Empirical evidence of stock return predictability obtained by financial ratios or macroeconomic factors has received substantial attention and remains a controversial topic to date. This is no surprise given that the existence of return predictability is not only of interest to practitioners but also introduces severe implications for financial models of risk and return. Founded on the assumption of efficient capital markets, research on capital asset pricing models has instigated this emergence of stock return predictability factors. Analysing these factors categorically, this paper will provide a balanced discussion of advocates as well as sceptics of stock return predictability. This essay will commence by firstly outlining the fundamental assumptions of an efficient capital market and its implications for return predictability. Subsequently, a thorough focus will be placed on the most significant predictability factors, including fundamental financial ratios and macroeconomic indicators as well as the validity of sampling methods used to attain return forecasts. Lastly this essay will reflect on the findings while proposing areas of further research.
Predictability and Efficient Capital Markets
It is often argued that if stock markets are considered efficient - i.e. that all security prices fully reflect all available information (Fama, 1991) - it is impossible to predict stock returns. Indeed, some researchers have even equated stock market efficiency with the non- predictability property. Originating in Samuelson’s (1965) random walk theory of asset prices of 1960, the efficient market hypothesis (EMH) proves that in an informationally efficient market, price changes must be unpredictable. Nonetheless, this hypothesis was firstly fully articulated by Fama (1970) in his influental review of „Efficient Capital Markets“. In fact, he distinguishes between three different forms of EMH - namely the “weak form“, “semi strong form“ and “strong form“ - each increasing in levels of available information for investors. At the core of the EMH lie three basic premises: (1) investor rationality - assuming that all investors act rationally - (2) arbitrage - assuming investment decisions satisfy arbitrage conditions - and (3) collective rationality - assuming random errors of investors cancel out in the market.
The assertion that prices fully reflect all available information remains heavily criticised,
particularly by Grossman and Stiglitz (1980), who point out that there must be “sufficient profit opportunities, i.e. inefficiencies, to compensate investors for the cost of trading and information-gathering“. Accordingly, Fama (1991) recognized that a weaker and economically more sensible version of the efficiency hypothesis was needed. In light of these difficulties, several advocates of the EMH have defined an efficient financial market as one that does not allow investors to earn above-average returns without accepting above-average risks (Malkiel, 2003).
In the financial industry, it is generally recognised that markets are predominantly efficient (Malkiel, 2003). However, inefficiencies can arise during periods of important institutional and technological changes (Persaran, 2005). An appropriate example is the valuation errors made during the 1999-2000 internet bubble. Such inefficiencies are hardly predictable but do occur occasionally. Nonetheless, one could argue that markets are still efficient even if many participants act quite irrationally, if efficient markets mean that investors do not earn aboveaverage returns without accepting above-average risks.
Market efficiency research has primarily emerged out of tests directed at asset pricing models, particularly Sharpe’s (1964) Capital Asset Pricing Model (CAPM). Findings show that the market beta itself is insufficient to describe the cross section of expected returns. Many of the most significant empirical anomalies in finance come out of tests directed at asset-pricing models. Depending on the desired emphasis, asset-pricing models are tested conditional on efficiency (Fama, 1991). This joint hypothesis problem questions whether such anomalies result from misspecified asset-pricing models or market inefficiency.
Stemming from the controversy about market efficiency there has been a variety of stock return predictability factors that have been put forth by researchers. Early work on return predictability focused on forecasting returns from past returns, using time series analysis. Former papers such as Jegadeesh (1990) and DeBondt and Thaler (1985) find that both short (less than one month) and long-term (three-to-five year) past returns are inversely related to future average returns. However, findings show that in intermediate horizons past returns are positively related to future average returns (Jegadeesh and Titman, 1993). Attempts at explaining this relationship include microstructure and data snooping biases, rational risk- based explanations (Conrad and Kaul, 1998), and irrational behavioural stories (DeBondt and Thaler, 1985). Nonetheless, the variation in results and the difficulties to reconcile theory with the exceptional profits generated by trading strategies that exploit past return patterns, cast doubt on the significance of past returns as a predictive power (Moskowitz and Grinblatt, 1999). It is for these reasons that the following section will focus on more recent and compelling predictability variables such as fundamental financial ratios and macroeconomic factors.
As a potential indicator of market valuations, the Price-to-Earnings ratio has extensively been used by both practitioners and academics to examine the predictability of real stock prices. In particular, Fama and French (1988), Keim (1990) and Bali et al. (2008) draw attention to the significant explanatory power of the Price-to-Earnings Ratio, which is identified as:
Price − to − earningst = Market price per sharet /Earnings per sharet
In general, the Price-to-Earnings ratio measures how much investors are prepared to pay per dollar of earnings generated by the company. Accordingly, a high P/E ratio suggests that investors expect higher growth in the future. Given this underlying level of investors’ confidence, numerous academics have attempted to elucidate the ratio’s strength of predictability regarding future stock returns.
Investigating the inverse of the P/E ratio, i.e. Earnings Yield, Shiller (1984) and Fama and French (1988) estimated regressions of returns on the lagged dividend yield or the lagged earnings yield, and concluded that both have explanatory power, but that the dividend yield has more. Fama and French (1988) account for these results by reasoning that earnings are more variable than dividends and, therefore, that the earnings yield ratio is a noisier measure of expected returns than dividend yield. On the contrary, Bali et al. (2008) used a portfolio of 48 industries to identify that earnings yield has significant explanatory power for the timeseries and cross-sectional variation in firm-level stock return. In fact, they attribute this forecasting power of earnings yield to the mean reversion of stock prices as well as the earnings’ correlation with expected stock returns.
In accordance with Bali et al. (2008), the American investment management company Vanguard (2012) asserts that valuation metrics like the price/earnings ratio have an inverse or mean-reverting relationship with future stock market returns, though meaningful only at long horizons. Nonetheless, Vanguard (2012: 2) cautions that the P/E ratio leaves a considerable portion of stock returns unexplained and, thus, advises that „expected stock returns are best stated in a probabilistic framework, not as a point forecast, and should not be forecast over short horizons. “Furthermore, alluding to the common notion that stock prices rise and decline in advance of corresponding changes in earnings, Beaver and Morse (1978) present empirical evidence that P/E ratios predict future earnings changes. Interestingly, Keim (1990) expands this hypothesis and contends that the strength of stock return predictability increases as firm size decreases and earnings yield increases. Employing a sample period from 1951 to 1986, Keim (1990) uncovers that, on one hand, both E/P and size effects are significant when estimated across all months, on the other hand, however, the influence of firm size and the earnings/price ratio on stock returns remains controversial.
A number of studies appear to show empirical evidence that the dividend-price-ratio, i.e. dividend yield (DY), can be used to estimate future stock returns, for example Rozeff (1984), Campell and Shiller (1988) and Fama and French (1988). Most of these studies used monthly data and defined the total return as:
Abbildung in dieser Leseprobe nicht enthalten.
R = Total Return P = Stock Price d = Dividends t = Time
According to Fama and French (1988), the basic idea is that stock prices are high relative to dividends when discount rates and expected returns are low and vice versa. Hence, dividend to price fluctuates with expected returns. They show that the forecasting power of the dividend yield increases in accordance with the return horizon. Furthermore, dividend yield reveals variation in return (Kothari and Shanken, 1997) and can forecast future return in 36 international markets (Choudhury, 2003). To explain the predictive power of dividend yield, Lewellen (2004), introduced a new test to improve the forecasting ability of financial ratios especially DY during 55 years. As many other studies, he used a simple regression model: