and beta value of the same quarter we needed to access daily crude oil spot prices during these quarters (periods 2010-01-01 to 2015-12-31) in order to collect high frequency data to compute the average quarterly volatility of crude oil. The volatility was computed in three steps. First, the daily returns of the spot oil prices was computed by the commodity’s logarithmic returns based on the same method for computing logarithmic returns of tradable securities.

Where *P* is the spot price of crude oil on day t and *P**t-1* is spot price of crude oil of the observable period and *X* is the daily return. Computing the logarithmic returns is beneficial when computing returns with a higher frequency (such as daily returns over a quarter) because this method is based on continuously compounded returns and so the frequency of compounding will not affect the sum of the singe period returns (Hudson & Gregoriou, 2015).

In the second step, the standard deviation of returns (daily volatility) was computed:

Where is the daily average volatility and *N* is the number of observations in one quarter.

At the final step, we ran a bivariate correlation analysis in SPSS to see how the stock values correlates oil price volatility to determine if volatile oil prices correlates to an upward pressure on stock (positive correlation), which was hypothesized to happen. Another analysis was made to check for correlation between hedging (as measured in percentage of total jet fuel purchases) and the airline stock beta values. In cases where only the annual hedging percentages could be accessed, these were set as an average for the quarterly hedging percentages in order to achieve the same frequency as the stock values. Before accessing the output from the correlation we used predetermined cut-off levels of correlation as follows:

*0 to -0.3 and 0 to 0.3 indicates a weak correlation **-0.3 to -0.7 and 0.3 to 0.7 indicates a moderate correlation -0.7 to -1 and 0.7 to 1 indicates a strong correlation*

**Empirical Findings**

*This chapter will present the empirical findings received from the research, which will answer the research questions presented in the purpose.*

The purpose of this research was to examine if there is a correlation between the volatility in the spot prices of jet fuel (highly correlated to the spot prices of crude oil which was used in the analysis) and the stock beta movements of airlines in Europe. The purpose was also to examine if hedging strategies pursued by airlines have a correlation with stock s.

**Findings: Correlation of Stock and Oil Price Volatility**

The first hypothesis was that there should be a correlation between price volatility of crude oil and stock . The output presented here from the bivariate correlation in SPSS shows an overall weak correlation between the two variables where *37,5%* of the sampled units (*36%* of the traditional carriers and *40%* of the LCCs) showed a significant correlation. Among the traditional carriers, the ones that showed significant correlation were Aeroflot (*r*=.383, *p*=-066), Croatia Airlines (*r*=-.580, *p=.*033), Icelandair (*r=*.510, *p*=.011) and SAS (*r*=.572, *p*=.003). Among LCCs, the ones that showed a significant correlation were EasyJet (*r*=– .454, *p*=.026) and Norwegian (*r*=-.400, *p*=.053).

The output showed a moderate correlation and goes in line with the hypothesis that there actually exists a correlation between the two variables. The correlation was overall *positive* for the Traditional carriers where the correlation was significant; an exemption from this is Croatia Airlines. The correlation was *negative *for the LCCs where the correlation was significant. The differences between these two subgroups of airlines could indicate that there are underlying differences that result in different behavior of the stock in response to price volatility of crude oil. The remainder of airlines (*p>*.1) showed an insignificant output and therefore there is no correlation that can be assessed. The output from the first correlation is shown below:

**Findings: Correlation stock and Hedging Percentages**

The second hypothesis was that there should be a negative correlation between airline stock and how much airlines hedged their purchases of jet fuel. The output presented here shows an overall positive correlation between the two variables where *71*% of the sampled units showed a significant correlation. The ones that showed a moderate correlation were Air France-KLM (*r*=.572, *p*=.003) and SAS (*r*=.413, *p*=.045). The ones that showed a strong correlation were Lufthansa (*r*=.702, *p*=.000), Turkish Airlines (*r*=.681, *p*=.000) and Air Berlin (*r=*-.855, *p*=.000). Air Berlin was the only airline that showed a negative correlation between hedging percentage and stock while the remainder showed a moderate to strong positive correlation. The output presented here is not in line with the second hypothesis. The output from the second correlation is shown below:

**Analysis**

*This chapter will analyze the empirical findings of the thesis. Further, interpretations from different perspectives will be made.*

The empirical findings from our research shows that oil volatility do not have a significant impact on airline stock s overall. This has been concluded before in research by Nandha and Faff (2013) where they found that oil volatility alone do not have a large impact on stock s of airlines, it is rather the effect of oil price volatility combined with changing oil regimes as a result of triggering events such as 9/11 and the gulf war crisis that do have an impact on the stock s. Their findings could support why oil price volatility in general did not show a strong correlation to the stock s in this sample.

The low overall correlation presented could also indicate that (1) there is a weak link between oil price volatility and cash flow volatility or (2) that there is a link between oil price volatility and cash flow volatility but that other variables affect stock s more than cash flow volatility. One example of this is that Berghöfer and Lucey (2014) found that airlines who use financial hedging reduce their exchange rate risk exposure with approximately 40%. Since European airlines purchase jet fuel in US dollars it could be argued that a large fraction of the risk exposure from european airlines comes from exchange rate risk and thus could be reflected in stock s. Another variable that could be reflected in the stock s is the impact on risk exposure from operational hedging; Berghöfer and Lucey (2014) found that operational hedging have an impact on overall risk exposure and that a reduction in fleet diversity affect risk exposure. This is especially true for the European airline market where fleet diversity has been reduced by 23,12%.

Morell and Swan (2006) argue that fuel prices cannot be passed to paying customers due to the large competition in the industry. However, they also state that large airlines that also handle cargo have been able to surcharge the jet fuel prices to paying cargo customers. This could be an explanation why the airlines with large cargo handling such as Lufthansa, British Airways (part of IAG) and Air France-KLM do not have a significant correlation between stock and oil price volatility; they do not face the same risk exposure to jet fuel prices and thus the stock is not responsive to oil price volatility.

The crude oil prices (and thus jet fuel prices) increased rapidly post 2008 financial crisis when the price of crude oil dropped. The reason for the overall low correlation in this case can be connected to the findings by Aggarwal *et al*. (2012). They conclude that risks are increased mainly by oil price declines, since the oil prices have increased compared to post 2008 financial crisis, the stock in our research might not respond to the risk exposure induced by oil price volatility as much.

Altough there was a weak correlation in general among the sampled airlines, some airlines showed a moderate correlation (*SAS, Croatia Airlines, Icelandair and* *Aeroflot)*. This is in line with the first hypothesis, which could mean that the oil price volatility during the sampled period could have induced cash flow volatility. In fact, when acessing the Amadeus database^{2} , SAS and Aeroflot showed high cash flow volatilites (*192% and 130% respectively)* and Icelandair had a cash flow volatility of *24%* during this time period. If this is due to the oil price volatility, the findings are in line with Carter *et al*. (2006), whom concludes that oil prices are correlated to cash flows. The link to the underinvestment theory (Froot *et al.* 1993) and the connection to obstructed EBIT growth (Lee and Hooy, 2012) is less apparent since these airlines showed significant average EBIT growth between *2010-2015*, which would not be the case if cash flow volatility induced obstructed EBIT growth. This means that oil price volatility could explain cash flow volatility but that cash flow volatility affect other variables than EBIT growth that could have an implication for stock . Lee and Hooy (2012) found that operational leverage is one determinant for stock in the European airline industry while Jones and Kaul (1996) found that stock returns are affected when oil price volatility have an implication for cash flows due to speculative dynamics, and according to CAPM stock returns could have an implication for stock .

**Table of Contents**

**1. Introduction **

1.1 Background

1.2 Problem Statement

1.3 Purpose

1.4 Thesis Outline

**2. Literature Review **

2.1 Oil Price Volatility and Stock βs

2.2 Hedging

2.3 Summary and Hypothesis Building

**3. Methodology and Method **

3.1 Methodology

3.2 Method

**4. Empirical Findings **

4.1 Findings: Correlation of Stock β and Oil Price Volatility

4.2 Findings: Correlation stock β and Hedging Percentages

**5. Analysis **

**6. Conclusion **

**7. Discussion**

7.1 Limitations

7.2 Implications

7.3 Further Research

**8. References**

GET THE COMPLETE PROJECT

Commodity Risk Management in The Airline Industry