Implications of Committed Capital Investments on firm decisions, risk, performance, and investor compensation : a case study of the airline industry

(1)

Implications of Committed Capital Investments on Firm

Decisions, Risk, Performance, and Investor Compensation:

[A Case study of the Airline Industry]

University of Amsterdam

Amsterdam Business School

MSc Business Economics

Master Specialisation Finance

Author:

J. B. Bergwerff

Student number:

10398171

Thesis supervisor: Dr. Tomislav Ladika

Finish date:

July, 2016

(2)

NON-PLAGIARISM STATEMENT

By submitting this thesis the author declares to have written this thesis completely by himself, and not to have used sources or resources other than the ones mentioned. All sources used, quotes and citations that were literally taken from publications, or that were in close accordance with the meaning of those

publications, are indicated as such.

COPYRIGHT STATEMENT

The author has copyright of this thesis, but also acknowledges the intellectual copyright of contributions made by the thesis supervisor, which may include important research ideas and data. Author and thesis supervisor will have made clear agreements about issues such as confidentiality.

Electronic versions of the thesis are in principle available for inclusion in any EUR thesis database and repository, such as the Master Thesis Repository of the Erasmus University Rotterdam

(3)

ii

Implications of Committed Capital Investments on Firm Decisions, Risk,

Performance, and Investor Compensation: A case study of the Airline Industry

Jasper Bergwerff

Abstract

The airline industry is a capital and labor intensive industry characterized by low profitability (Dempsey, 2008). Borenstein (2011) finds that on average, the US airline industry reported losses of 59 billion US dollars from 1971 until 2009 which is accompanied with 49 bankruptcies from 2001 to 2009. These phenomena emphasize the vulnerability of the airline industry. Still, airlines succeed in obtaining the capital required for investments, which are determined years in advance, and form commitments to the firm. Firstly, this thesis investigates whether airlines with relatively more investment commitments respond differently to a crisis than less committed airlines. It is found that airlines with relatively more commitments reduce their expenditure on staff, operating expenditure, and SGA expense compared to less committed airlines in a crisis. Furthermore, the higher committed airlines see their profit margins improve, but market capitalization and liquidity reduced compared to less committed airlines. The second half of this research addresses the investment returns of airlines. In order to attract capital, investors must be compensated according to the CAPM risk-return tradeoff. The results for Jensen’s Alpha suggest that for 57.2% of the airline-year observations, investors have a negative Alpha i.e. are under compensated. Regressing the Alpha’s on company characteristics convey that Alpha’s are statistically significant and negatively related to revenue growth and total assets.

JEL classifications: G01, G12, G32

Key Words: Committed investments, airlines, crisis, bankruptcy, labor union, volatility, CAPM, Jensen’s Alpha

* I would like to thank Dr. Tomislav Ladika for his constructive comments and for providing part of the committed investment data used in this research. I would also like to thank Mazars Paardekooper Hoffman N.V. for making the S&P Capital IQ database available during my employment at the firm, in order to obtain a universe of airline beta’s.

(4)

iii

The US airline industry became deregulated after the oil crisis in the 1970’s, which is argued to have been detrimental to airline profitability (Dempsey, 2008). Dempsey states that the industry has suffered from falling prices for airline services, meanwhile coping with increasing fuel prices, strong labor union presence and external shocks like the terrorist attacks of 9/11. These factors altogether have contributed to the bankruptcies of 122 US airlines between 1978 and 2008 (Dempsey, 2008). In general, the airline industry has been struggling and experiencing this plight ever since the end of the 20th century. On average, US airlines incurred losses every other year since 1981, reporting losses of 59 billion US dollars from 1971 until 2009 (Borenstein, 2011). Dempsey (2008) acknowledges that because losses remain prevalent for all airlines, even in states of economic (market) growth, airlines are likely to have insufficient capital to endure unexpected shocks without changing its corporate policy.

The existence of small capital buffers and vulnerability of airlines to unexpected shocks raises several questions. For instance, are airlines that invest in the acquisition of aircraft, more vulnerable to unexpected shocks than airlines that have lower levels of investment and capital commitments? This paper will bridge the gap in existing literature by investigating the implications of commited investments on a company’s business decisions and responses in a crisis, in addition to adressing the implications of committed investments on investor compensation. Thus far, empirical work treats capital expenditures as yearly decisions, however in reality these are determined years in advance. To explain, airlines determine the number of aircraft they wish to purchase from manufacturers such as Boeing years in advance. In spite of the industry uncertainty, investments, such as the acquisition of aircraft, are decisions that have to be made for the long run (Schefczyk, 1993). These are large committed orders that occur in a volatile industry, largely influenced by the economic environment (Hu & Zhang, 2015). Once the investment decision is made by an airline and an agreement is reached with the aircraft manufacturer, the orders are committed. 1 _{This implies that the investments become contractual} obligations, which will result in penalties for airlines if investments are cancelled or payments aren’t met (see appendix 5 for an example).

1_{The Securities and Exchange Commission (SEC) defines an investment commitment as a purchase obligation, which}

is: “an agreement to purchase goods or services that is enforceable and legally binding on the registrant and that specifies all significant terms.” https://www.sec.gov/rules/final/33-8182.htm)

(9)

2

The airline industry is characterized to have small capital buffers, meanwhile having a large cyclical exposure and sensitivity to unexpected shocks or crises (Ali, Hamson, Inglis & Sargeant, 2013). These factors altogether provide for a laboratory environment to investigate whether the business models for high and low committed investment firms are fundamentally different. The nature of the airline industry is such that an unexpected shock will likely require the airline to change its corporate policy and business decisions (Goyal & Negi, 2014). A continuation of the policy in place before the shock will likely trigger the company into default and distress (Goyal & Negi, 2014). On top of that, the airline industry is characterized by having a presence of special-interest stakeholders (unions) and returns that continue to be low relative to other industries, despite high leverage (risk) (Pearce, 2012). By demonstrating whether the investment commitments of airlines affect corporate policy and responses to unexpected shocks, answers may arise to several questions. Firstly, an unanswered question is why there is persistent equity financing of airlines despite shared beliefs of relatively frequent defaults.2_{It is possible} that highly committed airlines, or airlines with higher levels of operating leverage obtain higher returns for their equity investors? Secondly, there is a possibility that highly committed airlines manage to renegotiate new labor contracts according to the theory of Benmelech, Bergman, and Enriquez (2012). Will airlines with high committed investments become more profitable after a crisis due to cost cuts reductions such as staffing expenditure? The investigation is also relevant for policymakers, unions and other stakeholders who continue to debate about the degree of airline industry regulation in light of declining profitability and sustainability of the industry (Emerald, 2006). Bankruptcies and financial distress in airlines can result in layoffs, (termination of employment) which carry large economic and social costs (Sullivan & Von Wachter, 2009).

Firstly, this thesis models the yearly investment and leasing commitments of 37 airlines from the period 1993-2015, which became available early 2016.3_{In order to investigate whether} the degree of investment commitments result in different responses in a crisis, this thesis uses the difference in difference regression methodology. According to the commitment data, high committed airlines are designated to the treatment group, which contains airlines with relatively high commitments in a crisis. Other airlines are allocated to the control group, which contain airlines with relatively few commitments during a crisis. With airlines allocated in either the control or the treatment group, it is possible to investigate whether the groups respond differently to a crisis. More specifically, this research investigates whether unexpected shocks lead to disparities in corporate decisions and performance (measured by employment, expenses,

2_{It is stated by Dempsey that since deregulation in 1978, until the time of the publication of his work, a total of 122}

airlines went bankrupt. Not to mention that airlines such as US airways went bankrupt multiple times (2008)

3_{The data collection of this research was performed in early 2016 when the 2015 financial reports of some airlines}

(10)

3

profitability, liquidity, investment, and returns) for airlines with high and low levels of committed investments.

It is posited that committed investments affect the sustainability of airlines and hence to begin, lead to a disparity in the employment or staffing expenditure. It is assumed that highly committed airlines will have relatively more severe financial difficulties in a crisis because the committed investments form a liability to the company, and are recognized so by the Generally Accepted Accounting Principles (GAAP).4 This implies that commitments are debt-like

obligations, which in a crisis, increases the probability of defaulting on payment obligations (Berk & DeMarzo, 2011). The characteristic of having an increased probability of distress opens up the possibility to renegotiate the terms of labor with labor unions (Benmelech et. al., 2012). It is believed that these airlines have a credible threat of going bankrupt if no reduction in staff size or staffing expenses are realized. This phenomenon may force labor unions to do concessions since bankruptcy of an airline can result in a termination of employment for all employees. Hence, this research expects highly committed airlines to have relatively less employees, as well as less staff expenditure compared to less committed airlines.

Furthermore, it is predicted that operational expenditure, as well as selling, general and administrative (SGA) expenditure, are reduced compared to less committed airlines. This hypothesis comes from a similar line of reasoning as with the staffing expense reduction predicted earlier. Since a firm’s liquidity in a crisis is typically reduced (Berk & DeMarzo, 2011), and since purchase obligations can significantly impact a firm’s liquidity, it is predicted that highly committed airlines in a crisis are more likely to experience distress. 5 This may entail that a crisis

is more stringent on highly committed airlines, forcing these airlines to cut costs where possible. This is necessary in order to free up the funds that may allow airlines to continue operations. Hereby airlines may remain capable of paying for their debt and investment commitments. Doing so, airlines may avoid any penalties that are associated with a failure to meet the necessary payments of their investment commitments.

In spite of the cost reductions, this research predicts that highly committed airlines, relative to the low committed airlines, will have worse profit margins. In earlier research it is found that distressed airlines engage in price wars, at the cost of lower margins, in order to obtain the critically needed funds (Busse, 2002). As mentioned above, high committed airlines have more debt-like obligations in a crisis, which make these airlines more likely to enter distress in a market

4 _{It is stated by Dempsey that since deregulation in 1978, until the time of the publication of his work in 2008, a total}

of 122 airlines went bankrupt. Not to mention that https://www.sec.gov/rules/final/33-8182.htm)

5_{The Securities and Exchange Commission (SEC), due to the nature of investment commitments/purchase}

obligations which may have a significant impact on a firm’s liquidity, requires that companies report their investment commitments. https://www.sec.gov/rules/final/33-8182.htm

(11)

4

downturn (Berk & DeMarzo, 2011). Following the reasoning of Busse (2002), these airlines are then more likely to engage in price wars compared to their less committed competitors. This research therefore predicts that the profit margin reductions from price war behavior outweigh the profit margin improvements from cost reductions. Given the fact that a crisis raises the difficulty of generating a positive cash flow, existing cash and funds must be used to meet payments (Ali et al., 2013). As a result, this hypothesis is based on the same argumentation as before: a crisis dries up liquidity (Berk & DeMarzo, 2011), which is augmented in airlines with more investment commitments, or purchase obligations. For this reason, this research predicts that higher committed airlines in a crisis will have lower cash buffers compared to low committed airlines. It is argued that cash buffers are used to fulfill payment obligations on the commitments. The financial difficulties and increased probability of default for high committed airlines are also the reason why this research posits high committed airlines to have relatively larger decreases in market capitalization compared to less committed airlines. Investors face additional risk in highly committed airlines as debt and investment payments are fixed while cash buffers are reduced, increasing the risk of bankruptcy. This leads to the hypothesis that the highly committed airlines will be valued less according to its market capitalization compared to other, low investment airlines that remain more flexible (Jones & Ostroy, 1984). Lastly, it is predicted that highly committed airlines will realize lower stock return and invest less than low committed airlines.

Empirically, this research finds statistical support for the notion that airlines with higher commitments behave differently from low committed airlines following an unexpected shock. More specifically, support is found for the hypothesis that airlines with high commitments in a crisis reduce staffing expenditure and lay off more staff than airlines with lower commitments. This is in accordance with the theory that airlines in distress are more likely to renegotiate labor contracts. It is argued that distressed airlines use a credible threat of bankruptcy to win favorable terms of labor as unions do concessions (Benmelech et al., 2012). Furthermore, the conjecture on reducing other expenditures is also statistically supported. Not only are highly committed airlines more likely to reduce SGA expenditure, but they are also more likely to reduce operational expenses in a crisis, relative to less committed airlines. Moreover, contrary to the original hypothesis, due to the cost cuts relative to low committed airlines, the high committed airlines also realize larger, improved profit margins in the years following the crisis. Still, this does not necessarily refute the price war argument posed above. It does suggest that expenditure cost cuts improve margins and outweigh any potential margin decreases as a result of a price war strategy. In addition, statistical evidence is found for the idea that cash holdings, or liquidity, as well as market capitalization, is relatively lower among high committed airlines than less committed airlines in a crisis, which supports my initial hypothesis. Some concerns may be that years used to define the crisis, or the methdology used to determine whether an airline is high/low

(12)

5

committed is coincidentally showing the above results. However, robustness tests convey that alterations to the years that define the crisis, as well as alterations to the scaled commited investment measure, lead to similar (statistical) conclusions.

Apart from investigating the effect of committed investments on the airline decisions and policy in a crisis, this research also establishes predictions for the investment decision in the airline industry and its investor compensation. To explicate, there is a common belief that investments in the airline industry fail to generate the required returns of investors (Pearce, 2012). In addition, it is argued that the return on capital in the airline industry is below the cost of financing for the industry (Pearce, 2013). This characteristic suggests that the airline industry is value destroying, and according to financial theory, should result in poor stock performance.6 According to financial theory however, in order for a company to obtain the funds and external financing needed to invest, it should provide investors with a risk-adjusted compensation (Berk & DeMarzo, 2011). When this prediction is tested for the sub-sample of seventeen US-only airlines from 1995 to 2012, it is found that the average realized year-on-year stock return is 7.6% while the median amounts to 0.9%. While a return of 7.6% per year may still yield the required equity return on investments, a return of 0.9% on airline stock is likely insufficient for the level of risk, as this is a return below the risk free rate. This raises the question as to whether stock returns realized when investing in these airlines reflect the risk level of the investment for shareholders. This research posits that due to the volatility and risk present in the airline industry (Gritta, Chow & Shank, 1994), and because of the fact that the industry realized losses on average (Dempsey, 2008), investors fail to receive their risk-adjusted fair compensation. Therefore, it is anticipated that average and median value for Jensen’s Alpha in the research lie below zero. In addition, it is predicted that airlines with higher operating leverage realize lower Alpha’s due to excessive leveraged market (risk) exposure in an industry that is already highly levered (Dempsey, 2008). For the same reason, it is also predicted that Alpha’s are smaller for airlines with larger debt on the balance sheet than peers with less debt. Furthermore, this research posits that because the industry is shrinking and characterized by overcapacity (Pearce, 2013), larger airlines, or airlines that attempt to realize revenue growth, will have lower Alpha’s. Still, Alpha’s are expected to be larger for airlines with a higher return on their assets, since this increases the likelihood that the returns on capital (ROC) exceed the weighted average cost of capital (WACC) of the firm and thereby creates value (Damodaran, 2008).

A beta analysis on the airline’s excess stock return with the S&P 500 excess return reveals that the average beta of the sample is 1.16. By the CAPM equation, this yields an expected and required return of 9.9% on average and in the median. The mean and median values for Jensen’s

6_{Damodaran in “Damodaran on Valuation” refers to an industry being value destroying if its ROC < WACC. Even if}

(13)

6

Alpha amount to -2.5% and -7.2%, respectively. This indicates that on average, and in the median, the industry fails to compensate their equity investors with a return that compensates them for the risk, opportunity cost of capital, and time value of money. In fact, a total of 57.2% of the observations have a Jensen’s Alpha that is less than zero, i.e. shareholders investing in airlines of this sample, are under compensated for 57.2% of the years. Still this implies that 42.8% of the observations carry a positive Jensen’s Alpha. What this conveys is that for the sample, selective stock picking in the airline industry can still generate investor compensation equal to or above the required rate of return. Regressing the Jensen Alpha’s on company characteristics convey that Alpha’s are statistically significant and robust for revenue growth. The regressions suggest that higher revenue growth leads to smaller Alpha’s. The underlying reasoning is based on the idea that revenue growth is costly in an industry that is shrinking on average. This follows from the reasoning presented in Damodaran that in an industry with a WACC above the ROC, revenue growth is value destroying for shareholders. As a result, the regressions indicate a robust and statistically significant coefficient at the 1% level. Furthermore, robustness regressions indicate that larger firms, measured by the logarithm of total assets, also obtain smaller Alpha’s. Results are robust in that they have the same sign, but not statistically significant for debt to assets or return on assets.

The remainder of this paper is set up as follows. Chapter 2 addresses the existent theories available in investment literature. Despite the fact that the committed investments have not been modeled to date, the existing literature on investments aid in the formation of hypotheses for this research. Chapter 3 provides the methodology and model used to investigate whether the degree of committed investments have implications on the business decisions of firms in a crisis. Chapter 4 continues by introducing a model on testing the risk-return compensation in the airline industry. After this, chapter 5 will mention how the data for the investigations are constructed and discuss the descriptive statistics. The results to the investigations, for which the methodology is explained in chapters 3 and 4, are provided in chapter 6. Chapter 7 contains various alternative regression analyses in order to see if the regression results discussed in chapter 6 are robust. Lastly, chapter 8 provides concluding remarks.

CHAPTER 2 – Literature Review

In order to address the existent theories about investments, the first portion of this research paper is related to a large body of literature demonstrating that an investment decision affects corporate policy and returns both directly and indirectly. However, the existing literature focusses mostly on theories of capital expenditures. This implies that there is no explicit account for investment commitments and the distinction from capital expenditure in existing research

(14)

7

papers and academic articles. In other words, the committed investments of companies have not been modeled to date. Due to the fact that purchase obligations and investment commitments are part of an executory contract that have yet to be fulfilled, these obligations and commitments are not recorded on a company balance sheet (Lee, 2010). However, as argued above in footnote 5, the SEC acknowledges that investment commitments and obligations can greatly impact a firm’s liquidity. Hence, this paper will aid in bridging the gap in the universe of existing literature by explicitly accounting for the effect of capital (investment) and leasing commitments on corporate policy in a crisis. This investigation will therefore also inform about the informational content in the level of investment commitments that a firm has. Furthermore, the second half of this research will provide an insight into the common held perception that returns in the airline industry are insufficient and fall below the required return of investors based on risk exposure (Pearce, 2012). This second portion will investigate, among other things, whether the amount of investment commitments and lease obligations (operating leverage) influences investor compensation. The fundamental idea is based on the fact that the airline industry is volatile (Hu & Zhang, 2015). This implies that investors have an increased (market) risk exposure, which will be captured by a larger CAPM beta (Berk & DeMarzo, 2011). In turn, this will translate into a higher desired rate of return for the providers of equity. Having mentioned however that the airline industry has incurred several years of losses (Dempsey, 2008), it is questionable as to whether the airline industry succeeds in creating value for its shareholders.

The insights that arise from this research can be utilized to explain certain characteristics present in the airline industry as well as in other industries. Investment commitments, or purchase obligations, are present in almost all industries (Lee, 2010). This implies not only that this research can contribute to the universe of literature in the broadest terms, but also to contribute to specific industries characterized by purchase obligations or investment commitments. Hence, even though the investment commitments are not directly addressed in the existing literature, it can provide a theoretical framework and aid in formulating hypotheses for this research. In addition, the existing research on investments demonstrate that contradicting views exist, which provide no definitive answer as to what the effects of committed investments are on corporate policy. This highlights furthermore the novelty of this thesis topic and the relevance of performing research on committed investments.

2.1 Effects of Investment on Employment Decisions

First of all, there are opposing views in existing literature as to whether investments lead to differences in staff layoffs when experiencing exogenous shocks. As mentioned earlier in a reference to Sullivan and Von Wachter (2009), unemployment carries large social costs and thus

(15)

8

policy makers and unions are interested in knowing whether committed investments affect the employment decision. Also, as mentioned above, the airline industry is characterized to have strong labor union presence that makes wage reductions, as well as staff reductions, more difficult to enforce (Benmelech et al., 2012).

The authors Atanassov and Kim (2009) suggest that in distress or difficult times, instead of laying off staff, airlines are more likely to turn to other means of sustaining business operations. More explicitly, the authors demonstrate that weaker investor protection and strong unions, a feature of the airline industry (Dempsey, 2008), makes it more likely that poorly performing firms sell assets rather than lay-off workers to free up the funds for paying for commitments. When placing this into the context of investment commitments, this would suggest that airlines with larger commitments, though perhaps in a worse financial situation, turn to other cost cutting or cash generating measures. In other words, regardless of the degree of investment commitments an airline has, airlines would still maintain similar levels of employment, even after exogenous shocks. Earlier research by Pulvino (1998) objects this idea however. Airlines typically possess assets in the form of aircraft or aircraft parts and material (Darcy, 2010). Though at first sight a fire sale of aircraft seems feasible in distress, Pulvino finds that the trading of aircraft, the key asset to airlines (Darcy, 2010), is illiquid and only about twenty aircraft are traded every month (1998). The belief of Atanassov and Kim that airlines engage in the sale of assets when they enter a difficult financial situation, is also not supported by Hallock (1998).

In his article, Hallock (1998) finds that large firms are more likely to announce staff layoffs compared to smaller firms. Although this finding does not in itself say anything about investment commitments, it can be argued that large firms have bonds and alternative means of financing their activities, which facilitate large capital purchases such as the acquisition of aircraft (Berger & Udell, 1995). This entails that larger airlines are capable of sustaining and having higher investment commitments. Because of this phenomenon, larger airlines may be considered and characterized as highly committed airlines. In the event of a shock, this would suggest that highly committed airlines, because they are possibly also the larger airlines, will lay off more staff than less committed airline. Chen, Mehrotra, Sivakumar and Yu (2001) also have researched layoffs, finding that if a firm experiences poor stock performance and earnings announcements, it is likely to lay off staff. Investment commitments can be viewed as debt like payments that increase the financial burden of airlines during a crisis. According to financial theory, in the event of a crisis or distress, because of an increased bankruptcy risk, investors sell off their shares in the companies with increased financial burden. Following the terminology presented in the research of Chen et al. (2001), these high committed airlines then experience lower stock returns, which trigger them to have more staff layoffs.

(16)

9

Gittel, Cameron, Lim and Livas (2006) demonstrate that there is a disparity in employment layoffs among airlines following the 9/11 exogenous shock to the airline industry. Some airlines do not lay off any staff whereas airlines such as United Airlines and US Airways lay off up to 25% of staff. The authors attribute these layoffs to the superior business models of airlines with low layoffs. However, as with other authors, Gittel et al. (2006) provide no account of committed investments. It is therefore still ambiguous whether the level of investment commitments is affected by the business models of airlines or whether commitments affect business models themselves. If high committed investments result in superior airline business models, there would be support for the idea based on Gittel et al., that the degree of investment commitments affects the amount of employee layoffs.

Berk and DeMarzo (2011) introduce another argument as to why some airlines, despite strong labor union presence, succeed in reducing wage expense and the number of employees. The authors build on argumentation of Gittel et al. (2006) by stating that poor performance of airlines, perhaps due to an inferior business model, can facilitate staff layoffs or wage cuts. Berk and DeMarzo (2011) explicate that airlines that are at least partially financed by debt have a credible threat of bankruptcy, which can enhance the necessity of realizing wage and staff size reductions for airlines. As a result, American Airlines succeeded in winning wage concessions after talking to its labor unions in April of 2003 (Berk & DeMarzo, 2011). In a similar way, Delta Airlines managed to persuade pilots to accept a 33% wage cute in November of 2004 (Berk & Demarzo, 2011). This argumentation is confirmed further by Benmelech, Bergman and Enriquez (2012). Benmelech et al., (2012) state that airlines are more capable to renegotiate wage contracts, and thereby reduce employment (expenditure), when their financial position is weaker. According to the authors, the threat of bankruptcy, and thereby the inability to pay for pensions of workers in the future, is a credible threat to obtain concessions on labor. As mentioned above, the increased financial burden due to having investment commitments during a crisis increases the probability of bankruptcy, potentially allowing for wage renegotiation with employees. According to this theory, airlines with larger commitments in a crisis therefore succeed in reducing wags or laying off more staff than low committed airlines. This argumentation by Berk and DeMarzo (2011), as well as Benmelech et al. (2012), form the theory behind the hypothesis on employment for this research. It is posited that highly committed airlines will experience lower wages, employees and staffing expense relative to low committed airlines in a crisis. The Null hypothesis is that the coefficient for the Difference in Difference estimator (hereinafter “DD estimator”) for the employment decision regressions is not statistically significant from zero, versus the alternative that it is statistically different from zero.

(17)

10

2.2 Effects of Investment on Profitability

Airlines have other potential means to improve profitability than to reduce staffing expense. Therefore, after having argued the potential effects of commitments on the employment decision, the existing literature is investigated in order to develop a deeper understanding of commitment effects on the business models of airlines. There are contradicting theories as to whether the degree of (investment) commitments affects airline profitability after experiencing economic shocks. In their research, Harlan and Marjorie Platt (2002) state that companies with a higher fixed asset structure are more likely to enter distress. By definition, fixed assets are objects and means that provide long-term income and are not expected to be used up within a year or converted in to cash (Marshall, McManus & Viele, 2011). Aircraft acquisitions increase the fixed asset structure in an airline since aircraft acquisitions are large purchases of assets with a long duration and life. In essence the purchase cost of these aircraft is then spread over several years and covered by future income. However, this also implies that airlines with high investment commitments have an increased probability to enter distress. Aircraft purchases are large and requiring debt financing (Hu & Zhang, 2015). In addition, the authors state that income has to be sufficient in future years to cover interest and principal debt payments (Hu & Zhang, 2015). With this in mind, Busse (2002) demonstrates that airlines in financial distress engage in price wars. Airlines with higher leverage ratios are more likely to lower prices and risk losses to obtain critically needed cash for debt, interest, and commitment payments. Since high committed investment airlines have capital tied up in investments and since interest payments are a fixed cost that must be paid regardless of revenues, these airlines are more likely to enter distress following a shock due to the inability of meeting debt obligations (Freear, 1980). This research altogether suggests that high committed investment airlines are more likely to engage in price wars, leading to lower gross margins after an exogenous shock.

Based on this literature, there is reason to believe that there are differences in how high committed versus low committed investment airlines react to unexpected shocks. On the contrary, Vafeas and Shenoy (2006) state that profitability ratios are worse for high investment firms. However, in the event of low cash flows, which is a characteristic of the airline industry, the authors find no disparity in profitability and cost base exist for firms. This would suggest that even though in general higher (committed) investments lead to less profitability in terms of their margins, it this does not apply to airlines because of the low cash flow characteristic. The effect of commitments on the profitability ratio or margin of airlines therefore remains ambiguous. Still, the theory mentioned by Busse (2002) forms the basis of the hypothesis in this research. It is predicted that highly committed airlines will have lower profit margins than less committed airlines due to a price war strategy of high committed airlines. In other words, I expect to find a

(18)

11

coefficient that is negative for the DD estimator in regressions with profit margins as the dependent variable. As is done in all other regressions, the null hypothesis is that there is no statistical significance from zero for the Difference in Difference estimator versus the alternative that there is statistical significance from zero.

2.3 Effects of Investment on Return

As a precursor to investigating why investors choose to invest in the airline industry, the existing literature is used in order to develop an understanding of airline returns which investors use to judge the quality of an investment. To begin, Pearce (2013) finds that returns on capital (ROC) remain far below the weighted average cost of capital (WACC) for the whole airline industry. This implies that airlines destroy shareholder value and poor returns exist, independent of whether an airline has high or low committed investments (Koller, Goedhart, Wessels, Copeland & McKinsey, 2005). Financial theory states that firms with a WACC exceeding the ROC of their projects are value destroying for shareholders and have poor investments. Hence these firms shouldn’t be capable of receiving financing (Koller et al., 2005). Despite this, in general airlines are capable of obtaining financing, which becomes evident when witnessing the existence of aircraft purchases despite low cash flows (Carter, Roger & Simkins, 2006).

If the decision to buy or sell shares of an airline is related to the level of committed investments the airline has, there should be a disparity in the stock performance of airlines with high and low degrees of investment commitments. Existing literature provides no conclusive answer as to whether the degree of commitments affects (stock) performances of airlines. For instance, academic research is done by Titman, Wei and Xie (2003), claiming that companies with higher capital expenditure experience lower stock returns for five years. Firms that have substantial increases in capital investments, according to Titman et al. (2003), achieve negative adjusted returns because it is unsure whether these investments will provide a return with covers the costs of financing. Due to the fact that capital expenditure in airlines is primarily devoted to the acquisitions of aircraft (Carter et al., 2006), following the same line of reasoning suggests that airlines announcing increases in levels of committed investments should experience lower returns.

An underlying reasoning for airlines experiencing poor stock returns following increases in investment is also presented by Duchin, Ozbas and Sensoy (2010). The authors find that industries and firms relying heavily on financing, or firms that are financially constrained, experience lower stock returns (Duchin et al., 2010). In the case of airlines, companies with higher levels of investment commitments are obligated to fulfill larger payments and yearly installments on the orders for aircraft, facing fines if payment obligations aren’t met. The fact that the cash

(19)

12

flows are low in the airline industry and because besides debt payments, also installments on investments become mandatory in a crisis, stocks of airlines are sold in order to correct for increased risk exposure for the airline (Duchin et al., 2010). Therefore, it is presumed that highly committed airlines experience lower stock returns during an unexpected shock. Similarly, Pearce (2013) believes stock returns are lower in a crisis for highly committed firms. Pearce (2013) supports an earlier finding of Platt et al. (2002) by stating that aircraft purchases trigger a higher fixed asset and cost structures. This phenomenon, Pearce states, is also known as operating leverage and is reflected with an increased (levered) beta of a firm. The higher the beta, the more systematic risk exposure a firm has, implying lower expected returns for firms in downturns. Since airlines with high committed investments purchase more aircraft, the returns should be lower than for low committed investment airlines after a shock (Pearce, 2013).

Still there is literature that has a different argument as to why returns are lower for some airlines while others experience more positive returns. For instance, Carter and Simkins (2004) find that the 9/11 attacks (exogenous shock) plummeted stocks of airlines with lower cash reserves more than airlines with higher cash ratios. The argumentation in Carter and Simkins (2004) is based on the fact that high cash levels reduce probability of default. The authors state that cash holdings can serve as a buffer to meet interest or principal debt payments, as well as other obligations. The authors continue by explaining that cash holdings provide the airline with liquidity in the event that operating earnings of the company is insufficient to meet its financial obligations. However, it is not clear from their research whether committed investments drive the cash holdings of a firm. To explain, high cash reserves can arise from the need to fulfill large commitments in the future. An airline with commitments may choose to hold higher cash reserves to limit the risk of bankruptcy and anticipate weaker operational performance in a crisis (Berk & DeMarzo, 2011). On the other hand, the decision to maintain relatively high cash holdings compared to other airlines can also result from expectation to place investment orders in the near future. Airlines can therefore choose to hold relatively high cash levels if a company currently has low commitments but expects to place purchase orders, raising commitments in the future (Darcy, 2010). Hence, research of Carter and Simkins (2004) provides no answer to whether performance differs between airlines with high committed and low committed investments. Lastly, in their research, McConnell and Muscarella (1985) find that announcements to decrease capital expenditure are negatively received by investors. Thus as mentioned above, increases in capital expenditure are typically not received well by investors, but also announces to decrease investment leads to lower returns. However, the authors don’t focus specifically on committed investments or capital expenditure. As a result, there is no definitive answer as to what the stock returns of an airline would be if airlines decide to cancel or reduce committed investments. It is

(20)

13

not evident whether decreases of investment commitments in a crisis are well received by investors.

This research posits that the increased obligations in a crisis for highly committed airlines, raise the burden of these airlines and increase the systematic risk exposure according to the theories of Platt et al., (2002) and Pearce (2013). This altogether may lead to lower market capitalization and stock returns for high committed airlines compared to less committed airlines. Hence I predict that the coefficient for Difference in Difference estimator to be negative for stock returns and market capitalization. This idea is tested according to the null hypothesis that the coefficient on the difference in difference estimator (DD estimator) is not statistically different from zero versus the alternative that the coefficient is statistically different from zero.

2.4 Investor Compensation

In order to allow airlines to invest, as is the case with any company, firms need to be capable of obtaining the necessary funds. With this capital, airlines can place aircraft investment commitments. Meanwhile, companies are required to compensate the providers of capital for their risk and opportunity cost of capital (Koller et al., 2005). Peter Morrel (2013) states that in order to judge the airline’s performance, it is important to investigate the return on capital employed (hereinafter “ROCE”) sometimes referred to as the return on invested capital (ROIC). According to Peter Morrel (2013), this measure indicates how successful the airline is in its long term capital management and investment. With this in mind, the international air transport association estimated that the airline industry will require capital requirements of four to five trillion US dollars for the industry to acquire new commercial aircraft between the period 2011 and 2031. Pearce (2012) however addresses his concern that because the ROCE is lower than the cost of financing, airlines may have difficulty attracting the capital necessary for investment. Hence, in order to fully address why investors, choose to finance airlines for acquisition of aircraft, it is also necessary to investigate returns in the whole airline industry. Airlines with high returns, in particular returns equal to or above the risk-adjusted demanded return of the investment, are airlines that obtain financing if markets are efficient (Koller et al., 2005). Returns on stock that are insufficient to compensate for the level of risk, by financial theory, will not be capable of obtaining equity financing (Berk & DeMarzo, 2011). Perhaps the decision to provide airlines with capital for aircraft investment commitments is dependent on the degree of committed investments. Conceavably however, the existence of returns that are perceived to fall short of the risk-adjusted returns of other industries (Pearce, 2012) may be a characteristic of the airline industry and independent of the degree of commitments. Before investigating whether the returns on capital and on stock in this experiment comply with the general belief of poor

(21)

14

investment returns, existing literature can provide a foundation and build an idea as to whether poor returns are universal for the airline industry.

Dempsey (2008) investigated the non-existence of profitability on average for the US airline industry after deregulation. In his research, the author finds financial performance to be unsatisfactory for long term sustenance of the company. Furthermore he notes that in the mid 1990s, airline capital expenditure originated for 65% from debt issuance (Dempsey, 2008). Dempsey (2008) states that eveb though this value is above the average US corporate industry debt to capital financing for investments, it still implies that a large remainder of investments are financed by equity investments. For more recent years, the Dempsey (2008) notes that there has been a substitution to higher levels of equity financing, even though equity financing is still considerably less than in other US industries. In all cases, the fact that a company succeeds in obtaining equity financing, in the case of market efficiency, implies that investors succeed in pricing the risk of the asset and obtain a return that compensates them for this level of risk.

Dempsey (2008) acknowledges however that the airline industry experiences both severe business and financial risk and low returns on investment. The high debt financing present in airlines can result in exponential changes in shareholder returns from minor revenue changes (Gritta et al., 1994). This implies that stock returns are sensitive to the business environment and suggest that airline stocks are volatile. In addition, it conveys that airline stocks carry significant systematic risk exposure and beta’s, which according to Capital Asset Pricing Model (hereinafter “CAPM”) investment theory, should result in a higher expected and required return on stocks. This suggests that realized returns on stock for airlines with higher systematic risk exposure should be higher as well. Otherwise, these stocks would incur smaller, more negative, values for Jensen’s Alpha. Perhaps Jensen’s Alpha is more negative for airlines with higher Beta’s due to operating leverage. Wohjan (2012) mentions that the “legacy”airlines7 may overinvest due to

myopia, by taking on higher investment commitments, or operating leverage, than optimal. Later Hu and Zhang (2015) mentioned that flexibility for airlines is beneficial, which suggest that airlines less flexible airlines with higher operating leverage or commitments, will have more difficulty to create value for shareholders and subsequently realize lower Jensen’s Alpha’s.

In the event that the calculated beta and hence the required return for this research is larger than the realized return of investors, there is suggestive evidence that investors are under compensated and fail to receive the return that compensates for the level of risk of US airlines (Berk & DeMarzo, 2011). Jensen’s Alpha for these airlines in a given year would therefore indicate

7 _{In his work, Wohjan (2012) mentions the term “legacy” airlines. With this he means the airlines that were market}

leaders and largest in size before the onset of low-cost Asian carriers. Among these airlines were American Airlines, United Airlines, etc. Wohjan argues that these airlines still feel like they can still live up to their performances in the past and hence are overoptimistic, causing them to overinvest.

(22)

15

a negative coefficient. There is literature stating that risk assessment and hence return expectations according to CAPM need to be adjusted. Jin, Martin and Bodie (2006) highlight the need to adjust required returns calculated according to CAPM. According to the authors, a CAPM beta of US companies fails to take into account pension expenses and risk. As mentioned before, the airline industry is labor intensive, and wage, as well as pensions, form a large cost to airlines. This may entail that return calculations for the airline industry, which will be subject to large pension payments (Benmelech et al., 2012), may be misleading without adjusting for this. Jin et al. (2006), explicate that according to their findings, on average, the equity risk of US firms reflects the pension risk of companies. Still, they provide no explicit account whether these results are externally valid and will also apply to the US airline industry. Hence, if the results to the research indicate negative Alpha’s for the industry on average, it does not necessary imply under compensation for its investors, but there will be suggestive evidence for under compensation. A more thorough analysis will then have to be performed to investigate why investors continue to invest in the airline industry, which is beyond the scope of this research. One argument for this is presented above: maybe CAPM fails to reflect the pension risk in airline industry investments and investors succeed in estimating the true risk. It is however too soon to argue this belief. Regardless, for this research it is posited that investors in the airline industry are under compensated on average, for reasons such as low profitability and high risk exposure mentioned above.

I expect to find more under compensation (lower Jensen Alpha’s) for less flexible, more operating leveraged firms based on the theory of Hu and Zhang (2015) as well as Wojahn (2012) that flexibility is rewarded in a volatile industry. In addition, based on theory of Dempsey (2008), the industry is highly levered. According to Dempsey (2008), this implies that additional leverage causes increasing required returns for investors, which seem infeasible given the industry’s legacy of poor profitability. Using the same argument that the industry is highly levered (Dempsey, 2008), it is believed that the coefficient on variables for debt levels will also be negative. In addition, based on the argumentation mentioned earlier by Pearce (2013), since the weighted average cost of financing (WACC) is above the return on capital (ROC) in the industry, growth is likely to be value destroying. Hence it is predicted that the Alpha’s will be lower for the airline which experience higher growth. Using the same argument, airlines with a higher return on assets are more likely to have a return on capital above the WACC, thereby creating shareholder value (Koller, 2005). It is posited that airlines with a higher return on assets will have higher Alpha’s. These hypotheses are all tested against the null hypothesis that there is no statistically significant coefficient from zero on these coefficients. As is done for other variables, this is tested against the alternative that the coefficient on the variable is statistically significant from zero.

(23)

16

CHAPTER 3 – Classification of High Committed Airlines

The following section will discuss the underlying methodology as to how the differences in airlines responses for high and low committed airlines during a crisis will be tested. The tested model will be explained and discussed, as well as how this research will determine the airlines that have high commitment and low commitments in a crisis.

3.1 The Committed Investment Model

Airlines with different degrees of committed investments are presumed to have similar business models in times of economic growth. During these times, liquidity shortages are less existent and it is assumed that most firms meet their financial obligations (Berk & DeMarzo, 2011). However, as argued before, periods with crises may be more stringent for airlines with predetermined capital commitments, or higher operating leverage (Berk & DeMarzo, 2011). Due to the capital (investment) commitments, airlines with higher levels of investments may be less flexible in reacting to a crisis. During crises the liquidity dries up, making it more difficult for firms to obtain financing (Cornett, Mcnutt, Strahan & Tehranian, 2010). Hence, airlines are expected to have similar corporate policy in periods of economic growth, but the degree of committed investments can lead to different business decisions with exogenous shocks, or crises. This reasoning suggests an empirical investigation based on a difference in difference (DD) regression model:

Yit = α +βHighcommittedi +γPostt +δHighcommittedi ×Postt + Controlsit +εit (1)

The dependent variable “Y” in this DD regression represents measures of employment, costs profitability, liquidity, future investment, and stock returns. Hence dependent variables contain, but are not limited to measures such as: staffing expense and employment, EBITDA, operating profit, SGA expense, cash holdings, capital expenditure, market capitalization and stock returns. Since this research will investigate two crises, two specifications of the same regressions are made. Apart from the sample size, crisis years and three DD dummy variables, the regression model for both crises are identical. Hence dependent on the regression, the Post dummy variable covers either the period after the 2001 crisis (post =1 for years 2001 and after) or marks the period following the 2008 great recession (post = 1 for years 2008 and after).

Similarly, the treatment group will differ for both specifications of the regression model. This is a natural result of the fact that the years used to classify airlines into high or low committed groups differ for both specifications. As a result, the airlines that are high committed in the regression specification covering the 2001 crisis are not necessarily the same airlines in the high

(24)

17

committed group for the regression covering the 2008 crisis. The predominant coefficient of interest for both specifications however is δ, for the DD estimator, or interaction term, highcommitted*post. If high committed investment airlines have different responses to shocks relative to low investment commitment airlines for a given variable, the coefficient is nonzero. Since both the Post and the Highcommitted groups differ for both specifications, the interaction term will also investigate a similar phenomenon but cover a different time period and group of airlines in each of the tested specifications.

The motivation as to why the empirical investigation is performed by a difference in difference (DD) regression is because it is presumed that investments are planned and committed if the future prospects are positive. During periods of economic growth, on average the performance, employment and returns of airlines are assumed to be identical regardless of committed investment levels. Liquidity shortages are less prominent in economic booms and hence it is predicted that also business decisions are fairly similar, not dependent on the levels of operating leverage or investment commitments. This altogether hints at a common shared trend prior to the unexpected shock for all airlines. When hit by an unexpected shock, it is presumed that airlines with high committed investments will react differently than low committed airlines. The liquidity in the economy following shocks is expected to dry up and make it more difficult to obtain financing (Cornett et al., 2010). As one argument, Duchin et al. (2010) state that being financially constrained and having financial obligations leads to lower stock returns. At the same time, Pearce (2013) argued that aircraft acquisitions lead to worse exposure to unexpected shocks. Besides stock returns, a disparity between high and low committed investment airlines is also expected for expenditure on staff and employment levels, as well as gross margin, EBITDA, and other measures of profitability. This is also because of the possible preference for committed airlines to fulfill contractual obligations in order to avoid fines. Because these airlines have tied up capital and less liquidity buffers when hit by unexpected shocks, it is likely that these firms initiate layoffs (Benmelech et al., 2012) and price wars (Busse, 2002) to avoid distress. Hence the tested hypotheses in this part of the research tests H0 = δ=0, which suggests that airlines with different degrees of committed investments have similar reactions to shocks, versus the alternative, H1 = δ≠0 that the degree of commitments in a crisis influence lay-offs, profitability and returns for airlines.

Besides the committed investments, the control variables natural log of total assets, return on assets, airline nationality, liquidity, debt due in 1 year, revenue growth, volatility in growth, and staff expense per worker, are added to the regression. The inclusion of control variable firm characteristics related to size and returns can be important to prevent endogeneity in the regression model. A complete justification is provided in section 5.5.1

(25)

18

3.2 The Nature of Committed Investment Decisions

The investigation of the first portion of the thesis is built on the idea that some airlines have high committed investments at a given time, while others will have low commitments. This paper posits that airlines subject to high commitments during an unexpected shock will react differently to maneuver out of the shock than low committed airlines. As a result, an essential part of this investigation is to investigate whether airlines at any point in time are highly committed or low committed. Hence, a threshold is determined to categorize firms as airlines with high or low committed investments. This threshold is based on a one standard deviation change from the average commitment level of the airline. To begin applying this procedure and to categorize airlines into certain groups, the commitment measure must first be defined.

In order to quantify the commitments of an airline, I begin by taking a discounted weighted average (WACC) of the committed capital investments. This weighted average calculates a present value of the investment payments of a given year, which are due for the years t+1 to t+6+_. Next, this weighted average total of all future investments in a given year is divided by revenues, total assets or employees. Phrased differently, these commitment measures are calculated as the airline’s present value of commitments, divided by revenues, total assets or employees for all airline-year observations in the sample. It is essential for a proper investigation of commitments that the present value of commitments is scaled either by revenues, total assets, or employees. This procedure is taken in order to account for size differences among the airlines in my sample. Doing so ensures that the airline size doesn’t drive the decision of whether an airline is highly committed. This normalization potentially allows smaller airlines (by size) to fall into this category and allows airlines to be compared on the basis of similar measures or values. For reasons mentioned in the robustness section (chapter 7), the main DD regressions will use the commitments scaled by employees measure. Scaling the commitments by total assets and total revenues is also done to check for robustness (later in this research).

In analyzing the commitment data, it becomes apparent that airlines classified and placed into the high committed investment category for a certain year, are also likely to be categorized into the high commitment category for the following year. The commitments appear to occur in waves where large orders are made and where total commitments gradually decrease year by year before new orders are placed. This becomes evident when testing the serial year on year correlation of investment commitments (table 4). Regressing the investment commitments of a given year on the commitments of the year before indicates a statistically significant negative coefficient at the 10% level. This predicts that if commitments rose in a given year, commitments are lower for the following year (column (1)). In other words, following increases in purchase commitments, it is not predicted that an airline with also increase commitments in the following

(26)

19

year. Regressing the first year change in commitments on the change in commitments two years later also indicates a negative coefficient, though not statistically significant (column (2)). In other words, if total commitments increase in year one, it is likely that the total outstanding commitments decreases in the following years. Purchase orders or commitments, once placed, generate higher commitments for the airline, which try to spread out this commitment over several yearly payments in the coming years (Lee, 2010). What this entails is that once an agreement is made about investment commitments, airlines typically have several years where the commitments are gradually reduced. Relative to year one, if the commitments increased, the total commitments are predicted to decrease in year two as well as in year three and four. The fourth year after (year five) however shows a positive relation with the change of the first year, even though this coefficient not statistically significant. In other words, this may indicate that if a purchase order is placed in a given year, four years later a new order will be placed. This raises the belief that airlines do not place new investment commitments for two to three consecutive years. Phrased differently, if an airline would place purchase orders in two consecutive years, the change in year two would be positively related to year one, and hence the regression would indicate a positive coefficient for column (1) and (2).

This finding would support the idea that investment commitments are placed randomly by airlines and are spread out over several years. In doing so, certain airlines will be “unfortunate” to have placed purchase orders a year or two before the crisis. These airlines are then subject to increased commitments in a crisis, which in a time of low liquidity, can push airlines into (further) financial problems or distress (Berk & DeMarzo, 2011). Despite this, there is reason to believe that some airlines never place investment commitments, but rather decide to lease aircraft. The descriptive statistics to the present value of the commitments (PVcommitments) and thus also the commitments/employees measure shows that for about 10% of the observations, the values amount to 0. Since the commitment variable is composed of the present value of investment commitments over total employees, in order to operate aircraft, airlines must either operate existing older owned aircraft (and not place investment orders) or only lease aircraft. The dataset conveys that airlines that lease in certain years are fairly consistent in their behavior and tend to lease in other years as well. These airlines will therefore categorize into the low investment commitment category (control group) regardless of the value for the investment commitment measure. For these airlines, the commitment measure will amount to zero for all years in which these airlines engage in a lease-only strategy to acquire aircraft. Please note that the robustness section (chapter 7) does investigate a different form of the scaled commitment variable, which does include leasing commitments.

(27)

20

3.3 Identifying High and Low Committed Airlines

Now that the nature of the commitments is investigated, it is possible to allocate airlines to either the treatment or control group. The methodology of determining whether an airline falls into the high committed (treatment) group or the control group is as follows.

As shown, airlines appeared to place purchase orders in waves or every few years. Also argued above was the idea that once commitments are placed, airlines maintain similar levels of commitments in the following years. Only a portion of the total outstanding commitments level is reduced in the years after a purchase order. This implies that those airlines placing purchase orders one to three years before the crisis will have high commitments during the crisis. As a result, the research will look at those airlines with high commitments for the years 1999-2001 for the 2002 crisis and for the years 2005-2007 for the Great recession of 2008. Having justified the years that are used to place airlines in the high committed (treatment) group or into the control group, this paper proceeds by mentioning the criterion used to allocate airlines in the groups. Keeping in mind that airlines place aircraft orders that will keep the commitment ratio higher for several years, it is necessary to look at changes in the commitment ratio. Investigating the changes in the commitment ratio will inform when a purchase order is placed since the change in the change in commitment ratio will be positive for those years. Hence airlines are placed in the highcommitted group when the changes (increases) in the commitment ratio are equal to or larger than one standard deviation. Doing so also controls for the possibility that some airlines may be able to sustain slightly larger commitments since allocating airlines is done on the basis of a standard deviation change and not absolute levels.

An analysis across the sample indicates that almost all airlines are considered highly committed at least some point across the sample, while a few airlines have such standard deviation changes more than once every four years. Therefore, to summarize, airlines with a one standard deviation increase in the commitment ratio for the years 1999-2001 are considered the high committed airlines in the 2002 crisis. Airlines with a one standard deviation increase in the commitment measure for the years 2005-2007 are considered high committed airlines for the 2008 crisis.

CHAPTER 4 – A Model of Investor Risk-Adjusted Compensation

As a first step in developing an understanding, as well as judging the realized investor returns in the airline industry, it is necessary to determine the fair8_{, risk-adjusted compensation. As is}

8_{Note, fair compensation implies that the investor receives an appropriate CAPM based compensation for the level}

of risk. It is assumed that investors are well diversified and therefore should receive compensation based on the risk free rate and a premium for every additional unit of systematic risk exposure. Any account to “fair” or “fairly” in the remainder of this thesis will imply the above mentioned, CAPM risk-adjusted return.