«Equity Style Investing RONG, WU How to cite: Equity Style Investing, Durham theses, Durham University. RONG, WU (2013) Available at Durham E-Theses ...»
The dividend yield also predicts the style allocation along both size and value dimensions. While this predictor has more significant and positive impact on return distributions for small-cap stocks and value stocks than for large-cap stocks or growth stocks, a positive shock to short-term interest rate would induce investors to tilt towards large stocks or growth stocks and tilt away from small stocks or value stocks. The term spread also exerts significant impact on the style allocation process. A positive shock to the term spread would induce investors to overweight small-cap stocks or growth stocks in general.
Overall, Chapter 5 concludes that business predictors such as short term interest rate, term spread, dividend yield and default spread exert a strong influence on the shape or location of a mean-variance investor’s optimal style investing frontier. Investors who can capitalise on the conditional business cycle information consistently beat those disregarding business cycle influence, both in-sample and out-of-sample.
1.5 Research structure The remainder of the thesis is structured as follows: Chapter 2 reviews the literature. Starting from the equity style investing history, this chapter first reviews the firm characteristics documented to be related to the cross-sectional average stock returns. Since investors categorise stocks based on firm characteristics, some typical style investing advocated by such characteristics are explained and the time-series of the style performance over the business cycles are analysed. Chapter 2 also reviews some competing explanations for typical style premiums as advocated by traditional and behavioural finance. Following the time-varying style return dynamics, the benefits of style rotation strategies are reviewed. In response to the business cycle effect in the predictability of style return dynamics, the optimal style allocation in a mean-variance framework is also extensively reviewed.
Chapter 3 examines the relative importance of the style driving sources that determines the differentials of style returns in the UK market. Using U.K. stock data, this chapter explicitly tests how firmspecific characteristics and the business cycle conditions function separately to drive the dynamics of stock performance based on the categorisation of size and value-growth dimensions. Specifically, Chapter 3 aims to answer a central question: what is the dominant driving force to determine the relative style performance, the firm characteristics or the business cycle risk?
Chapter 4 investigates an adaptive tactical style investing problem by applying a binomial approach to focus on the shifting between pairs of equity styles. At each given point of time investors extrapolate the relative expected performance of different asset classes like value versus growth stocks or small versus large stocks according to their past performance and bet 100% of investing on the ‘winner’ style financed by shorting the ‘loser’ style. By exploring the profitability of style momentum the evidence of equity style cycles in the U.K. stock market is examined. More importantly, by examining the profitability of such style momentum strategies after controlling for the stocklevel momentum and the industry-level momentum effects, Chapter 4 further tests whether style effects are unique in affecting the crosssection of stock returns.
In response to the increasing popularity of using macro information to aid optimal style selection for the quant circles in the investment community, based on the methodology of Brandt and Santa-Clara (2006), this chapter approximates the solution of a mean-variance multi-style investor’s optimal style rotation question incorporating the business cycle predictability. The approach is parsimonious as the optimal style weights are parameterised directly on a set of pervasive business cycle predictors. By exploring how the directions of the expected style returns as well as the location and shape of the optimal style allocations are affected by given shocks to business cycle variables, Chapter 5 demonstrates how business cycle volatility affects asset return volatility and in turn investor’s optimal style allocation.
Finally, Chapter 6 concludes the thesis and offers recommendations in the areas for relevant future research.
2.1 Introduction The research of equity styles began in 1970s when the investment community began to gather and analyse market data and money managers. Financial analysts have long observed clusters of stocks with similar characteristics and performance patterns in the U.S.
markets. Early studies such as King (1966) and Farrell (1975) use cluster analysis to identify natural groupings of stocks and portfolios.
They find that some groups of stocks and portfolios with similar characteristics demonstrate similar return patterns. Other studies such as LeClair (1974) suggest that groups of fund managers with similar investment philosophies could also lead clustering portfolio returns. The most prominent study in the context of investment style and mutual fund performance analysis was conducted by Sharpe (1988, 1992). Sharpe developed a returns-based technique that is rooted in analysing the covariance structure in manager return patterns. It is proposed that managers with different styles would behave differently and this behaviour could be determined by looking at the underlying fund’s ‘effective asset mix’ in terms of a predefined set of style indices. In addition to Sharpe’s returns-based approach to assess the style characteristics of a portfolio, the portfolios-based approach based on the actual holdings is also popular in the investment industry. For example, Grinblatt and Titman (1989) employs the quarterly holdings of a sample of mutual funds to construct an estimate of their gross returns. Daniel et al. (1997) also evaluate the portfolio performance based on the characteristics of stocks held by the portfolios. In their study, the benchmarks are constructed from the returns of some passive portfolios matched with stocks held in the underlying portfolio based on market value, book-to-market ratios and past relative returns.
The heightened attention of style and style investing in today’s investment community is perhaps driven by several motives. First, academic studies suggest that investment style shapes the pattern of portfolio returns more than any other factors in the investment process. It is argued that the philosophy of how to select stocks trumps what individual stocks are selected in determining the overall portfolio performance. Brinson et al. (1986) document that asset allocation decision accounts for about 90% of the return variations in large pension funds. Likewise, Hansen (1992) argues that return differentials due to investment style accounts for approximately 60% of the performance over short and medium term. More specifically, Sharpe (1992) proposes that 90% of the performance of equity funds is due to the overall style of the fund, the remaining 10% is due to the individual characteristics of the specific securities hold.
Second, in recent years, money managers have been required by the consultants and trustees to identify their investment styles. For marketing purposes, fund managers generally define their fund products into different style classifications to meet different investors with dedicated risk preference. Hence in today’s asset management industry, style has been widely recognised as a tool for portfolio management and performance evaluation. Style analysis is important for portfolio management as it can simplify the portfolio selection problem and the process of diversification (c.f. Barneby et al.
(1986), Mullainathan (2000)). Hence professional money managers can benefit from the style analysis to build portfolios, while plan sponsors or individual investors can obtain information regarding the managers’ area of expertise and therefore become more knowledgeable about how to allocate their money across funds with different investment styles.
2.2 The dimension of equity styles
The concept of equity style in the stock markets can be defined as a systematic classification by market segments sharing distinguishing characteristics. Such characteristics can be quantified by a number of descriptors like measures of return volatility, the firm size, values of corporate growth rate and the quality of the underlying company etc. These common factors are recognisable components of equity portfolio styles box widely accepted in the investment community (e.g.
Morningstar equity style box). Correspondingly, equity style investing can take different forms based on the underlying framework.
A popular style investing approach is to form portfolios based on firm characteristics. Style investing based on firm-specific characteristic factors uses firm size or other valuation multiples as criteria to sort stocks to construct portfolios. In addition to the general categorybased methods, investors may also follow positive feedback trading according to the relative returns, namely to long (short) past winners and short (long) past losers based on the stock’s performance. Such momentum or contrarian investing is well recognised in the market practice. Characteristic related and feedback investing strategies aim to exploit and benefit from market deficiency. The massive existence of anomalies in the financial markets implies that investors chasing such strategies may have good opportunity to add value through efficient style rotations. Today, equity style investing such as value, growth, contrarian and momentum are familiar and well-considered concepts in the asset management industry. The following sections conduct an extensive review regarding these investing strategies.
2.3 Size, value and growth investing
The stock market as a whole can be broadly divided into different types of stocks based on their similarities along some dimensions like firm values, valuation multiples and risk exposures etc. The size (measured by the market capitalisation) and value-growth (defined by valuation multiplier) are earliest style categories recognised in the investment community. Although stocks can be sorted based on other dimensions, to categorise the stock universe into the broad category of value and growth class is more natural because empirical studies generally suggest that the return differentials to these stocks are more likely driven by the economic fundamentals.
While it is intuitive to understand that small and large stocks differ in that they have different market values, the characteristics of value and growth stocks can differ in a number of ways. Value and growth stocks generally share common characteristics of valuation multiples.
Value stocks generally have relatively low prices as compared with the underlying fundamentals. Such stocks normally have low priceto-earning (PE) ratios, price-to-book (PB) ratios or price-to-cash flow (PC) ratios and high book-to-market ratios (BM). Value stocks also have higher dividend yield (DY) and lower price-to-net tangible asset ratio. In contrast, growth stocks have opposite characteristics, such stocks typically have high PE, PB or PC and low BM values relative to their stock prices, they also tend to have lower dividend yield or higher price-to-net tangible asset ratio.
A large financial literature relates stock returns with firm-specific characteristics. Since the introduction of asset pricing model (CAPM) of Sharpe (1964) and Lintner (1965), academic researchers find that CAPM cannot fully explain the stock returns with market risk along.
Researchers have therefore identified factors other than market risk to interpret the stock returns. The published papers document that firm-specific characteristics like size and value-growth descriptors are significantly related to expected stock returns. Early pioneering works of Basu (1977) and Banz (1981) use PE ratios and firm size to explore the cross-section of average stock returns on U.S. equities and document the evidence of ‘PE effect’ and ‘size effect’. Chan et al.
(1991) find the explanatory power of book-to-market (BM) ratio to the Japanese stock returns. Studies such as Rosenberg et al. (1985), Lakonishok et al. (1994) find that other factors, such as cash flow-toprice ratio and the past sales growth rate, are also significant to stock returns. The prominent study of Fama and French (1992, 1993) use a multifactor asset pricing model supplementing the standard market risk premium with factors related to the firm size and BM ratio and find that their three-factor model can capture large fractions of the variability of cross-sectional average stock returns in the U.S. stock markets. These papers and many others have served to deepen our understanding in the role that firm characteristics played in explaining the average stock returns in the international framework5. The pervasive influence of these empirical findings has
Partial list of other papers in this literature includes Ball (1978), Sharpe (1982),
Chen et al. (1986), Bhandari (1988), Jaffe et al. (1989), Capaul et al. (1993), Breen and Korajczyk (1995), Chan et al. (1995) and Kothari et al. (1995), among many others.
been such that it is now a common practice to define the investment styles along two basic dimensions, namely the small-large and valuegrowth, in today’s asset management industry.
A considerable literature exists to explore the relative performance of basic equity style investing in the global stock markets. The general findings are, first, investing in smaller firm stocks tend to outperform over the long-run but with higher risk than investing in the large-cap stocks. For example, over the period of 1926 to 2002, investing in small-cap companies could outperform large-cap strategy by almost 5% annually despite of the large stocks’ dominance during 1950s and 1980s in the U.S stock markets (State Street Research (2003)).