«Equity Style Investing RONG, WU How to cite: Equity Style Investing, Durham theses, Durham University. RONG, WU (2013) Available at Durham E-Theses ...»
Recently, a number of empirical studies provide the evidence for reversals on the style level. Barberis and Shleifer (2003), for example, propose a theoretical style-level positive feedback trading model in an economy with two types of investors: Switchers (positive feedback traders) and Fundamental Traders (arbitrageurs). They assume that Switchers invest in styles that have performed well in the recent past and their behaviour could trigger style level momentum. In contrast, Fundamental Traders build portfolios by buying recent losers that look cheaper according to the estimated cash flows information. This model would imply that asset returns are less correlated than cash flows. Moreover, when an asset is classified into a style, its correlation with other assets already in that style would increases.
Hence regardless of its cash flow characteristics, when a stock is admitted as a constituent in an index, the underlying stock becomes more correlated with that index. The conclusions of this model are supported by Teo and Woo (2004). Teo and Woo investigate the style effects in the cross-section of stock returns in the U.S. markets and find the evidence for style-level reversals, style-level momentum and positive feedback trading at the style level. Likewise, and perhaps more prominently, Chen and De Bondt (2004) investigate the style momentum payoffs for large U.S. companies in the S&P-500 index over the period 1977-2000. They find that Style momentum effect is distinct from the price and industry momentum, and investors pursuing strategies of buying stocks with past winner characteristics and selling stocks with past loser characteristics could outperform for periods up to one year and possibly longer.
2.6 The cyclicality of style returns and macro cycle Style investing is a common investment strategy advocated by both fundamental and technical investors. But just like other strategies it can suffer during certain investment periods. It is observed that the performance of small size stocks and value stocks go through cycles, and such cycles may not coincide with the overall stock market. The time-variation or cyclical nature of style performance and volatility has raised many interests from both the academics and practitioners.
Studies show that the size premium varies over time or disappears for some periods. Fama and French (1992), Eleswarapu and Reinganum (1993), Dichev (1998), Chan et al. (2000), Horowitz et al.
(2000a, b), Amihud (2002) and Roll (2003), among others, document that the size effect has diminished or cease to exist since 1980s in the U.S. markets. Similarly, Dimson and Marsh (1999), Michou et al.
(2010) show that no size effect is found in the U.K. market in later 1980s. Internationally, Barry et al. (2002) also fail to find the size effect in global emerging markets. Most recently, Fama and French (2012) find that no size premium exist in any of the 4 global markets (i.e. North America, Europe, Japan and Asia Pacific) for 20 years investing period since 1990. A number of other studies also suggest that the size premium demonstrates cyclical nature. For example, Horowitz et al. (2000a) find that the size effect changes over time and it is more pronounced in one period but not for the other or it can even reverse.
While the findings of value premium are relevant from a perspective of the long horizon, over short investment periods the performance of value investing is not reliable and also time-varying (Oertmann (1999)). The tech rally in 1990s and the recent market turmoil in 2007 are perhaps two episodes for the poor performance of value investing (c.f. Owyong (2011). Empirical findings generally suggest that the annual value-growth return pread can vary considerably with respect to both signs and magnitudes (c.f. Arshanapalli et al.
(1998), Lucas et al. (2002)). Oertmann (1999) and Zhang (2005) also find that the U.S. value premium and the volatility of value-growth style investing returns are closely related with market states and business cycles. Likewise, Zhang et al. (2008) establish a strong link between size and value premium with macroeconomic state in the context of U.K. market.
2.7 Time-varying style returns and business cycle variables
The economic interpretations for above mentioned time-variation of equity style returns are twofold. The first focuses on the behaviour of market participants such as noise traders and speculators. There is large literature reporting that speculative trading behaviour causes fads, bubbles or even market crashes. The second explanation relates stock price movements to the macroeconomic fundamentals.
The expected stock returns evolve over time in response to cyclical and structural changes in macro-economy. However, macroeconomic conditions do not affect all stocks in the same manners. Different stocks tend to behave differently in various stages of a business cycle.
For example, consumer staple (known as defensive stocks) generally have inelastic demand and are therefore not much affected by peaks and troughs of the business cycle. There are other stocks, however, can lead the economic cycle and are quite sensitive to the state of the economy. For instance, capital goods yield good performance during the recovery phrase, while luxury stocks generally offer best returns during boom time in the business cycles.
Bolten and Weigand (1998), DeStefano (2004) demonstrate that the determinants of stock value defined by the equity valuation models can possess time-varying patterns related with business cycles.
Indeed, the relative performance of equity styles has been observed to be closely associated with the cyclicality of macro-economy. The rationale behind such divergent performance of style investing stems from the different sensitivity of asset value or return determinants to different business conditions. It is suggested that the returns of small stocks investing is more pronounced during recessions. Similarly, Kwag and Lee (2006) argue that the benefit of value investing is even greater during periods of contraction than expansion. Indeed, value stocks tend to be more sensitive to the cyclical strength of the overall business environment. They generally outperform growth stocks when the macro-economy changes from the sustained period of weakness to transitions into an accelerated recovery period. Conversely, growth stocks are favoured by investors in a slowing economy states and are therefore more likely to be able to beat value stocks when the economy transitions into a period of steady growth or simply begins to weaken.
There is overwhelming evidence to suggest that some business cycle pervasive variables such as the changes in GDP rate, inflation rate, the slope of the yield curve or the term structure of the interest rates and the default premium are important economic variables to determine future stock returns. Recent literature on the relation between stock returns and business cycles have focused on 4 variables due to their indicator nature that predict the future business cycle fluctuations. The 4 underlying variables are 1) the short-term interest rate; 2) the dividend yield on the overall market;
3) the default spread and 4) the term spread.
The short-term interest rate (yld hereafter) can be proxied by the yield on the 3-month T-bills. Fama and Schwert (1977), Fama (1981) show that this variable is negatively related to the future market returns. More specifically, Choi and Jean (1991) find that the risk relating to yld for small stocks is a significant source of the investing risk, while yld risk for large stocks is ‘negative’. Choi and Jean (1991) argue that the variable yld explains a significant portion of the size premium for the NYSE and AMEX stocks.
The dividend yield on the overall stock market (div hereafter) is one of the oldest variables recognised to affect the expected stock returns.
Studies such as Keim and Stambaugh (1986), Cambell and Shiller (1988), Fama and French (1988), Hodrick (1992) and Nelson and Kim (1993) all show that dividend yield is associated with slow mean reversion in stock returns over the business cycles. Fama (1990) argues that stock prices are low relative to the dividends when the discount rate and expected returns are high, and vice versa. More recently, Ait-Sahalia and Brandt (2001) argue that div should forecast returns on the basis of the present value formula (since div does not appear to predict dividend growth).
The default spread (def hereafter) is measured by the yield spread between the lower-yield to higher-yield bond. This variable measures the credit market conditions, a change in def can be generally interpreted to signal the market’s revisions of expectation of worsening credit market conditions. The use of def is motivated by the studies of Stock and Watson (1989) and Bernanke (1990). By doing the horse race research in predicting future business conditions, these authors find that a variable similar to def does the best job. Hence the variable def is a leading indicator of the state of the economy.
Keim and Stambaugh (1986) use def to predict stock and bond returns. Chen et al. (1986) find that def is an indicator to the business cycles. They argue that the def is likely to be high when the economy is in good condition, and vice versa. Likewise, Fama and French (1989) and Fama (1990) show that def tracks the long-term business cycle conditions and therefore captures variations in expected returns within the business cycles. Daniel and Torous (1991) further suggest that the variable def contains information about future production volatility. Jagannathan and Wang (1996) also report that this variable may capture investor’s hedging concerns associated with time-varying risk premia.
Chan and Chen (1991), Gertler and Gilchrist (1994) and Perez-Quiros and Timmermann (2000) suggest that small and large size stocks have different accessibility to credit markets. Compared to the large firms, small firms are vulnerable to the variation of credit market conditions over the business cycles. Fama and French (1992, 1995) contend that value stocks tend to have high financial leverage and cash flow problems than growth stocks. Hence it is expected that def may be closely related to the size and value premiums.
The term spread (term hereafter) is defined as the long-term interest rate minus the short-term interest rate. This variable can be poxied by the spread between the yield of long-term government bond and the yield of 3-month T-bills. term is considered as one of the most widely used indicators for market's expectation about future interest rates, it also arguably captures the hedging demands to investors associated with changes in interest rates. The term spread tends to decreases in an expanding economy as short-term rates generally rise more the long-term rates. Conversely, term generally increases when economy is in contraction (Lucas et al. (2002)). Indeed, Fama and French (1989), Hahn and Lee (2006) all show that the slope of the yield curve moves in tandem with the business cycle fluctuations. They show that the term spread tends to be low near business cycle peaks and be high when the economy troughs. Daniel and Torous (1991) also provide evidence that this variable is primarily informative about the future growth prospects. Overall, it is argued that positive shocks to the term spread happen at bad times while the negative shocks happen at good times. Since the expected stock returns are low when the economy peaks and high when the economy troughs, the variable term positively predicts expected returns by the effect on the expected company earnings and in term the value of the stock in the context of the dividend discount or cash-flow discount valuation models. Chen (1991) use the term spread to predict excess returns. Recent study of Ait-Sahalia and Brandt (2001) confirms that term is positively related with expected returns.
In summary, the above 4 variables are standard macro-economic variables containing rich information of business cycle risks. The predictability of these variables is due to their business cycle indictors that contain information about the current and future economic conditions. In particular, def and term have long been regarded as proxies for credit market conditions and the stance of monetary policy, indicating that innovations in these variables would capture changes in the financial market's expectation regarding future credit market conditions and the interest rates environment, and would ultimately transitions to the expectations of company earnings and the stock value in the dividend discount or cash-flow discount valuation framework.
2.8 Equity multi-style rotation strategy
The existence of the cyclicality of style returns and business cycle effect highlights the importance of capitalising on the time-varying characteristics of style returns and volatility in the investment process. Such dynamics of stock returns and its relationship with the underlying macroeconomic variables that vary over business cycles would represent significant opportunity as well as significant risks for investors. The evidence of relative style returns under different economy regimes indicates that investors who successfully exploit the variability of multi-style premiums based on different market conditions are likely to be able to obtain better performance than active strategies based on single style investing only. Although some previous studies suggest that the ability to beat a benchmark by market timing or style timing remains debatable (c.f. Henriksson (1984), Connor and Korajczyk (1991), Ferson and Schadi (1996), Chan et al. (2004)), and the implication of style timing strategies is constrained by the inherent difficulties (c.f. Levis (2003)), in market practice, however, investors still have strong incentives to capitalise on the benefit of style rotations in the multi-period asset allocation process.
Style rotation strategies have been attractive to money managers as potential source of adding value. Such strategy could be arguably implemented by, but not limited to, the use of an adaptive approach.