Beta is a widely accepted measure of systematic risk and is used by practitioners for capital budgeting, portfolio formation, and performance evaluation. It is important to know whether beta is stationary overtime and to identify the factors that may help forecast it. Since beta represents co-movement with the market, it is pertinent to check if betas of individual stocks and portfolios change over bull and bear market phases. Regression analysis, paired t-tests, and correlation analysis were used to study beta for 158 stocks and 15 portfolios in the Indian stock market over 12 years from 1991 to 2002. Regression analysis indicates that alpha and beta were not significantly different for majority of the individual stocks and portfolios during the alternating market phases. Paired t-tests, on the other hand, shows evidence of non-stationarity in some of the alternating periods and stationarity between all bull periods. Analysis of the control groups also reveal that the overall period was not stationary. Although correlation between pairs of periods was fairly high and significant in some cases, it was not consistent group-wise and was at odds with the t-tests in some cases. Overall, the evidence of non-stationarity of beta between bull and bear periods was not consistent and cannot be put to practical use.

It is used to construct portfolios, measure the performance of investment managers, develop project screening rates for capital budgeting, and value companies. An essential prerequisite for using beta for decision making is a reasonable degree of stationarity and predictability overtime. The objective of this study is to examine the effect of market phases on the stationarity of beta in the Indian stock market from 1991 to 2002. This study briefly reviews the relevant literature and presents evidences, which show that the effect of alternating bull and bear market phases on beta is not consistent and significant, and therefore beta cannot be put to practical use.

Beta Stocks are subject to systematic risks, which are common to all, and unsystematic risks which are industry- or firm-specific. Systematic risk includes economic and political conditions, which are expected to affect all stocks in a similar fashion, though the extent may differ. Unsystematic risk includes factors such as government policies regarding a particular industry, labor availability, the financial structure of a firm, etc. Unsystematic risk is considered as diversifiable through the creation of portfolios. According to Gooding and O’Malley (1977), stationarity refers to the absence of period to period fluctuation in beta, while stability is concerned with fluctuations in beta caused by varying the differencing intervals (weeks, months) in calculating returns.

Research on stationarity of beta has covered various aspects. Blume (1971) studied stationarity using beta sorted portfolios; he found correlation between periods increased with the size of portfolios. Baesel (1974) studied the risk class changes using a simple transition matrix and found that the stationarity increases with increase in period length. Vasicek (1973) suggested a Bayesian approach to the adjustment of security and portfolio betas. Blume (1975) examined in detail the tendencies of betas to regress towards the mean over time. Scholes and Williams (1977) highlighted the problem of non-synchronous trading for calculation of beta. Fabozzi and Francis (1977) investigated the effect of bull and bear markets on alpha and beta. A modified version of the single index market model was tested using alternative definitions of bull and bear market conditions on monthly returns of 700 stocks on the NYSE from 1966 to 1971. They found no significant effect of the alternating forces of bull and bear markets. Gooding and O’Malley (1977) studied monthly returns for portfolios created from 200 of the largest US stocks from November 1966 to 1974 and used correlation analysis and paired t-tests to check the stationarity of beta. They found that beta coefficients were influenced by major market trends.

IUP Journal of Applied Finance, Beta Stationarity , Bull and Bear Markets , Portfolio Management, Regression analysis, correlation analysis, Capital Asset Pricing Model , CAPM, Bombay Stock Exchange, BSX, National Stock Exchange, NSX, New York Stock Exchange, O'Malley, Empirical Testing of Capital Asset Pricing Model, Financial Management and Analysis.