Uppsala university
Department of Economics D-level thesis
Supervisor: Sebastian Arslanogullari Spring term 2006
What Characterises Successful Stocks?
-A case study of Swedish companies between 1995 and 2005.
Gabriel Forss 800718-4735 Djäknegatan 40 018-25 80 19
gforzz@hotmail.com
Abstract
This paper discusses the indicators of financial success for Swedish companies from 1995 until 2005. Quarterly data on 42 Swedish companies were collected from the Datastream data base and analysed by using both portfolio analyses and parametric analysis. In this study, financial success is measured by using the acclaimed concepts of the Sharpe ratio and the Jensen’s Alpha. The Sharpe ratios of the companies are studied between 1995-2005 and this discussion is complemented by analysis of the Jensen’s Alpha in the second half of that time period i.e. 2000-2005. The relationship between these performance metrics and certain company-characteristics such as the book-to-market ratio, the ROA measure and capital structure is studied. The conclusion is that companies that have a high degree of profitability and maintain high book-to-market ratios outperform other companies in terms of generating excess returns to shareholders. Another interesting observation is the fact that company size does not have any significant relationship to company performance.
Keywords: financial success, Swedish companies, profitability, stock returns, book-to-market, Sharpe ratio, Jensen’s Alpha
Table of Contents
Chapter 1. Introduction ... 1
1.1 Aim... 2
1.2 Limitations ... 3
1.3 Methodology and data ... 4
1.4 Previous research... 5
1.5 Structure ... 6
Chapter 2. Independent Variables ... 7
2.1 The book-to-market ratio ... 7
2.2 Size ... 9
2.3 Return on Assets (ROA) ... 11
2.4 Capital Structure... 12
2.5 Liquidity ... 14
2.6 Cash Conversion Cycle ... 15
Chapter 3. Proxies for financial success... 17
3.1 The Sharpe Ratio... 17
3.2 The Jensen’s Alpha ... 18
Chapter 4. Results and discussion ... 20
4.1 Portfolio Analyses ... 20
4.1.1 ROA ... 20
4.1.2 Book-to-market ratio ... 21
4.1.3 Size ... 23
4.1.4 Liquidity ... 24
4.1.5 Capital structure and the cash conversion cycle... 26
4.2 Parametric analysis... 26
4.2.1 Reliability ... 28
Chapter 5. Concluding remarks... 30
5.1 Suggestions for future research ... 31
Bibliography... 32
Appendices ... 34
Appendix A: List of Companies... 34
Appendix B: Additional portfolio analyses... 35
Appendix C: Complete regression analyses... 37
Appendix D: Tests of Heteroskedasticity and autocorrelation... 39
Chapter 1. Introduction
Why are some companies more successful than others in terms of generating returns on stocks? That is undoubtedly a question that many an investor has contemplated over the years.
The question is as simple as it is intriguing. Indeed, this problem seems quite straightforward but at the same time it is obvious that there is no easy answer to it. Being able to firmly assert which companies would outperform others in terms of stock returns of course on the outset seems quite difficult. Nevertheless, the area opens up for much interesting research and innovative approaches. The central question contemplated is whether or not there are certain indicators of successful companies. Can success be broken down into different measurable factors and scrutinised? Are there in fact certain company-specific characteristics that would add up to a statement on what constitutes a so-called “winning bet” on the stock market?
These thought-stimulating postulations certainly provide scholars with many interesting starting points of analysis. Different frameworks of analysing success-related variables are also possible to set up. To be able to create and evaluate such frameworks is indeed an enticing thought to many scholars of financial economics and indeed to investors everywhere.
Such models are of course immensely difficult to construct, especially bearing in mind that these models hopefully in the end should serve somewhat of a predictive function. Predictive in the sense that one from such a model should be able to conclude that companies of a certain kind would yield larger returns than others. Obviously, reaching firm conclusions in such a matter is quite difficult a task. Nevertheless, the usefulness of a reliable model of this kind would undoubtedly be great.
Recently, the work of economists interested in company valuation has also been facilitated through the greater availability and user friendliness of comprehensive data bases which constitutes valuable sources of information. This evolution has to a great extent simplified the process of data collection of many researchers and has contributed greatly to the production of interesting articles and theses within the academic community.
Every rational investor is obviously interested in achieving the highest possible return on his investment. Consequently much attention within this field is aimed at determining what factors drive above stock market performance of certain stocks. Simply by quickly glancing through an economic journal it is evident that some stocks fare better than other and create returns that are in excess of the average on the stock market. Therefore it has for a long time
been of interest not just to the financial economics academic community but the actual financial practitioners, who are doing business and working in different consulting firms.
Building on this interest in achieving financial success, much research has been performed with regards to this phenomenon. Economists and financial advisors are constantly searching for winning stocks and this search has spurred a continuously growing field of research.
Therefore, there is certainly a huge bulk of literature to benefit from and to also be able to make a contribution into.
1.1 Aim
This study aims at investigating Swedish companies and determining what factors are key in terms of generating returns to shareholders. More specifically, the objective is to assert what impact certain variables have in terms of generating above-average stock returns. A conclusion will thus be made as to which are the indicators of successful companies (throughout this paper, success of a company will refer to the notion of creating excess wealth for shareholders in relation to other companies). Therefore it is assumed that the success of a company is not due to chance but that there is a relationship between some important company-specific factors and shareholder wealth. Perhaps these factors will not completely explain the variation in stock returns but certainly to a great extent. Great effort has therefore been made in choosing the explanatory variables and this will be discussed in greater detail later on. At the moment it suffices to simply mention which variables will be tested in terms of explaining the financial success of companies. As such, the variables have been chosen to reflect certain important financial aspect of the companies and the variables are: size, book-to- market ration, return on assets, capital structure, liquidity, cash conversion cycle. These concepts thus serve as potential indicators of superior financial performance. Indeed, these variables must also be assumed to illustrate important financial features of different companies and these should also be sustainable in the sense that they apply equally well across industries and types of businesses. This is important considering the wide range of businesses that are investigated in the study.
On the other side of the analysis one should at this stage mention that the variables that will define success are the Sharpe Ratio, and the Jensen Alpha. Financial success will thus be measured using these two different methods. These concepts will be dealt with in great detail later on but should nevertheless at least be mentioned now. As with the explanatory variables, these variables have been chosen from practice developed within earlier studies and they
serve as widely recognised and renowned measures of accumulation of excess shareholder wealth. All of these variables have been computed from information found in the Datastream data base.
1.2 Limitations
The companies used in this study are all companies that are registered on the Swedish stock market. Data for the companies are without exception taken from the data base called Datastream.1 For the selection of companies the starting point was all the companies that were registered at the stock exchange at the first of January 2005, which amounted to 271. The next step was to choose a year from which to start the study. Many aspects were considered, mainly time specific events such as tax reforms, regulations of the financial markets and the presence of the financial crisis of Sweden in the beginning of the 90s. All of these instances would surely affect the financial data of the companies but in the end the year 1995 was chosen as a good time of departure since many of the above-mentioned events had taken place and been internalised into the market. Thus the time period selected for the study ranges from 1995 until 2005. Selecting this time period undoubtedly had some precise implications for the availability of data and the selection of units to be studied. It was contemplated to choose a somewhat shorter time period. Surely, such an outlook would have simplified the matters of finding complete data sets, but in the end it was nevertheless decided to stick with the original period and instead cope with having fewer companies present in the study. Recognising that many overwhelming changes occurred within the financial system and the capital market in the early years of the 1990s, for the reasons mentioned above 1995 was chosen as a good year of departure for this study.
With regards to the selection of companies, it was essentially restricted by the availability of data. Finding complete data sets for the companies was necessary albeit tedious work.
Evidently the companies had to exist for the entire time period and complete sets of data on the variables in the study were needed .Working backwards from 2005 to 1995, many lapses and instances of missing data in the data sets in Datastream narrowed the selection down to 42 companies. This process of natural selection so to speak, although not optimal is nevertheless a factor which shouldn’t hamper the study to any great extent, especially not statistically anyway.
1 Datastream available at Uppsala university.
1.3 Methodology and data
This paper is an empirical analysis of a selection of Swedish companies. Being quantitative in its design, the study uses quarterly data and assesses the importance of six company-specific characteristics on the ability of companies to generate shareholder wealth. More specifically, the study is divided into two parts. A more practically-oriented part consisting of investment strategies will be accompanied with a parametric method of regression analyses.
The first part is based on a method commonly used within the business of stock valuation.
The procedure is as follows. At time t, the companies will be sorted into seven different portfolios according to their score on one of the six indicators used to affect financial success.
This means that the companies with the highest book-to-market ratio (for instance) will be placed in portfolio number one whereas companies with lower book-to-market ratios will be divided into corresponding groups with portfolio number seven consisting of the companies with the lowest ratios. At t+1 the objective is then to study how these different groups of companies fare with regards to generating shareholder returns (measured by the Sharpe Ratio and the Jensen’s Alpha). At time t+1 the procedure is repeated with new groups of companies being constructed due to their book-to-market ratios. Similarly the creation of shareholder returns is studied over the subsequent time period and so on. This process will be repeated with all of our indicator variables. A difference in shareholder returns should hopefully be distinguishable, at least between portfolio number one and portfolio number seven. Albeit quite straightforward and investment-based, this valuation method suffers from not being able to ascertain whether the results are statistically robust or not. Therefore further statistical analysis needs to be incorporated into the study.
To accompany this broad method of company valuation a parametric analysis will follow.
Thus, the relationship between the indicator variables and the two measures of financial success will be studied. The method used to evaluate this relationship will be a plain and simple ordinary least squares (OLS) regression model. The intuition and workings of such an estimation is well-known and there is no need to further dwell on it here. The use of it is indeed quite conventional and should be familiar to all scholars of financial economics.
Consequently, the x-variables will thus contain the quarterly data collected for the 40 companies on the six indicator variables. These are subsequently regressed on the measures of financial success, first on the Sharpe Ratio and then on the Jensen’s Alpha. Computing the
quarterly numbers for the Sharpe Ratio is easily done but generating time series of Jensen Alphas is a bit trickier. Essentially, the Jensen Alpha is the alpha that appears in the Capital Asset Pricing Model (CAPM) equation2 which means that if one is supposed to compute time series on alpha, there is a need to know the beta for the company. Therefore, the beta values for the different companies during 1995-2005 have been estimated using the CAPM equation.
This estimation of beta has then been used to generate the company-specific values of the Jensen Alpha on a quarterly basis.
Furthermore, in order to make the regression analysis interesting from an investing point of view, the indicator variables have been lagged by one time period. This means that for instance the ROA value at time t will be coupled with the Sharpe Ratio of time t+1. This will produce a model where the predictive functions of the indicator variables will be thoroughly tested.
1.4 Previous research
This paper benefits greatly from an earlier study by Johnson and Soenen (2003). In fact their article is the chief instance of motivation for this present paper. In that article the authors perform a study of 478 American companies during a time period of 16 years and investigate what are good and reliable indicators of success of those companies. Both in terms of reference literature and other practical issues this study has been of great inspiration to me.
Nevertheless, due to certain restrictions, such as time and scope and what not, I have not been able to perform as comprehensive a study as Johnson and Soenen. The similarities between our work should however be acknowledged. As to my knowledge no previous study of this kind has been performed on Swedish companies and therein lays the contribution of this paper to the scientific community. Hopefully further insights into the characteristics of Swedish companies and the Swedish stock market will thus be generated.
The explanatory variables, which will be discussed in the following chapter, have been selected from those used by Johnson and Soenen and from other similar studies. They represent a broad spectrum of company-specific characteristics and thoroughly portray the financial situations of the different companies. A couple of other studies could be mentioned as more frequently cited in this paper and among these the influential work of Fama and French (1992) and (1998a) is greatly used. Their work on the importance of the much
2 Ri-Rf=alpha+beta(Rm-Rf)
publicised book-to-market ratio has been quite important for this study. Also the work of Sharpe (1994) and Jensen (1969) have been used to a great extent for the calculation of the different measures used to operationalise financial success.
1.5 Structure
The structure of this paper is quite straightforward. This introduction will be followed by two chapters that comprehensively describe and discuss and the choice of variables in the study.
The fourth chapter will contain the results of the OLS regression analysis and will also contain a discussion of the results obtained. Finally, a concluding chapter will round off the paper with a summary of the conclusions reached by the study. Ideas of further research will also be mentioned.
Chapter 2. Independent Variables
This chapter will give a thorough description of the choice of independent variables for the study. These variables have been selected from the vast literature of previous studies on this subject. The discussion in this chapter will also mention the ways in which these variables have been operationalised.
2.1 The book-to-market ratio
In a much renowned article Fama and French (1992) investigate the importance of company size and the book-to-market ratio in order to explain the returns of individual stocks. One of their important conclusions is that the book-to-market ratio is indeed quite instrumental in explaining average stock returns. In fact they state that there is “a strong cross-sectional relationship between average returns and book-to-market equity.”3 The usefulness of the book-to-market ratio in order to explain and predict the success of a company is further acknowledged by other studies. For example Rosenberg, Reid and Lanstein (1985) show that average returns on U.S. stocks are positively related to the companies’ ratio of book values of common equity to their market value. This observation holds for the Japanese firms as well as Chan, Hamao and Lakonishok (1991) report that the book-to-market ratio in fact is equally important in explaining the financial success of the companies registered on the Japanese stock exchange. Nevertheless, the emergence of the book-to-market ratio as an important determining factor in order to predict stock returns was firmly established by Fama and French (1992) and (1998a,b).
The book-to market ratio is also extensively used as an indicator of a firm’s growth opportunities. Kothari and Shanken (1997) study the significance of the book-to-market ratio and the notion of dividend yield as indicators to predict stock returns. Their investigation show that the book-to-market ratio indeed provides a good measure of financial success of the Dow Jones Industrial Average for the time period 1926-1991. Even though dividend yield holds a slightly stronger explanatory position in certain sub periods, the book-to-market ratio remains the stronger of the two variables for the entire period. They thus provide reliable evidence that the book-to-market ratio track time series variation in stock returns.
3 Fama and French (1992), p. 440.
Further evidence of the usefulness of the book-to-market ratio as an indicator of stock returns is given by Pontiff and Schall (1998). In a similar vein as Kothari and Schanken (1997) they investigate the book-to-market ratios of the Dow Jones Industrial Average. Their conclusion is that the book-to-market ratio indeed serves as a good forecast of future market return on stocks. In their study they also include and control for other variables that have previously been assumed to predict market returns, such as default spreads, interest rates, term structure slopes, and dividend yields. These indicators do not share the same predictive and forecasting power of the book-to-market ratio however and the usefulness of the book-to-market ratio as an indicator of stock return is once again emphasised.
Fama and French (1998b) ascertain that investors classify stocks that have high ratios of earnings to price, cash flow to price or indeed book-to-market as value stocks. A value stock is a stock that tends to trade at a lower price in relation to its fundamentals; such a stock is therefore considered undervalued by investors. Value stocks are often contrasted against growth stocks. Growth stocks (sometimes referred to as ‘glamour stocks’) are found in companies that are growing. These stocks do not usually generate dividends as the company prefers to put their retained earnings into new investments. Stocks from growing companies are often overvalued, having low book-to-market ratios.
Nevertheless, interested in the book-to-market ratio a value investor would thus believe that the market is inefficient and that it is possible to find stocks that are traded for less than they are worth. Indeed several studies, including Fama and French (1998b) and Lakonishok, Schleifer and Vishny (1994) contend that there is a significant value premium for U.S. stocks.
Value stocks, i.e. stocks with high ratios of book-to-market (B/M), earnings to price (E/P), and cash flow to price (C/P) do create higher returns on the average than stocks with low B/M, E/P, C/P. There is also evidence that this value premium exists in other markets as well, for instance Chan, Hamao, and Lakonishok (1991) reports corresponding findings is Japan.
Fama and French (1998b), studying the returns generate by value stocks on the one hand and growth stocks on the other conclude that “value stocks tend to have higher returns than growth stocks in markets around the world” and “sorting on book-to-market equity, value stocks outperform growth stocks in twelve of thirteen different major markets during the 1975-1995 period.”4
4 Fama and French (1998b), p. 1997.
In this study, the book-to-market ratio is reported as an indicator of value stocks. The book-to- market ratio is calculated in a straightforward manner following Fama and French (1992), taking the book value of common equity in the firm divided by their corresponding market value.
2.2 Size
As Johnson and Soenen (2003) states, company size is “the second most publicized variable to explain stock returns.”5 The preoccupation of company size as a proxy to depict stock returns, hence creating shareholder value, is thus evident in much research. Fama and French (1992) maintain that in the presence of rational pricing of assets, stock risks are multidimensional. An aspect of this multidimensionality is consequently that of company size.
Since Fama and French (1992) were very much interested in deriving the relationship between financial leverage and security returns, they excluded financial firms in their selection of sample firms. Because high leverage is quite normal for financial firms and not necessarily a signal of financial distress (as it typically is for non-financial firms), financial firms were left out of their study. The Fama and French study did however single out the variables of company size and the book-to-market ratio as being the most important ones in terms of generating security returns. The chief finding of their study is that stock returns are negatively related to size but positively related to the book-to-market ratio.
In subsequent research performed by Barber and Lyon (1997) the robustness of the results of Fama and French (1992) are tested by examining the applicability of size and book-to-market ratio as proxies explaining the stock returns of financial firms which, as noted, were excluded in the previous study. Barber and Lyon’s conclusion is that financial and non-financial firms indeed have very similar return patterns and that “both financial and nonfinancial firms exhibit a significant size and book-to-market premium.”6 Small companies with high book-to- market ratios tend to achieve higher stock returns. Barber and Lyon are also unable to reject the hypothesis that financial and non-financial firms would differ in terms of their premium on the variables of size and book-to-market ratio. Therefore, “firm size and book-to-market ratios have similar meanings for financial and nonfinancial firms – at least as they relate to security returns.”7
5 Johson and Soenen (2003), p. 365.
6 Barber and Lyon (1997), p. 876.
7 Id.
Laporta et al. (1997) find similar evidence of the similarity between financial and non- financial companies with regards to the usefulness of size and book-to-market ratio as indicators of superior stock returns. Evidently, these two variables explain the stock returns of companies in an economically meaningful way, both financial and non-financial ones.
Furthermore, Laporta et al. (1997) discusse the role of expectational errors by investors as a possible explanation why value stocks outperform growth stocks. These behaviouristic explanations are undoubtedly gaining ground in the world of finance these days. There is no doubt however, that the size and book-to-market ratio of a company represent two core fundamentals that are highly correlated to stock returns.
Rouwenhoust (1999) is also preoccupied with size and the distinction between value stocks and growth stocks. As has already been noted, numerous studies confirm that American stocks exhibit a significant relationship both to size and the book-to-market ratio.
Rouwenhoust reports that similar studies have been performed on developed equity markets outside of the United States with findings similar to those of American stocks. In an attempt to extend this analysis to emerging markets, Rouwenhousts study shows that the many of the characteristics of the developed equity markets are prevalent in emerging markets as well.
More importantly, Rouwenhoust concludes that “averaged across all emerging markets, stocks exhibit momentum, small stocks outperform large stocks, and value stocks outperform growth stocks.”8
It is interesting to note that, whereas Fama and French (1992), Rouwenhoust (1999), and Barber and Lyon (1997) report a negative relationship between size and stock returns, Johnson and Soenen (2003) find just the opposite. Johnson and Soenen in their study find that size indeed shows a significant positive relationship to stock returns. This positive relationship is also documented in Shefrin and Statman (1995).
There are several different ways of measuring company size. There is the possibility of computing company size by taking the price per share times the shares outstanding or simply using the market value of the firm as estimation where such information is available. In this study, the market value of the company will be used as a proxy for company size.
8 Rouwenhoust (1999), p.1441.
2.3 Return on Assets (ROA)
As a measure of profitability the concept of return on assets (ROA) is commonly used. ROA is thus a proxy for how profitable a company is relative to its total assets. As Johnson and Soenen (2003) state, ROA “is an asset utilization ratio that indicates how effectively or efficiently a firm uses its assets.”9 Thus, ROA indeed gives an idea of how efficient management is at generating earnings in relation to the firm’s total assets. Their study also reports a significant positive relationship between stock returns and ROA.
Furthermore, as highlighted by Johnson and Soenen, the “effectiveness with which a fixed capital, working capital and other assets are employed obviously is a driver of growth.”10 Therefore the profitability of the firm, exemplified by a greater level of ROA, gives an idea of the growth possibilities of the firm. With regards to investment strategies, the profitability of the firm indeed is an important characteristic which is important to pay attention to.
In this study, the ROA is computed as the ratio between the net income of the firm relative to their total assets. Thus it is displayed as a percentage which shows what earnings are generated from invested capital (assets). The ROA ratio is thus a good indication of whether or not management are continuing to earn increasing profits on investments. Investors would thus expect good management to strive for a greater ROA ratio, since that would mean that greater profit on each invested assets is being extracted.
Another common indicator of profitability which is commonly used by financial analysts is the return on equity (the ROE). Though still perhaps being the most widely employed metric of profitability there is evidently some shortcomings to it. Evidently, the ROE does not tell investors anything about whether or not a company has excessive debt or is using debt to drive returns. Such factors are obviously important to take into considerations when planning investments. In such situations, it is important to note that the ROA is computed with total assets in the denominator. Total assets are of course calculated as the sum of liabilities and shareholders equity. Therefore, the debt situation of the firm is incorporated into this measure and thus resolves one of the problems typically associated with the ROE measure.
9 Johnson and Soenen (2003), p.365.
10 Id.
2.4 Capital Structure
Ross, Westerfeld and Jaffe (2005) suggest that “the theories of capital structure are among the most elegant and sophisticated in the field of finance.”11 This bold statement hints at the importance of a firm’s capital structure for achieving greater financial success for the company. In an influential paper Bradley, Jarrell, and Kim (1984) assert that the notion of finding an optimal capital structure for the firm has been “one of the most contentious issues in the theory of finance during the past quarter century.”12 This interest and preoccupation with capital structure has been huge and it is obvious that this field of study has continued to generate great interest in recent years. Nevertheless, as highlighted by Ross, Westerfeld and Jaffe it is difficult to prescribe any exact formula for computing the optimal rate of leverage for the firm. Evidently, the debt-to-equity ratio is quite an elusive concept. Nevertheless, an intriguing relationship between the leverage and the profitability of a firm seems to exist.
In this study the capital structure of the firms refer their employment of leverage; therefore it is measured as the ratio of long-term debt to total assets. Financial leverage is often believed to comprise many advantages. If there are investment opportunities available to a firm which would generate a higher return than the required interest rate for borrowed funds, additional leverage would thus be beneficial. These differences generate greater profits for shareholders and increase the return on equity. As Johnson and Soenen (2003) points out, this argument “is largely based on the tax deductibility of interest expenses making borrowing a ‘cheaper’
source of financing than equity.”13
Using debt as a tool to drive returns hinges on the assumptions behind the so-called ‘trade-off theory’ of capital structure. This theory posits that there is a trade-off between the “tax advantages of debt and various leverage related costs.”14 Such costs are those often associated with financial distress and include both direct and indirect costs. The trade-off theory thus implies that there is an optimal level of debt to every company which consequently should be pursued by financial managers.
11 Ross, Westerfeld and Jaffe (2005), p. 461.
12 Bradley, Jarrell, and Kim (1984), p. 857.
13 Johnson and Soenen (2003), p. 365.
14 Bradley, Jarrell, and Kim (1984), p. 857.
A line of thought which holds several objections to the trade-off theory of capital structure is the one commonly attributed to the work of S.C. Myers15, namely the pecking-order theory.
The main point of difference compared to the trade-off theory is that a target level of leverage for the firm does not exist and that profitable firms use less debt. Therefore a negative relationship between profitability and leverage is implied. As described by Ross, Westerfeld and Jaffe profitable firms generate internal funds which can be used to finance projects. This type of financial slack is often quite beneficial and desirable for the firm. The source of financing available to firms can therefore be ranked in order, hence the name pecking-order theory.
As described by Leary and Roberts (2004) internal funds are at the bottom of this pecking order with equity being at the top. The firms’ preference order of financing sources is due to asymmetric information between managers and investors with timing being an essential attribute of financial managers. Leary and Roberts (2004) assert that “since internal funds avoid informational problems entirely, they are at the bottom of the pecking order.”16 Accordingly, equity holds the largest adverse selection costs and “when internal funds are insufficient to meet financing needs (i.e. financing deficit), firms turn first to risk-free debt, then risky debt, and finally equity, which is at the top of the pecking order.”17 Leary and Roberts also do point out that the empirical evidence concerning the pecking-order theory is conflicting, with several studies suffering from statistical power problems. Therefore it is subject to extensive research which will determine which way to go concerning capital structure.
Nevertheless, with regards to the capital structure of firms there seems to be contradictory results stemming from different empirical studies. Bhandari (1988) finds evidence that “the expected returns on common stocks are positively related to the debt/equity ratio (DER), controlling for the beta and firm size.”18 Fama and French (1998a) on the other hand, provide conclusions which are at odds with the generally accepted view of the relationship between debt and firm value. Instead of the assumed positive relationship between debt and firm value, they find no reliable evidence of the supposed tax effect. Their results point to a negative
15 Myers,S.C. (1984) The Capital Structure Puzzle, Journal of Finance 39.
16 Leary and Roberts (2004), p.5.
17 Id.
18 Bhandari (1988), p.527.
relationship between these two notions exposing the fact that increased leverage is bad news for firm value.
2.5 Liquidity
A notion that is undoubtedly related to the previous discussion on capital structure is that of company liquidity (previously referred to as financial slack). Measuring the liquidity of a company is usually done by taking the most liquid assets, i.e. cash and other marketable securities, as a fraction of the total assets. This method is also employed in this study where liquidity is computed as the ratio between common assets and total assets. The term common assets is used in the Datastream data base referring to notions such as cash and short term securities. This term is deemed to be an adequate measure of the liquid resources of the companies.
The liquidity of a company is important in several respects. First and foremost, having a stock of cash at hand greatly facilitates the process of being able to perform new investments.
Obviously, a greater degree of liquidity means that the company is likely to have sufficient funds ready for upcoming investment projects. When projects with positive net present values emerge it is often desirable to act quickly and in such situations internally generated funds are of the utmost importance. Johnson and Soenen (2003) highlight this aspect of liquidity by stressing that having cash at hand is “most valuable to firms with plenty of positive-NPV growth opportunities.”19 Therefore, the ability of being able to use internal funds as the chief resources of funding for new investments is the most important advantage of having a high degree of liquidity within the company.
However, there are also some documented downsides to having a large amount of liquid resources at the company’s disposal. An excessive amount of liquid resources is not always beneficial for the firm. According to Johnson and Soenen such a situation might signal “slack management practises.” From the onset there is obviously already an agency problem between management and share holders which can be augmented by an increased amount of cash at the management’s disposal. The relationship between shareholders and management can be described as a principal – agent problem from which certain typical features can be discerned.
Jensen (1984) describes the relationship between shareholders and corporate managers as one
‘fraught with conflicting interests.’ It can be assumed that both the shareholders (the
19 Johnson and Soenen (2003), p.366.
principals) and management (the agents) share some fundamental interests, that the company is successful etc. Be that as it may, there are undoubtedly some diverging views with regards to deal with any potential financial slack. The directors of the company might use such funds for expanding their on-the-job consumption, having a greater spending account or other job perks and requisites. The shareholders would perhaps rather have the financial slack returned to them in form of dividends or invested in new projects. Situations as the one described above is undoubtedly quite frequent in the business world. Jensen (1984) discusses the negative aspects of having plenty of free cash flow and stresses that managers in such situations tend to make bad acquisitions. Managers might thus be encouraged to take it easy when the free cash flow situation is excessive.
2.6 Cash Conversion Cycle
Johnson and Soenen (2003) assert that “efficient working capital management is an integral part of overall corporate strategy to create shareholder value.”20 Working capital refers to the capital that a company needs in order to handle its day-to-day operations. It is commonly accepted as an indicator of a company’s efficiency and its short term financial health. The concept of the cash conversion cycle typically refers to “the continuing flow of cash from suppliers to inventory to accounts receivables and back into cash.”21 Originally, the notion of the cash cycle stems from the work of Gitman (1974). In this article Gitman provides a complement to original cash flow analysis for establishing the minimum liquidity requirements of the company. The technique he derives makes it possible to make quick estimates of company specific liquidity preferences.
Nevertheless, building on the work of Gitman (1974), the concept of the cash conversion cycle was further developed by Richards and Laughlin (1980). Essentially, Richards and Laughlin were concerned with the quite static view of handling the balance sheet liquidity ratios in order to calculate the company’s liquidity position. In their view, the cash conversion cycle was best conceptualised as “reflecting the net time interval between actual cash expenditures on a firm's purchase of productive resources and the ultimate recovery of cash receipts from product sales.”22 It was also necessary to adopt a payables turnover concept in
20 Johnson and Soenen (2003), p.366.
21 Id.
22 Richards and Laughlin (1980), p.34.
order to extend “the operating cycle analysis to incorporate both the relevant outflow and inflow components.”23
In this study the cash conversion cycle is measured following Johson and Soenen (2003) and is computed as (inventories + accounts receivables - accounts payables) *360 / total sales.
Their calculation is very much based on the concept established by Gitman (1974) and Richards and Laughlin (1980).
23 Richards and Laughlin (1980), p.36.
Chapter 3. Proxies for financial success
This chapter deals with the two variables defining financial success in the study, namely the Sharpe Ratio and the Jensen’s Alpha. A general background to the functioning of these variables will be accompanied with a description of the computations used in order to derive them.
Before we turn to the measures of financial success used in this study it is of interest to say a few words of the selection of these concepts. Evidently, there are several measures available which are used to measure financial success. Examples of such concepts are for instance the Treynor ratio, and the method of computing the Economic Value Added (EVA). The Treynor ratio is quite similar to the Sharpe ratio but differs in the respect that the Treynor ratio uses the beta of the portfolio as a measure of its volatility whereas the Sharpe ratio uses the portfolios standard deviation. Taking the market risk into account obviously reveals much information, and it should be accounted for in some way. In this study, the market risk aspect of the portfolios is captured in the use of the Jensen’s Alpha as a complement to the Sharpe ratio.
The EVA measure has received a lot of attention in recent years.24 In essence this concept describes what amount of shareholder value the company is creating. Thus, this performance metric is computed by subtracting the cost of a company’s capital (both equity and debt) from its operating profits. Johnson and Soenen (2003) use the EVA as a performance metric of financial success in their study and also provide an outline of how to compute it using data from the Compusat data base. Though, similar to the data base used in this study of Swedish companies, the concepts used in Compusat are not completely comparable to the ones used in Datastream. For that matter it was also much a question of availability of data. Therefore, the EVA was not incorporated into this study and the two metrics of performance that are used are consequently that of the Sharpe ratio and the Jensen’s Alpha.
3.1 The Sharpe Ratio
The Sharpe ratio is essentially a measure of a portfolio’s risk-adjusted return. In fact the ratio measures performance relative to total risk. It is a very practical proxy because it reveals information whether or not the excess returns of a portfolio come with much additional risk.
24 see for example Hecking (2002)
Therefore, the Sharpe ratio is often used to distinguish ‘good investments’ from ‘bad investments’ because although a portfolio may generate greater returns than its peers it is important to determine if those excess returns do not come with too much extra risk. In Sharpe (1994) the user friendliness of the Sharpe ratio is discussed and the ratio is calculated by using the following formula.
In order to compute the Sharpe ratio it is thus necessary to take the risk free interest rate into account. In this study, the risk free rate is calculated as the average of the three months market interest rate of the period. Data on the three months rate were taken from the SCB25 and computed accordingly.
Evidently, the Sharpe ratio represents a useful method for evaluating the financial performance of companies. It does, however, have some shortcomings which highlight the need to complement it with an additional performance metric.
3.2 The Jensen’s Alpha
The Jensen measure is based on the Capital Asset Pricing Model (CAPM). More specifically, it is a risk-adjusted performance measure that represents the average return above the return predicted by the CAPM, given the portfolio’s beta and the average market return. This rate of return is the portfolios alpha, hence the name Jensen’s alpha.
25 www.scb.se
The interesting aspect conveyed by Jensen’s Alpha, as compared to the Sharpe Ratio is the fact that that Jensen’s Alpha brings the market risk into consideration. Therefore, the measure accounts how well portfolio manager is at dealing with the systematic risk. Essentially, the Jensen’s Alpha gives an indication of the degree to which the portfolios are earning significant returns after accounting for market risk, as measured by beta. If the portfolio is earning a fair return, for the given market risk, then alpha would be zero. A positive value of the Jensen’s Alpha is thus a sign of a portfolio earning greater returns than would be expected by the CAPM.
With regards to the computation of the Jensen’s Alphas in this study the market return has been computed by using the general index. It was thus decided to use a broad index for the bench-mark portfolio and the general index was a top candidate for that position. Also, for this study to work, it was necessary to calculate time series of the Jensen measure to comply with the other collected data. In order to generate series of Jensen Alphas on a quarterly basis it was thus necessary to determine the value of beta for the different companies used in the study. It was contemplated to use the measures found in the publication “Börsguiden” but this idea was discarded due to lack of available information. Instead, it was necessary to calculate these beta values using regression analyses. A problem that emerged was the fact that it was impossible to calculate statistically correct beta values for the entire time period 1995-2005.
In order to avoid the statistical problem of “errors in variables”26 it was needed to limit the values of Jensen Alpha to the latter part of the time period. More specifically this was done by dividing the sample companies into two different parts with respect to two different time periods. The first group consists of the time period 1995-2000 and the second group thus deals with the time period 2000-2005. Subsequently, using regression analyses on the first time period, beta values were obtained for the companies. These beta values were consequently used to generate quarterly time series of alpha for the second time period.
Therefore, analysing the indicator variables in connection with the Jensen Alphas will only be performed for the time period 2000-2005.
26 see Campbell, Lo, and Mackinley, (1997) for a discussion of this phenomenon.
Chapter 4. Results and discussion
This section conveys the results of the analysis of what are the indicators of financial success for Swedish companies during the time period 1995-2005. The method of portfolio analysis which was introduced earlier constitutes a point of departure for presenting the outcome of the investigation.
4.1 Portfolio Analyses
4.1.1 ROA
The first portfolio analysis concerns the profitability measure ROA. Not surprisingly, the portfolios that exhibited higher ratios of ROA generated superior returns in the following quarter. The results are presented in the table below.
Table 1
This table presents the average returns (provided by the Sharpe Ratio) of seven different portfolios which are ranked according to their score on the ROA measure. Portfolios range from 1995-2005 and are reformed every quarter.
Portfolio 1 (high ROA) 0,067319
Portfolio 2 0,011477
Portfolio 3 0,087027
Portfolio 4 0,065428
Portfolio 5 -0,02084
Portfolio 6 -0,12875
Portfolio 7 (low ROA) -0,26444
The Sharpe ratios of the different portfolios exhibit a clear and visible pattern with the top four portfolios generating positive values of the Sharpe ratio whereas the bottom three perform very poorly. Especially, the portfolio consisting of the least profitable companies achieve low returns on investment. By comparing the two extreme portfolios it is obvious that the companies with a high degree of profitability outperform the low-profitability companies.
This result is in line with the conclusions drawn by Johnson and Soenen (2003) who find that the ROA measure is the variable which has the greatest impact on the performance measures for financial success. They report a relatively strong, statistically significant (at the 99% level) relationship between ROA and the Sharpe ratio and Jensen’s Alpha respectively.
Turning now to the additional portfolio analysis, which illustrates the performance of the portfolios in relation to the Jensen Alpha, the same pattern as in the Sharpe ratio discussion is distinguishable. Portfolios with greater levels of ROA, do fare better than less profitable companies. The results from the portfolio analysis concerning Jensen’s Alpha are posted in the following table.
Table 2
This table presents the average returns (provided by Jensen’s Alpha) of seven different portfolios which are ranked according to their score on the ROA measure. Portfolios range from 2000-2005 and are reformed every quarter.
Portfolio 1 (high ROA) 0,032716
Portfolio 2 0,026646
Portfolio 3 0,023681
Portfolio 4 0,040685
Portfolio 5 0,00507
Portfolio 6 -0,01828
Portfolio 7 (low ROA) -0,03173
When comparing the results from the Sharpe ratio analysis with the Jensen Alpha measures from the second half of the time period it is obvious that the Sharpe ratios are more differentiated. Compared to the Jensen Alphas, the Sharpe ratios span over a wider spectrum, ranging all the way down to quite large negative returns. Obviously, the Sharpe ratios are measured over a longer time period which make them more susceptible to different sorts of time specific events and other historic occurrences. The financial development of the companies on the Stockholm stock exchange might quite plausibly have been somewhat volatile during the latter half of the 1990’s. The start of the 20th century does similarly not contain such stocks as was prevalent during the 1990’s. Such matters are of course of the utmost importance when comparing time series of this kind.
4.1.2 Book-to-market ratio
Similar to the ROA measure the same pattern is visible when analysing the impact of the book-to-market ratio on financial success. The top portfolio, which consists of the companies with the largest proportion of common equity in relation to their market value, generates a
Sharpe ratio of 0,046 whereas the portfolio with the lowest book-to-market value performs worse. The results are presented in the following table.
Table 3
This table presents the average returns (provided by the Sharpe Ratio) of seven different portfolios which are ranked according to their book-to-market ratio. Portfolios range from 1995-2005 and are reformed every quarter.
Portfolio 1 (high book-to-market ratio) 0,046
Portfolio 2 -0,097
Portfolio 3 0,022
Portfolio 4 0,0003
Portfolio 5 -0,07
Portfolio 6 -0,083
Portfolio 7 (low book-to-market ratio) -0,0006
As in Fama and French (1992), the book-to-market ratio exhibits a positive relationship to the Sharpe ratio. It is somewhat peculiar that portfolio number two records such a surprisingly bad return. It should be noted, however, that there are studies that predict a negative relationship between the book-to-market ratio and financial success. Johnson and Soenen (2003) and Shefrin and Statman (1995) actually find a negative relationship between the book-to-market ratio and financial success.
By contrast to Shefrin and Statman, the Jensen’s Alpha analysis suggests that companies with larger ratios of common equity in relation to market value give greater returns than companies with low book-to-market ratios. The numbers are presented in the following table.
Table 4
This table presents the average returns (provided by Jensen’s Alpha) of seven different portfolios which are ranked according to their book-to-market ratio. Portfolios range from 2000-2005 and are reformed every quarter.
Portfolio 1 (high book-to-market ratio) 0,027096
Portfolio 2 0,008439
Portfolio 3 0,045816
Portfolio 4 0,008084
Portfolio 5 0,000983
Portfolio 6 0,002171
Portfolio 7 (low book-to-market ratio) -0,01381
These numbers show a clear pattern of a positive relationship between the book-to-market ratio and the Jensen’s Alphas for the companies. Whereas the top portfolio yields positive returns, the bottom portfolio does not reproduce the same positive numbers. These results are in support of the Fama and French (1992) study and suggest that Swedish firms exhibit a positive relationship between book-to-market and financial success.
4.1.3 Size
A much studied relationship is that between financial success and firm size. Several studies, including Shefrin and Statman (1995) and Johnson and Soenen (2003), report that large companies fare better than small ones in terms of generating excess returns to shareholders.
Their results, albeit statistically significant, are at odds with previous investigations. Fama and French (1992), for example, find that stock returns are negatively related to size. Because several studies point to a robust relationship between size and financial success, it was expected to find similar patterns in the analysis of Swedish companies. The results are somewhat inconclusive, however, as is evident from the following table.
Table 5
This table presents the average returns (provided by the Sharpe Ratio) of seven different portfolios which are ranked according to their market value. Portfolios range from 1995-2005 and are reformed every quarter.
Portfolio 1 (high market value) -0,014
Portfolio 2 -0,036
Portfolio 3 0,0042
Portfolio 4 0,053
Portfolio 5 -0,074
Portfolio 6 0,0003 Portfolio 7 (low market value) -0,12
Evidently, the portfolio analysis does not present any evidence of company size being either positively or negatively correlated to financial success. The numbers are quite random, with middle-sized companies being the best bet for investment. Nevertheless, it is evident that no robust relationship can be observed.
When contemplating the results from the Jensen’s Alpha analysis, the same uncertain pattern as in the Sharpe ratio investigation is discernable. Even though the portfolio consisting of the biggest companies outperform the portfolios with the low market value companies, the top return-yielding companies seem to be the middle-sized ones. Table 6 presents the numbers.
Table 6
This table presents the average returns (provided by Jensen’s Alpha) of seven different portfolios which are ranked according to their market value. Portfolios range from 2000-2005 and are reformed every quarter.
Portfolio 1 (high market value) 0,005288
Portfolio 2 0,007548
Portfolio 3 0,015978
Portfolio 4 0,024614
Portfolio 5 0,018236
Portfolio 6 0,017704
Portfolio 7 (low market value) -0,01059
From these numbers, there is obviously no distinguishable relationship between company size and financial success for Swedish companies, whether measured by the Sharpe ratio or by the Jensen’s Alpha. Indeed, the results do seem quite uncertain.
4.1.4 Liquidity
Company liquidity does exhibit somewhat of a positive pattern to the returns measured in the Sharpe ratio. This is highlighted by a recorded positive return from the portfolio with the high
liquidity companies and a negative return from the portfolio consisting of low liquidity companies. This positive relationship is not clear-cut however. Regard the following table.
Table 7
This table presents the average returns (provided by the Sharpe Ratio) of seven different portfolios which are ranked according to their rate of liquidity. Portfolios range from 1995- 2005 and are reformed every quarter.
Portfolio 1 (high degree of liquidity) 0,023
Portfolio 2 -0,10
Portfolio 3 0,007
Portfolio 4 -0,007
Portfolio 5 -0,03
Portfolio 6 -0,016
Portfolio 7 (low degree of liquidity) -0,057
Turning now to the latter part of the time period, it exhibits a similar pattern. Even though portfolio number two records a surprising result both with regards to the Sharpe ratio and to the Jensen’s Alpha, the same pattern emerges. Portfolios consisting of companies with larger degrees of liquidity tend to outperform low-liquidity companies. The resuults from the Jensen Alpha analysis is presented in table 8.
Table 8
This table presents the average returns (provided by Jensen’s Alpha) of seven different portfolios which are ranked according to their liquidity. Portfolios range from 2000-2005 and are reformed every quarter.
Portfolio 1 (high degree of liquidity) 0,059775
Portfolio 2 -0,01578
Portfolio 3 0,025279
Portfolio 4 0,020152
Portfolio 5 0,010211
Portfolio 6 -0,01637
Portfolio 7 (low degree of liquidity) -0,00449
With regards to previous research, Johnson and Soenen (2003) do not find any significant relationships between the level of liquidity of the companies and their financial performance.
Somewhat surprisingly, their study also records negative relationships between liquidity and the performance metrics. At the same time, however, their analysis does record poor p-values for this variable with 0,87 for the Jensen’s Alpha and 0,32 for the Sharpe ratio. Therefore it will be interesting to see if the positive relationship discerned from the Swedish companies will hold in the parametric analysis that will follow.
4.1.5 Capital structure and the cash conversion cycle
These two variables do not record any discernable nor apparent pattern in relation to the performance measures. In a similar vein as with the analysis on company size these variables do not seem to exhibit any clear-cut correlation to the Swedish companies’ Sharpe ratios or their Jensen’s Alphas. Therefore, the portfolio analyses of these variables have been placed in Appendix B.
4.2 Parametric analysis
In order to provide further evidence of the relationship between the chosen indicator variables and financial success an OLS regression is performed. The following table presents the results of the time period 1995-2005 where the indicator variables are regressed on the corresponding Sharpe ratios for the companies.
Sharpe Ratio regression Coefficients Standard Error t- ratio p-value
Constant -0,313170942 0,131216669 -2,38667 0,017115
Market Value -1,66397E-07 3,51313E-07 -0,47364 0,635817 Book-to-Market ratio 0,140010717 0,043623171 3,209549 0,001355 Return on Assets 2,621615564 0,42608279 6,152831 9,56E-10
Liquidity 0,226817826 0,155333559 1,460198 0,144428
Capital Structure -0,281748743 0,219213232 -1,28527 0,19888 Cash Conversion Cycle -0,000340069 0,000718532 -0,47328 0,636075
The regression shows a particularly strong relationship between the ROA and the Sharpe ratio. This relationship is also statistically significant and supports the conclusions drawn
from the portfolio analysis performed earlier. In addition, the book-to-market ratio records a positive relationship to the Sharpe ratio albeit smaller than the relationship recorded by the ROA. This observation, being statistically significant, is also one which further strengthens the evidence that was presented by the portfolio analysis. In accordance with the portfolio analysis, company liquidity records a positive relationship to the Sharpe ratio. This relationship is not statistically significant, however, with a p-value of 0,144.
These observations, with a positive relationship between the ROA and the book-to-market value are also observable in the results from the Jensen’s Alpha regression (2000-2005).
However, the positive relationship between the Jensen’s Alphas and the book-to-market ratio is not statistically significant until the 93% level. The complete set of coefficients is presented below.
Jensen Alpha regression Coefficients
Standard
Error t-ratio p-value
Constant -0,07551 0,028639 -2,63648 0,008533
Market Value -8,4E-09 6,38E-08 -0,13178 0,895188
Book-to-Market ratio 0,016199 0,00884 1,832604 0,067218
Return on Assets 0,407863 0,085277 4,782801 2,04E-06
Liquidity 0,088285 0,033939 2,60131 0,009452
Capital Structure -0,03745 0,049204 -0,76107 0,446833 Cash Conversion Cycle 0,000209 0,000174 1,198744 0,230969
Aside from the remarks made in relation to the book-to-market ratio and the ROA measure, a couple of interesting observations can be made from these two regression analyses. First, both regressions record a negative relationship between size and financial success but this relationship is highly insignificant (statistically speaking) in both instances. This is indeed quite a surprising result, since size has recorded statistically significant relationships in much previous research. Second, with regards to the Jensen’s Alpha analysis for 2000-2005 the role of company liquidity exhibits a positive and statistically significant relationship. This is in line with the results from the portfolio analysis performed earlier. Evidently, the shorter time period studied for the Jensen’s Alpha provides for more precise parametric analysis.
Moreover, the variable of capital structure and the cash conversion cycle does not generate any significant relationships to neither the Sharpe ratio nor the Jensen’s Alpha. This result
was anticipated due to the quite inconclusive evidence from the portfolio analyses performed on these variables.
4.2.1 Reliability
In order to ascertain the reliability of the parametric analysis some tests are necessary. An aspect which is important, especially regarding time series analysis, is that of autocorrelation.
We can get an approximation of the degree of autocorrelation by looking at the Durbin- Watson statistic. A rule of thumb is that values near 2 indicate that the null hypothesis of no autocorrelation cannot be rejected. If that statistic is close to the number 2 we therefore assume that no autocorrelation is present. In this study the Sharpe ratio regression records a Durbin-Watson statistic of 2,00 whereas the Jensen’s Alpha regression displays 1,86 for the same statistic. This suggests that there is no autocorrelation present, a conclusion which is further strengthened by looking at the residual plots for the two regression analyses. These plotted residuals are found in appendix d and do not display any signs of autocorrelation.
Furthermore, it is important to determine that the independent variables in the study are not severely correlated to each other, i.e. that the study does not suffer from multicolinearity.
Thus, a correlation matrix for the X-variables was incorporated and the conclusion is that no disturbing degree of multicolinearity can be distinguished. These numbers are found in appendix c.
Finally a test of heteroskedasticity is necessary. Heteroskedasticity refers to unequal variance in the regression errors which might upset the results of the regression analysis. The most used method for testing for heteroskedasticity is the test labelled as White’s test.
Consequently, such a test was performed for the two regression analyses. The Sharpe ratio regression posted a p-value of 0,79 and the Jensen’s Alpha regression got a p-value of 0,11.
These findings suggest that no heteroskedasticity is present in the regression analyses.
The conclusion from these three tests is that the statistical reliability of the parametric analysis is quite robust. There is no evidence of multicolinearity, autocorrelation, or heteroskedasticity.
Therefore, the results from the regression analyses should indeed be reliable. On a final note, however, it should be noted that the two models both exhibit quite poor R-square and adjusted R-square values, reported in appendix c. Nevertheless, since Johnson and Soenen (2003) do not present the R-square values for their investigation, there is no point of comparison.
Studies of this kind might in fact be characterised by low R-square values. Still, the relatively poor fit of the models seems a bit surprising.
Chapter 5. Concluding remarks
The search for winning bets on the stock market has generated huge interest within the research community. Both academics and practitioners alike are preoccupied with finding models that would accurately predict stock returns. Evidently, much remains to be done before any assertive models are found. Still, being able to define the characteristics of companies that beat the overall performance of the stock market is a major driving force of financial economists. Obviously, the quest for correct indicators of financial success is also fraught with many obstacles and the solutions are not always as straightforward as one would have hoped for.
This study has analysed the relationship between certain specific financial indicators and financial success of Swedish companies. The indicator variables were chosen from previous studies and the predictive power of these variables in terms of generating returns to shareholders was examined. In terms of measuring financial success, the Sharpe ratio and the Jensen’s Alpha were used. In order to determine the relationships between financial success and the financial indicator both a method of portfolio analyses and a parametric method were used.
The results point to the fact that profitable companies with a high book-to-market ratio are successful in generating excess returns to shareholders. In addition, the liquidity of the companies seems to be a good indicator of financial success, particularly in recent years (2000-2005). Contrary to previous studies, the size of the companies does not present any clear-cut nor statistically significant relationship to stock returns. Evidently, it is always necessary to keep in mind that we are comparing this relatively modest study of 42 Swedish companies to studies that analyse other markets (predominantly the American stock market) which are characterised by other types of companies. It should therefore not be surprising that somewhat diverging results might emerge, which is also the case in respect to contemporary research. The fact that no statistical significance is reached with regards to company size is still surprising however.
Nevertheless, this study has hopefully provided some insights into the existing literature on which are the indicators of financial success. Furthermore, the application of the model onto the Swedish stock market provides many thought-stimulating topics for discussion. As was
evidenced in the preceding discussion, there is always a difficulty when dealing with time series that stretch over a long period of time. Such time series are always susceptible to historical occurrences that might upset the results. At the same time, such instances also provide many opportunities to remodel the investigation and find ideas to new research. Such research will undoubtedly be performed and the quest for successful stock picks will go on.
This paper constitutes a humble addition to contemporary research within this field.
5.1 Suggestions for future research
As previously noted, this field of study has generated enormous interest in recent years.
Nevertheless, it would undoubtedly be interesting to see further elaborations on the models used within existing research. It could be contemplated to use other measures of both the explanatory and the explaining variables. Concentrating on a shorter time period and using a larger sample of companies could also be worthwhile. Further expansions of the statistical methods used could also be considered.
