T. Daniel Coggin

Multifactor Asset Pricing Analysis of International Value Investment Strategies

Using a large international equity market database that has not been previously used for such a purpose, this paper documents that value (i.e., high book-to-market ) stocks outperform growth (i.e., low book-to-market ) stocks, on average, in most countries during the January 1975 December 1995 period, both absolutely and after adjusting for risk. The international evidence confirms the findings of previous work reported for the U.S.. For 1975-1995, the annual difference between the average returns on portfolios of high and low book-to-market stocks is 12.94% in North America, 10.42% in Europe, 17.26% in Pacific-Rim per year, and value stocks outperform growth stocks in 17 out of 18 national capital markets. Our analysis also shows that a three-factor model explains most of the cross-sectional variation in average returns on industry portfolios across countries and that the superior performance of the value investing strategy, documented in this study, is a manifestation of size and book-to-market effects. These results are consistent with those reported by Fama and French (1994, 1996) that show that the value-growth pattern in stock returns is largely explained by a three-factor asset pricing model. Our results suggest that the Fama and French (1996) three-factor asset pricing model is not limited to the U.S. stock market. Several recent studies have documented that value strategies (i.e., investing in stocks that have low prices relative to earnings, dividends, historical prices, book assets, or other measures of value) produce higher returns. Among these studies are Basu (1977), Rosenberg, Reid, and Lanstein (1985), De Bondt and Thaler (1985, 1987), Jaffe, Keim, and Westerfield (1989), Chan, Hamao, and Lakonishok (1991), Fama and French (1992), and Lakonishok, Shleifer, and Vishny (1994), all of which show that stocks with high earnings/price ratios or high book-to-market values of equity earn higher returns. A number of alternative explanations for the observed superior returns of value strategies exist. Fama and French (1992, 1993) argue that value strategies are fundamentally riskier and therefore the higher average returns associated with high book-to-market stocks reflect compensation for bearing this risk. A similar argument has been made by Chan (1988) and Ball and Kothari (1989). They suggest that the market overreaction (i.e., winner-loser effect) result of De Bondt and Thaler (1985) is due almost entirely to intertemporal changes in risks and expected returns. Lakonishok, Shleifer, and Vishny (1994), however, argue that value strategies yield higher returns because investors are able to identify mispriced stocks and not because they are fundamentally riskier. Ball, Kothari and Shanken (1995) report that the profitability of the value investing strategy is driven by performance measurement problems and microstructure effects. Kothari, Shanken and Sloan (1995) attribute the superior performance of value strategies to the research design and database used to conduct these studies [i.e., survivorship bias (see Davis (1994)], look-ahead bias [see Banz and Breen (1986)] and data snooping [see Lo and MacKinlay (1990)] in the selection of firms that are included in the CRSP and Compustat databases. Chan, Jagadeesh and Lakonishok (1995) show that this is not the case. Davis (1994) reports a value premium in U.S.stock returns prior to 1963 as well. In a recent study, Fama and French (1996) document that the superior performance of the value investing strategy is a manifestation of size and book-to-market effects. Empirical work has 3 discovered some stylized facts on the behavior of stock prices that cannot be explained by the capital asset pricing model (CAPM) of Sharpe (1964) and Lintner (1965). However, this evidence is largely based on firms in the U.S., and it is not at all clear how these facts relate to different countries. Without testing the robustness of these findings outside the environment in which they were found, it is hard to determine whether these empirical regularities are merely spurious correlations that may not be confirmed across capital markets. This study fills this gap in the literature. In addition, we document for the first time the cross-sectional relationship between beta, size (SMB), book-to-market (HML) and average industry portfolio returns in a sample of 2,629 stocks in 18 equity markets ( including the U.S.) over the 1975-1995 period. The first objective of this article is to examine the robustness of the value-investing strategy using monthly data for 18 equity markets and four regions of the world economy (i.e., North America, Europe, Pacific Basin and International) obtained from the Independence International Associates, Inc.(IIA) database for the 1975-1995 period. We note that in this study the term “international” refers to both the U.S. and non-U.S. stock markets. There are 1,554-2,629 stocks in this database that are tracked by Morgan Stanley Capital International (MSCI) throughout this period. For each country there is a set of five portfolio returns: market, value, growth, small and large. The sample covers more than 75 percent of each country’s market capitalization. There is no survivorship bias in the data set ( as defined by the MSCI database ) since each portfolio is calculated based on the companies that were actually in the MSCI database as of the Januaryrebalance date of each year. The use of such a broad international data set provides a unique opportunity for this analysis. By focusing on the 1975 to 1995 period, this paper studies the behavior of stock returns across countries using a large and updated database that has not been previously used for such a 4 purpose. To the extent that other countries are similar to the U.S., they provide an independent sample to reproduce the regularities found in the U.S. and compare the results to those reported in earlier studies . To the extent that our sample contains countries that are not similar to the U.S., it will increase our ability to shed additional light and help us understand the forces behind the superiority of value strategies. The second objective of the article is to investigate whether value stocks are riskier than growth stocks, since this issue remains controversial among researchers. We focus on betas, coefficients of variation and Sharpe ratios for value and growth strategies. Consistent with the empirical findings of Lakonishok, Shleifer, and Vishny (1994), our analysis provides evidence in support of (i) the superior performance of the value-investing strategy and (ii) the view that such strategies are not fundamentally riskier in 18 equity markets. However, this pattern in international stock returns cannot rule out the possibility that the superior performance of the value-investing strategy is a manifestation of size and book-to-market effects, as reported for the U.S. by Fama and French (1996). The third objective of this study is to examine whether the superior performance of valueinvesting in 18 capital markes is a CAPM related anomaly as argued by Fama and French (1996). This issue is addressed by implementing the Fama and French (1996) three-factor model internationally. However, our empirical investigation is based on portfolios that allow the slopes of the factors to vary over time as opposed to forming portfolios on size, book-to-market and other measures of value that result in factor loadings that are essentially non-time varying [ see Fama and French (1996 ) ]. Our evidence shows that the three-factor model explains most of the cross-sectional variation in average returns on industry portfolios across countries, and that the superior performance 5 of value investing documented in eighteen stock markets ( including the U.S.) is driven by relative size (SMB ) and distress (HML ) effects. Our results are consistent with those found for the U.S. by Fama (1994) and Fama and French (1996). Furthermore, the stock market evidence from around the world suggests that the Fama and French (1996) multi factor asset pricing model is not limited to the U.S. capital market. It holds across capital markets and regions of the world, although it does not uniformly explain portfolio returns in all markets. The rest of the paper is organized as follows. Section I describes the data used in this study. Section II presents annualized return spreads for value and growth strategies based on value and growth portfolios formed on an annual basis for three different investment horizons. Section III examines whether the superior performance of value stocks is related to the upward movements in stock markets. Section IV investigates whether the arbitrage portfolio formed by buying value stocks and selling growth stocks is associated with the effects of firm size. Section V analyzes the robustness of value investing strategies using the Fama and French (1996) three-factor asset pricing framework, and Section VI concludes the paper.

An analysis of the diversification benefit from international equity investment

John E . Hunter and T . Daniel Coggin N ow that national stock markets have developed in a number of countries, investors are increasingly attempting to diversify risk by spreading their portfolios across different national stock markets. Recent comprehensive summaries of the growing international portfolio selection literature are given in Adler and Dumas [1983] and Solnik [1988]. The extent to which investment risk can be diversified depends upon the degree to which national markets are dependent upon each other. If all national markets were completely dominated by a single world market factor (i.e., if all cross-national correlations were 1 .OO), then international diversification would have no benefit. If all national markets were completely independent (that is, if all cross-national zorrelations were zero), then international diversifization over an infinite number of countries would zompletely eliminate the effect of variation in national .narkets. Studies of stocks within the United States have :ound that the correlation structure of returns is .argely explained by a “single index” model. That is, .here is one dominant causal agent or ”market factor” *n the correlation structure of stock returns (see, e.g., .<ing [1966], Cohen and Pogue [1967], and Elton and 3ruber [1973]). Other factors, such as industry group, xere present but have been found to be of relatively ninor importance. In recent research that has sparked considerible debate, multi-factor models derived from the aritrage pricing theory still find a single dominant ‘actor in U.S. and Canadian stock return correlation natrices (see Kryzanowski and To [1983], Conway 33 5 3 8 2

A meta-analysis of mutual fund performance

The purpose of this article is to introduce the statistical technique of meta-analysis of regression results using as our example the Lee and Rahmann (1990) study of the performance of 93 mutual funds. Specifically, we derive and estimate the meta-analysis formulas, explicitly adjusted for correlated regression residuals, which quantify the effect of sampling error on their reported regression results. Our analysis of selectivity reveals some real variation around a mean risk-adjusted excess return of about 1% per year; while our analysis of market timing reveals some real variation around a negative mean value and confirms that the correction for heteroscedasticity does make a difference. An examination of the 80% probability interval for the mean selectivity value indicates that the best mutual funds can deliver substantial risk-adjusted excess returns.

THE DIMENSIONS OF INTERNATIONAL EQUITY STYLE

his article provides a comprehensive summary of the dunensions of international equity style in the developed world from T January 1975 through December 1996. It describes eight dunensions of equity style using twenty-two years of returns on twenty countries grouped into four time periods and s i x geographc regions. Other studies have looked at various aspects of international equity style for smaller samples of countries and shorter time periods. See, e.g., Capaul, Rowley, and Sharpe [1993], Umstead [1993a, 1993b, 1995, 19971, and Sinquefield [1996]. To our knowledge, ours is the first study to present such a comprehensive overview of international equity style.

A Panel Study of U.S. Equity Pension Fund Manager Style Performance

The authors examine the investment performance of 292 U.S. equity pension funds in three major style categories for two consecutive twelve-quarter time periods, including survivors and non-survivors. They offer three major conclusions. First, the choice of an equity benchmark affects the magnitude of alpha. Second, it is difficult to find investment managers within equity styles who consistently add value relative to the S&P 500 and the appropriate style benchmark. Finally, there is no evidence that the number of funds outperforming that appropriate style benchmark index (after fees) exceeds random chance. Including non-survivors and applying selected control variables does not affect the basic results.

Some Tests of the Risk-Return Relationship Using Alternative Asset Pricing Models and Observed Expected Returns

This study examines the performance of three asset pricing models: the CAPM, the APT and the UAPT using observed expected returns from a three-phase dividend discount model with Value Line analyst estimates of future company-level earnings, dividends and growth rates. Our study is the first we know of to test the three major asset pricing models using observed expected returns. Our results are similar to prior research using ex post (realized) returns in that we find that the UAPT using macroeconomic factors is the best performing model, followed by the APT and the CAPM. However, our results also suggest that the importance of macroeconomic factors is much greater to expected returns than to realized returns, and the corresponding R2 values for models using expected returns are much higher than for models using realized returns. Combining our results for the UAPT with those of Marston and Harris (1993) for the CAPM suggests that these models are more successful in tests using observed expected returns than in tests using realized returns as proxies for expected returns. Unit root tests suggest that monthly observed expected returns follow the classic “random walk without drift” model while monthly realized returns do not.

The January Effect and the Global Value-Growth Premium

Here are important findings on abnormal returns for value stocks and the January effect, two challenges to the efficient market model. There is a January effect on the value-growth spread in the U.S. for both large- and small-capitalization stocks for the period 1975–2001. Use of four different benchmark indexes adds robustness to this result. Examination of five major non-U.S. markets—EuroPacific, France, Germany, Japan, and the U.K.—reveals a January effect in the large-cap value-growth spread for three markets (France, Germany, and U.K.) and in the small-cap value-growth spread for all five markets.

Change in Political Party and Issue Attitudes

This chapter discusses the change in political party and issue attitudes. Political voting is largely determined by three factors: the political party of the candidate, the stance of the candidate toward salient political issues, and the candidates personal characteristics, such as warmth or trustworthiness. Multiple regression analysis has usually suggested that party identification and political attitudes are the strongest predictors of vote; though these two variables are themselves correlated. The focus of this chapter is on the causal determination of these two attitudes and the explanation for the correlation between them. The causal theories to be tested assume that the primary determinant of political identification and political attitudes is the messages emitted by the parties as to their stands on various issues. The specific theories are derived from the theories of attitude change. In these theories, the political party is given the role of the source of messages, where the content of the message is the stand taken by the party. The voter is given the role of the receiver and his attitude is his own position toward the issue addressed by the party message under consideration. The overall impact of the set of messages presented by the party can be assessed by cumulating the impacts across the various issues.