Financial theory describes risk assessment as one of the most important part in an investment decision making process. However, for a risk to be known, it is important for investors to interpret information flowing on the market. This study aims to examine the association between accounting information and the market risk over time. It also evaluates how far the beta value and accounting variables can be useful for investors in Mauritius. Beta estimates are calculated using Capital asset pricing model and accounting risk variables are derived from theoretical foundations and prior empirical findings. The relationship between the financial ratios and the level of systematic risk is obtained by regressing the variation in the beta against changes in the accounting variable.
The empirical evidence shows that beta is valid on the Stock Exchange of Mauritius (SEM). However, the power of beta is relatively low in capturing the systematic risk. This finding is in line with Campbell (1995) who obtained similar observation for emerging equity market and with Bundoo (2000) who noted same result. Finally the result shows that a strong association exist between accounting variables and market risk and it also observed that this relationship is consistent over time. Accounting variables like growth rate, debt ratio, asset size, liquidity, profit margin and accounting beta are able to capture market risk where beta generally provides a high explanatory power of systematic risk. The findings contradict the some of the association between the market risk measures and accounting risk measure obtained Beaver et al (1979).
The growth experienced in the Stock Exchange of Mauritius (SEM) during the years 1989 to 2007 was with no precedence. Stock prices of quoted companies on the SEM boomed, causing a high influx of capital which caused the market to rise to its peak with a net market capitalisation of MUR 173 billion in the end of the financial year 2007. Local investors who had investments in fixed deposits from local commercial banks shifted some of their investments to the SEM, with view of higher return. But Stock prices started to fall soon after the end of the month of February 2008 and within a year the SEMDEX reached a position which was a low as the values experienced in September 2006.
While this fall was largely attributed to the morose international situation, as a result of the international financial crisis; there is also the question whether the SEM effectively capture risk which is inherent by companies quoted and how far investors in Mauritius used the publish financial information to evaluate and predict the level of risk in the operating environment.
Financial markets serve a key purpose in an economy by allocating productive resources among various areas so as to enable an efficient resource allocation, across different firms, investors assess the security and market expected prospects and risks and form a portfolio of investments based on their assessment. Security analysis usually involves an evaluation of the financial position and performance obtained from the financial statements published periodically by companies. In an efficient financial market the share prices is expected change to the fair value of the firm as new information flows into the market.
Financial theory describes risk assessment as one of the most important part in an investment decision making process. The return of a stock is often considered to be narrowly related with the risk which the investor is taking while holding that stock. This makes the generally accepted principle that the higher is the risk in investing in an asset, the higher should be the asset’s expected return. This implies that there is a positive correlation between risk and expected return in holding a stock.
1.1 Problem Statement
The analysis of stocks return is intricately linked with the analysis of risk. Empirical studies carried by Graham et al (2001) has shown that the Capital Asset Pricing Model (CAPM), (an asset pricing tool which uses risk as a basis to calculate assets return) is used, by more than seventy five percent of the chief financial officers, as primary tools in the portfolio selection process. However some authors in the capital markets literature (Campbell (1995) and Chan et al (1991)) have argued that in the case of emerging stock exchanges the CAPM is inapplicable and beta is not significant.
However, for a risk to be known, it is important for investors to interpret information flowing to the market. Fama (1963) described three generic forms of market efficiency based on the market reaction to inflow of information. Markets which react to all past information are said to be in its weak form, those markets which react to all past and publicly available information are referred to as semi-strong efficient markets and those which react to all past, public and private information are considered as strongly efficient markets. A study made by Bundoo (2008) showed that Stock Exchange of Mauritius (SEM) has the characteristics of a market in its weak form. This implies that the SEM effectively responds to past information. Yet there is absence of empirical research which evaluates whether market return and risk are effectively pictured through accounting ratios.
1.2 Aims and objectives
This paper aims at analysing the share prices in the SEM and key accounting ratios to evaluate the financial position, performance of a sample of companies quoted across various economic sectors of the SEM with the view of answering the above question. It also seeks to test whether investors can trust beta in their decision-making process on the SEM.
The paper also aims at:
understanding the relationship between the financial ratios, market return and risk; estimating the level of systematic for different business segment where financial market information is not available; and to guide investment in measuring the systematic in private and non listed companies in Mauritius.
1.3 Organisation of this paper
The paper is organised as follows: Chapter 2 provides a summary of literatures concerning risk measures, accounting tools and market-based models to measure the performance and risk; It also surveys the empirical researches on the SEM and similar markets; Chapter 3 develops the models which are to be used in the analysis of the relationship between systematic risk and accounting ratios; It also outline the methodology and sample data which is used in the analysis; Chapter 4 presents the key findings from the study and Chapter 5 concludes the paper.
2 Literature review
Risk and return of a firm are the two most important factors in the development of financial strategy for both individual investors and firms. Risk is inherently multi-dimensional and as such it has multiple characteristics which may be classified as financial and non financial. These characteristics make up the risk profile of a security, which is generally observed as changing with time and at different levels of a market. These changes in turn, impact on the return of the investors either by creating value or destroying the initial value before the investment.
Modern financial theories have proposed different models which are founded on sound theoretical analysis which can be used to estimate the different degree of riskiness of a particular security. These risk measures are then used in valuation models to estimate the return which an investor, with a defined risk attitude, can expect from an investment. As described in chapter 1, above, the applicability of such financial theories remain untested in many emerging markets.
This chapter reviews the financial models which are commonly used by practitioners for estimating of the risk of stocks and stock market and their corresponding returns. It also summarises the main financial ratios which are used to analyse the financial risk, financial performance and the value of the firm. Finally a summary of the accounting tools and market-based models to measure return is also presented.
It has always been difficult for practitioners to reach a consensus on the definition of risk. Moles (2004), nevertheless, provides a simple definition which is taken in this paper as basis for risk measurement. He defines risk as “the chance (or probability) of a deviation from an anticipated outcome”. With this definition it is implied that risk is made up of at least these 3 elements:
1. probability: which means that risk can be quantified and expressed as a parameter, number of value;
2. deviation from anticipated outcome: which is extent to which the actual result may deviate from that which is expected;
3. anticipated outcome: this means that it is the consequence of the actual results deviating from the expected results that leads to risk. Newbold et al (2003) states that probability can be measured using past data by considering the proportion of times that an event occurred. For the case of an investor the anticipated event would be the financial return which he or she can expect by holding an asset. The measurement of the deviation from the anticipated return is normally done using the standard deviation of returns generated by an asset with regard to the expected return.
2.1.1 Systematic and unsystematic risks
The deviation from the anticipated return is caused by is explained by 2 levels of risk: systematic risk and unsystematic risk. The sum of these two main categories of risk is the total risk to which an investor is exposed to.
Systematic risk is associated with overall movements in the general market or economy and therefore is often referred to as the market risk. The market risk is the component of the total risk that cannot be eliminated through portfolio diversification.
Unsystematic risk which is a component of the portfolio risk that can be eliminated by increasing the portfolio size, the reason being that risks that are specific to an individual security such as business or financial risk can be eliminated by constructing a well-diversified portfolio.
2.2 The Capital asset pricing model
Markowitz (1952) constructed a mean-variance model to observe the trade-off between risks and return. The model mathematically proved that return can be maximised, while minimising the overall risk, by holding a diversified portfolio. The idea was based on the concept that securities that are inversely correlated or having coefficients which are less than one. Such negative or low correlation coefficient results in a low covariance between securities in the portfolio. The low covariance implies a comparatively low level risk. However, Sing et al, (2001) observed that the model ignore the general risk-averse attitude of most investors.
The Capital Asset Pricing Model (CAPM), developed by Sharpe (1964), is based on the framework set out by Markowitz (1952) which considers that investors invest their money in a portfolio of assets. The CAPM states that the return which a risk averse can expect from investing in a risky asset is a risk premium over the risk free rate. The formula 1 below states the formula which can be used to calculate the expected return.
E(Ri) = Rf + i ( E(Rm) – Rf ) (2.1)
E(Ri) – expected rate return of stock I;
i – relative risk of share I;
E(Rm) – expected rate return of the market portfolio; and
Rf – risk-free interest rate.
Sharpe (1964) and Lintner (1965) explained that the correct measure of risk of an asset is its beta factor, a standardised measure of the systematic risk and that the risk premium per unit of riskiness is the same across all assets.
CAPM has been developed by considering some assumptions such as normal distribution of assets return, perfect divisibility of assets and return, the existence of a risk free rate, perfect market conditions, inter alia, which might not exist in the real world. Despite the fact that most of the above assumptions are neither valid nor fulfilled, the CAPM has become an important tool in finance. It is widely used by finance practitioners for assessment of cost of capital, portfolio performance, portfolio diversification, valuing investments and choosing portfolio strategy among others.
The β factor in the equation 2.1 measures the volatility of the specific asset with regard to the volatility in the market, that is, the market risk. Mathematically it is expressed as in equation 2, below:
systematic_riskasset = covariance of the asset and that of the market
market_risk is the volatility in the market portfolio, it is measured by the standard deviation of prices of the market portfolio.
2.2.1 Empirical review of Capital asset pricing model
The empirical studies undertaken by Jensen et al. (1972) found supportive evidence for CAPM. The authors found that the actual return, for a sample of companies quoted on the New York Stock Exchange (NYSE), were consistent with the predictions of the CAPM. They noted that the relationship between the average return and beta was very close to a linear one and that portfolios with high betas had high average returns. The same result was confirmed by Black et al. (1972), who studied of all the stocks on the NYSE over the period 1931-1965.
Black et al. (1972) formed portfolios of stocks and analysed the abnormal return with regard to the beta factor, and found a linear relationship between the average excess portfolio return and the beta. Black et al (1972) observed that the beta factor measured the responsiveness of the share return to changes in the returns of the market. Stocks with high positive betas had stock price which rose faster than the market. This implies that high beta stocks bear a higher degree of risk compared to stocks which have their beta factor as negative. Stock with negative beta behave negatively to changes in the market, as such, in a bearish market, it is more attractive to invest in these stocks as it helps to preserve the value of the investor.
Fama et al. (1973) also observed a larger intercept than the risk-free rate when analyzing the return against risk. They confirmed that there is a linear relationship between the average return and the beta, even over longer period. They further investigated whether the squared value of the beta and the volatility of assets returns explained the residual variation in the average returns across asset and found that, in addition to portfolio risk, there are other variables that affect expected return.
2.2.2 Critics against Capital asset pricing model
There has been also several criticism of the applicability of the CAPM in many markets. Empirical research undertaken by Basu (1977) proposed other factors which have to be considered instead of relying wholly on a single variable, beta. According to Basu (1977) the price earnings ratio has a great influence in market return. Banz (1981) challenged the model by indicating that firm size have a considerable impact on the average returns of a particular stock and thus firm size could better explain the volatility than the market beta.
The author observed that the average return of small firms were higher than the average returns on stocks of large firms. Chan et al (1991) made a further observation, on the Japanese market, that stocks with high ratios of book value of common equity have significantly higher returns than stocks with low book to market equity. In this respect, book to market equity started to be regarded as being an important variable that could produce dispersion in average returns.
Fama and French (1992) came up with the conclusion that a more realistic approach of the risk in the market is the multi-index models. Their study concluded the findings of Basu(1977), Stattman (1980), Banz (1981) and Chan et al (1991) who argued that size of the firm and the books to market equity ratio are far superior in explaining asset returns.
In contrast with CAPM which can be considered as a single factor model, Ross (1976) proposed a multifactor arbitrage pricing theory (APT). Groenewold et al (1997) examined the validity of the model for Australian data and compared the performance of the empirical version of the APT and the CAPM. They concluded that APT outperforms the CAPM in terms of within-sample explanatory power. The APT, however, is a generic model and does not specify any factor which has to be considered in analysing return with regard to risk.
2.2.3 The ongoing debate on the applicability of Capital asset pricing model
Nevertheless, there is no consensus in favour of CAPM due to the disparities in the empirical findings and the debate continues. In general, the studies challenge the data used by Fama et al (1993). Kothari et al (1995) argue that the findings of Fama et al (1993) depend essentially on how the statistical findings are interpreted.
Amihudm et al (1992) and Black (1993) supported the idea that the data are too noisy to invalidate the CAPM and showed that when a more efficient statistical model is used, the relationship between average return and beta is positive and significant. The author further suggested the findings in respect of size effect could be simply in a sample period effect and that it may not be noted in another period.
Similarly, Berk (1995) questioned the findings of Chan and Chen (1991). The author emphasised that stock prices (and market value of the equity (MVE)) depend on the expected future cash flows which is used by investor to estimate the risk and the required rate of return. Therefore, if two companies have a higher discount rate and consequently its price and MVE will be lower. In this sense, MVE captures the information about the company’s risk, since any change in investors’ perceptions of risk is immediately reflected in the stock prices.
Furthermore, when the expected return of a firm is defined as the expected cash flow divided by its MVE, the relationship between MVE and return is clearly negative for companies with equivalent cash flows. Berk concludes that for companies of similar cash flows, the higher the risk of the cash flow, the higher the discount rate investors apply to it, which causes price to decrease and expected return to increase. This concept has contradicted the findings of Chan and al (1991), which attribute higher returns to smaller companies.
Owing to its intuitive appeal, the CAPM has become an important tool in finance for assessment of cost of capital, portfolio performance, portfolio diversification, valuing investments and choosing portfolio strategy among others. However, there is no consensus in the literature as to what a suitable measure of risk is, and consequently, as to what is a suitable measure for evaluating risk-adjusted performance (Galagedera, 2007). As such, the debate for robust asset pricing models continues. Other studies (Ball and Brown (1969) and Beaver, et al (1970)) have focussed on accounting variable to convey information about the market risk.
2.3 Accounting variables as a measure of systematic risk
Research in accounting variable as a measure of risk has increased considerably since the last forty years with a number of published papers by Beaver et al (1970), Lev et al (1974) , Bernard (1989), Ohlson (1995), and Kothari (2001). Beta measures the relative risk whereby risk itself is determined by some combination of firm characteristics, market conditions, and the sensitivity of the firm stock to market conditions. As such, understanding the relationship between the accounting variable and the systematic risk can provide an alternative basis to a market based estimation and prediction which will in turn guide the accounting policy formulation and investment decision making (Brimble et al, 2007).
The study by Beaver et al (1970) was the most quoted research in accounting and financial research. The author had improved the perdition of systematic risk by considering the firm specific characteristic and they identified significant association between market risk and firm specific accounting information.
The financial statements of firms were mostly used in providing considerable information that could be used to measure the inherent risk. In fact, the Financial Accounting Standards Board (1983) stated that the objective of financial reporting is to provide information that is useful to present and potential investors and creditors and other users in making rational investment, credit, and similar decisions.
A number of studies investigated how financial information becomes impounded in security prices and affects investment decisions. These accounting data are converted into the financial constructs, such as growth, operating leverage, profitability, liquidity, and efficiency. There is considerable evidence that since the late 1800’s ratio analysis has been widely used in the valuation of published financial data (Connor, 1973). Researchers and investors use mainly financial ratios for risk modelling purposes based on different criteria of comparison which are discussed as follows:
Time series analysis: It also known as trend analysis and it is used to compare financial ratios over a period of time. Ratio analysis for one year may not present an accurate picture of the firm (Rao, 1989). As such, to appraise a firm’s performance, the present ratios need to be compared with the past ratios.
Cross-sectional analysis: This method compares ratios of one firm to the ratios of some other selected firms operating in the same industry at the same point in time (Pandey, 1999). Such comparison indicates the comparative financial position and performance of the particular firm.
Industry analysis: According to Pandey this type of analysis helps to ascertain the firm’s financial standings and capacity vis-à-vis other firms in the same industry. A study conducted by Beneda (2006) indicated that commercial lenders often consider the use of industry ratio analysis to be critical with regard to the potential success of the business. The main shortcoming of this analysis is that it is difficult to obtain the average ratio of an industry and if available the average ratio is composed of both strong and weak firms.
Financial ratios were used for locating possible takeovers and mostly to predict major events such as corporate failures (Scott, 2004). Other studies reported on an association between accounting ratios and market risk measures, and proposed that certain accounting ratios can be used as proxies in predicting future security (Beaver et al. 1970; Elgers and Murray, 1982).
2.3.1 Usefulness of accounting variables
The use accounting as means of estimating the systematic risk will allow the user of the financial statement to assess the investment alternative in terms risk, return and the value of the firms. Ryan (1997) has widely discussed the motive for relating accounting research to measures of market risk:
The volatility of market betas over time indicates that the ex post measure of systematic risk is does not provide meaning full information in estimating the future risk. As such, understanding the relationship between accounting variables and systematic risk could indeed be useful in measuring and predicting the actual and upcoming market risk.
Market based measures of risk, like the capital asset pricing model, fail to consider most of the firm specific characteristic such as the operational factors and environmental contingencies which influence risk. The accounting risk based information gets closer to the identification these economic fundamentals. Therefore accounting model provides an actual risk determinants rather than just determining the level of risk.
Accounting risk model overcome the conventional problem were ex post measure of risk can not be applied due the fact that historical security returns is not available or insufficient like in the case non listed entities and for initial public offering
Accounting variable are not affected by the noise found in traditional risk estimates which rely on past trading histories whereby significant variation in one period subsequently affect the overall risk level ;
The development of trading strategies and the construction of portfolios with the desired level of risk.
2.3.2 Theoretical and empirical review of the relationship between individual accounting variable and systematic risk.
Researchers on the association between systematic risk and accounting ratios were primarily initiated by Beaver (1970). The ratios used by the author were dividend payout, growth rate and leverage ratio, liquidity ratio, variability of earnings and co-variability of earnings. Other studies have further elaborated on these ratios and they also added other accounting based to measure the systematic risk. All these ratios aim at measuring the operating risk, financing risk and growth risk. The theories and empirical finding between these two variables are discussed as follows:
Corporate dividend policy has been the object of lively discussions in finance literature. The debate has revolved around the question of whether companies with generous distribution policies are less risky and whether there exists an optimal payout ratio. Theoretically, it is often asserted that firms with low payout ratios are more risky. This is because that cost for external finance is relatively high for risky firm than firm with low risk. In this respect, risky firms rely on the utilization of their own reserves to carry out business activities.
Dividend payout also affects the systematic risk by the information perceived by variation in the dividend policy. The original idea behind the information content of dividends, was developed by Lintner (1956) who claimed that managers only increased dividends when they believe that the levels of the firm’s earnings have permanently increased. He argued that decrease in dividend may be interpreted as cash flow or liquidity problem. Miller and Modigliani (1961) have argued, on the other hand, that dividend policy is irrelevant to the market value of shares. In a model which disregards taxes, they conclude that the payout policy which the corporation adopts, has no effect on the price of shares. Similarly Watts (1973) and Gonedes (1978) found no evidence that changes in dividend policy contain new information regarding firms’ future earnings.
Gordon (1963) further pointed out that an increase in the proportion of retained profit now means higher cash dividends in the future and therefore conservative dividend policy has no effect on the risk factor. Still, Veikko (1967) explained that the higher the retention rate, the further in the future cash dividends are moved and the greater the uncertainty about their actual amount. Empirical evidence by Edward et al (1998) further showed that a significant negative relationship exists between the dividend pay out ratio and risk element.
Growth affects the systematic risk in two main ways as identified by Beaver et al (1973). Firstly, where a firm earns excessive earning opportunities, that is, where the expected rate is higher than the cost of capital. Growth is normally attained by an expansion in the assets size either through the acquisition of new plants or by creating new product line or by takeovers. The excessive earnings stream derived from these operations is argued to be more uncertain (i.e. volatile) than the “normal” earnings stream of the firm. In this respect the authors stated that a positive association exists between growth rates and risk.
However, Harrigan (1984, 1986) have deepened this analysis and the author has observed different level of association over different industry life cycle characteristics. Harrigan argued that growth strategies, through takeovers and new product development, may be quite risky during an embryonic stage due to the high degree of product, process, and market uncertainty. In contrast, growth strategies may be less risky during times when demand conditions are growing in a stable manner. Finally, growth strategies are expected to become quite risky again as an industry is in transition to maturity because of the cut in the excessive earning streams.
The second argument is related to the logic developed about the dividend payout ratio. Additional capital, utilized in the growth of the firm, would reduce the firm earnings in two main ways. If the expansion in asset is financed by the external debt, the firm earning would be eroded through finance cost. Whereas if the growth is financed through the retained earning, a sharp cut in earning attributable to the shareholder is expected. Both methods will ultimately lead to a reduction in dividend payout and thus increase the systematic risk.
Theoretically, larger firms are less risky than smaller firms. This is because large firms have better access to capital market, management skills and expertise and greater market liquidity. These factors provide opportunities to diversify and to seize new market opportunities to reduce operating risk which will impact on a lower beta than small firms. The studies of Dun et al (1970) reveal that the frequencies of failure are lower for large size firm than firm with low asset capitalization. Horrigan (1966) has shown that the most single important financial statement variable used to predict the bond rating of a firm was total assets.
The author observed that if the asset returns are independent, the variance will decrease in direct proportion to the difference in asset size that is, as firm size doubles, the variance of the rate of return will be cut in half. Empirical work by Alexander (1949) observed that as firm size increase, the volatility in the earning streams decrease accordingly.
Moreover firm with wide operating activities are required to make more disclosure. For example the Mauritian companies act, 2001, stipulate that firms with Turnover above MUR 30 Million are required to file a complete set of financial statements with the Registrar of Companies. This information may be consulted by the members of the public upon payment of a nominal fee. Thus, more information is available to evaluate risk level. Collins et al (1987) have identified that small and recently incorporated firms have a high probability of financial distress.
Research about the association between the market based beta and an accounting beta originated with Ball and Brown (1969). Accounting beta measures the degree of co-variability of firm earnings and the market earnings. Beaver et al (1970) argue that, if beta is being the used as the market determined concept of risk, then the most direct approach would be to compute the beta value on accounting earnings. Bowman (1969) demonstrated that the higher the accounting beta, the higher the systematic risk. Hence a positive relationship is expected between the two variables.
The important relationship between earnings and the market beta is their covariability, accounting beta, is shown in the above. However, the empirical research has generally shown earnings variability to be superior to an accounting beta. Beaver et al (1970) found in a model that use accounting variables to forecast market risk that earnings variability was the most significant variable and that accounting beta did not make a statistically significant contribution.
The relationship established by Ball and Brown (1969) is therefore theoretical. Empirical results may differ from theory for two main reasons as advanced by Bowman (1969). The assumptions (i.e there are only pure equity firms