Serial correlation of stock returns

 

Result and Discussion

 

Direct measures of the serial correlation of stock returns, or equivalent direct measures of the mean and variance of long-horizon returns, depend a lot on the period studied and the econometric method (Cochrane, 1999). Hence, the long-run variance of prices must be the same as the long-run variance of dividends, and this extra piece of information helps to measure the long-run variance of returns. (Cochrane and Sbordone, 1986; Cochrane, 1994; Campbell, Lo, and MacKinlay, 1997)

 

Briefly, there are strategies that result in high average returns without large betas, that is, with no strong tendency for the strategy’s returns to move up and down with the market as a whole. Multifactor models have supplanted the capital asset pricing model (CAPM) in describing these phenomena. Stock and bond returns, once thought to be independent over time, turn out to be predictable at long horizons. All of these phenomena seem to reflect a premium for holding macroeconomic risks associated with the business cycle and for holding assets that do poorly in times of financial distress. They also all reflect the information in prices – high prices lead to low returns and low prices lead to high returns.

 

The world of investment opportunities has also changed. Where once an investor faced a fairly straightforward choice between managed mutual funds, index funds, and relatively expensive trading on his own account, now he must choose among a bewildering variety of fund styles (such as value, growth, small cap, balanced, income, global, emerging market, and convergence), as well as more complex claims of active fund managers with customized styles and strategies, and electronic trading via the Internet.

 

The new academic portfolio advice reacts to the new facts. An investor should hold, in addition to the market portfolio and risk-free bonds, a number of passively managed “style” funds that capture the broad (nondiversifiable) risks common to large numbers of investors. In addition to the overall level of risk aversion, his exposure to or aversion to the various additional risk factors matters as well. For example, an investor who owns a small steel company should shade his investments away from a steel industry portfolio, or cyclical stocks in general; a wealthy investor with no other business or labor income can afford to take on the “value” and other stocks that seem to offer a premium in return for potentially poor performance in times of financial distress. The stock market is a way of transferring risks; those exposed to risks can hedge them by proper investments, and those who are not exposed to risks can earn a premium by taking on risks that others do not wish to shoulder.

 

A substantial amount of research in finance is directed toward understanding why different financial assets earn different returns on average and why a given asset may be expected to earn different returns at different points in time (Jagannathan, Skoulakis and Wang, 2002). Various asset pricing models that explain how prices of financial claims are determined in financial market have been proposed in the literature to address these issues. These models differ from one another due to the nature of the assumptions that they make regarding investor characteristics, that is, preferences, endowments, and information sets; the stochastic process governing the arrival of information in financial markets; and the nature of the transactions technology for exchanging financial and real claims among different agents in the economy. Each asset-pricing model specifies what the expected return on a financial asset should be in terms of observable variables and model parameters at each point in time.

 

Most of the models start by studying the first-order conditions to the optimal consumption, investment, and portfolio choice problem faced by a model investor. That leads to the stochastic discount factor representation of these models. The price assigned by a model to a financial asset equals the conditional expectation of its future payoff multiplied by a model-specific stochastic discount factor. In an informationally efficient market where the econometrician has less information than the model investor, it should not be possible to explain the difference between the market price of a security and the price assigned to it by a model based on information available to the econometrician.

Herings (2003) argued that in realistically calibrated two period general equilibrium models with incomplete markets CAPM-pricing provides a good benchmark for equilibrium prices even when agents are not mean-variance optimizers and returns are not normally distributed. Herings (2003) numerically approximate equilibria for a variety of different specifications for preferences, endowments and dividends and compare the equilibrium prices and portfolio-holdings to the predictions of the CAPM. While the CAPM does not hold exactly for the chosen specification, it turns out that pricing-errors are extremely small.

Merton (1973) and Chang, Hung and Lee (2003) have demonstrated that if an investor anticipates information shifts, he will adjust his portfolio choice today in an attempt to hedge these shifts.

 

Moreover, an econometric methodology is developed to simultaneously estimate the magnitudes of these three portfolio performance evaluation measures. The results show that mutual fund managers are on average with positive security selection and negative market timing ability.

 

Direct measures of the serial correlation of stock returns, or equivalent direct measures of the mean and variance of long-horizon returns, depend a lot on the period studied and the econometric method. Hence, the long-run variance of prices must be the same as the long-run variance of dividends, and this extra piece of information helps to measure the long-run variance of returns. (Cochrane and Sbordone, 1986; Cochrane, 1994; Campbell, Lo, and MacKinlay, 1997)

 

Credit:ivythesis.typepad.com