«Equity Style Investing RONG, WU How to cite: Equity Style Investing, Durham theses, Durham University. RONG, WU (2013) Available at Durham E-Theses ...»
The term spread (term) also exerts significant impact on the style allocation process. The regression coefficients of term are all positive, and in the style allocation context it has significant positive tilt for small stocks at both monthly and quarterly horizons, and significant positive sign for growth stocks on monthly frequency. This suggests that a positive shock to variable term would encourage investors to overweight small cap stocks or growth stocks. Fama and French (1989) point out that term spread tracks the short term fluctuations of business cycle and its value to signal expected returns are high during recessions and low during expansions. It is argued that positive shocks to term happen at bad times while the negative shocks happen at good times. Hence investors are induced to hold more small stocks or growth stocks when economic situations are bad. This conclusion seems intuitively contradicts to the results documented by Chan and Chen (1991) for small size stocks but is consistent with Petkova and Zhang (2004). Chan and Chen (1991) argue that small firms tend to be marginal firms that have generally lost market value due to poor performance. Such firms have high financial leverage and cash flow problems and hence are difficult to survive to bad times. In light of this argument, it is reasonable to assume that investors would underweight small cap stocks when economy is in recession.
As a summary, business cycle variables exert different mechanisms to the conditional style return distributions and the style investing implementations. Variables such as def, yld, div and term convey useful information about the current and future directions in the broad economy and business cycle environment assumed to determine the inter-temporal behaviour of equity style dynamics. As Petkova (2006) points out, these variables model the two aspects of the time-varying investment opportunity, the yield curve and the distribution of stock returns. Investment strategies incorporating such business cycle predictors typically yield better performance relative to strategies disregarding the stock return predictability. Traditional portfolio selection generally first specifies a model for the moments of stock returns and then implementing the optimal allocation using plugged estimates that is based on partial information for expected returns forecasting. In contrast, investing strategies directly parameterising on the business cycle variables can arguably capture the time variations of all the moments of asset returns and therefore generate higher returns. Such outperformance is arguably driven by the different mechanisms that business cycle information affects in the investment process. Namely, shocks to the variables are found to be transmitted very differently in asset pricing and asset allocation process. It is found that apart from their predictability on return distributions, variables such as yld, div and term exert significant impact on style allocation on both size and value dimensions.
Interestingly, the optimal asset allocation policy derived by such variables often contradicts to empirical asset pricing predictions. The optimal style investing strategies significantly tilt to holding small-cap and growth stocks during economic bad times despite small stocks may have financial difficulties in recessions and lower expected returns results from positive shocks to the variables. These results are consistent with Avramov and Chordia (2006) who also find that their outperforming strategies in NYSE-AMEX stocks hold small cap, growth and momentum stocks. Since mean-variance optimal investing uses asset returns and volatility as inputs, it is suggested that style volatility, not the expected style returns, plays a key role in the optimal style investing framework.
To get a more clear perspective as how information of style volatility affects the allocation process, Figure 5-2 shows the time-series of style allocation weights based on these two approaches. Indeed, conditional investing capitalising on the information of the business cycle exhibits significant difference and tends to bet more extreme positions on both long and short directions. Such investing tends to long more for the long side and short more for the short side as compared to the traditional optimal investing. Investors following the conditional investing (the Doctrinaires) directly predict their optimal style investing weights with business cycle predictors and hence benefiting from capturing more information beyond the first and second moments of stock returns that affect asset allocation decision, and therefore yield more extreme tilts but better in- and out-of-sample performance.
Figure 5-2 The time-series of style weights based on traditional and unconditional (regression-based) style investing
5.6 Summary and conclusions Extant literature documents the benefits of incorporating business cycle effects on investor’s asset allocation process. However, the transmission mechanism of such business cycle volatility to portfolio selection is not extensively studied. Meanwhile, prior studies generally unrealistically focus on all the stocks in the market. When dealing with optimal portfolio selection problem, prior studies take the tradition approach of Markowitz (1952) and focus more on the timevarying nature of return distributions driven by different business cycle predictors. However, the role such predictive variables play on determining optimal portfolio allocation is less directly explored.
Chapter 5 contributes to the literature by allowing the hypothesised investors to have access to different market segments and implement different equity style investing without the restriction of long or short.
Such investors can be regarded as hypothesised “fund of hedge funds” investors. It is understandable that investors care more about how the economic exogenous forces directly determine the ultimate investing choices (i.e. optimal style timing weights). Following the methodology proposed by Brandt and Santa-Clara (2006), this chapter implement an optimisation framework to investigate several equity style investing strategies based on business cycle information and examine their ex ante in-sample and ex post out-sample performance. By answering questions like if business cycle predictor variable x increases, should the investor move to/away from y style, this chapter gives multi-style investors an intuitive manner to understand their asset allocation process when incorporating business cycle predictability.
The empirical results in this chapter first suggest that regardless of return horizons, investors tend to significantly long value stocks or small stocks, and short growth stocks or large stocks in their optimal style allocation process. The U.K. market data shows that investors tend to buy small value, large blend and large value stocks in the long position, and short sell small growth, middle growth and large growth stocks. In particular the small value stocks are overwhelmingly to be held as it best captures the interaction of size and value effects.
It is found that the conditional style investing incorporating business cycle information and the unconditional style investing disregarding business cycle effect is much different. Specifically, sceptical investors who disregard business cycle predictability are conservative regarding their overall net equity exposures relative to the Doctrinaires who have strong prior beliefs about the business cycle information. The latter tend to be aggressive and generally end up with extreme positions to some styles and often financed by leverage. One reason for such extreme tilt is because the Doctrinaires believe the return differential of these styles can be estimated using business cycle predictors thus the exposure can be reduced at bad times when expected returns are low or volatility is high.
This chapter shows that business cycle variables affect the conditional
style returns and the optimal style investing in quite a different way:
First, default spread (def) plays is similar role in both expected returns and style allocation, however its significance declines in the style investing process despite of its significant role in the expected return distributions. In addition, it is predicted that a positive shock to the short-term interest rate (yld) would induces investors to overweight small stocks and underweight large stocks despite the lower expected returns for small stocks and higher expected returns for large stocks are estimated. In addition, a positive shock to yld would lead investors to tilt to growth stocks, which matches their higher expected returns signalled by changes of yld.
Second, the dividend yield (div) predicts the style allocation along both size and value dimensions. Although div has more significant (positive) impact on returns for small cap stocks (value stocks) than for large cap stocks (growth stocks), a positive shock to this variable would induce investors to overweight large stocks (growth stocks) and underweight small stocks (value stocks). The term spread (term) also exerts significant impact on the style allocation process. Generally a positive shock to term would induce investors to overweight small cap stocks or growth stocks.
Overall, it is concluded that business cycle predictability benefits investors’ dynamic optimal style investing. Variables such as yld, term, div and def exert a strong influence on the shape or location of investor’s optimal style investing frontier. Smart investors who can capitalise on the conditional business cycle information consistently beat those disregarding business cycle influence, both in-sample and out-of-sample.
Chapter 6 Summary, conclusions, implementations and recommendations for future research
6.1 Summary of the research Human beings are capable of classifying objects into categories to simplify the decision-making process. The idea of categorisation is also pervasive in today’s financial market. Investors generally classify all the assets in the market into several groups like equity, cash, real estate etc. Within each asset class they also define some subgroups that share properties similar to the major asset class but are unique along specific dimension. For example, stocks can be subdivided according to market values as small-caps and large-caps. In addition, they can also be classified as value stocks and growth stocks based on some valuation multipliers. According to the relative returns, stocks can be labelled as ‘winners’ or ‘losers’. In the investment world, ‘style’ refers to such systematic classification of investing assets by market segments. The definition of style is not fixed, due to market innovation or academic research findings, new styles may evolve and old styles may die off as time goes by. Equity style investing is an investment strategy based on stock classifications. In today’s investment industry, style investing is well recognised and has gained growing popularity.
The concept of equity style and style investing offers an example of the exchange of ideas between academic research and investing practice.
Style investing changes the way academics and practitioners think about investment. Recent empirical studies suggest that Institutional investors like fiduciaries of pension and endowment funds follow specific investment styles (Brown and Goetzmann (1997), Fung and Hsieh (1997), Chan et al. (2002)). For these institutional investors, the control of investment style has become a critical aspect of investment monitoring and decision-making process. Despite the obvious simplicity of following style investing in the asset allocation process, money manager’s incentive for equity style investing also stems from capitalising on the relative performance across equity styles.
Financial markets have long observed the style return differentials together with the tremendous swings of equity style dynamics. Overall, empirical findings have shown that over the long term small-cap investing and value investing have been more advantageous in most equity markets around the world, but there are periods where smalllarge returns and value-growth returns reverses dramatically. The dynamics of equity style returns have introduced the new risk-return structure for active portfolio management. But to capitalise on the style effect, money managers would need to not only be able to identify the underlying drivers that determine the relative style performance, but also to capture the mechanisms through which those underlying driving forces work. Most importantly, active managers must be able to capture the dynamic properties of those driving forces to forecast the future style trends in order to optimise their investment process.
Over the years, although the benefits of style investing have been well recognised, the academic view of the cause for such benefits is very much debatable. There is still no general consensus as why some asset classes earn better returns than others do in the same period.