Return-generating Process Research Articles

Behavioral Determinants of Home Bias - Theory and Experiment

We study portfolio diversification in an experimental decision task, where asset returns depend on a draw from an ambiguous urn. Holding other information identical and controlling for the level of ambiguity, we find that labeling assets as being familiar or from the homeland of subjects increases portfolio weights by around 25%, respectively; although the return-generating process remains unaffected. Importantly, we only find these effects when the returns of assets are highly ambiguous. Our ambiguity robust mean-variance model accurately predicts benchmark portfolio weights of the experimental control group, where assets are not labeled: subjects allocate more wealth to assets with low ambiguity. For treatment group portfolios, which show a bias towards assets with a familiar or homeland label, the model does not hold. This misdiversification against the benchmark portfolio can be rationalized via the concept of source dependence of uncertainty attitudes.

On the Pricing Of Chinese Stocks

This study identifies the leading risk attributes to Chinese stock returns. We demonstrate that the forecasting ability of a multifactor expression that includes micro (fundamental) risk factors conditioned by time-varying macro global and local risk factors is significantly superior to the forecasting ability of simpler nested unconditional models. We conclude that micro and macro local and global risks are instrumental in describing the return-generating process of Chinese equities. Using an attribution analysis, we further show that the valuation of Chinese equities is largely conditioned by expected changes in local and global macro risks, and less by unconditional micro risk premiums.

Stock Selection Skills of Indian Mutual Fund Managers during 2000-2012

Mutual funds work on the basis of two maxims maximization of returns and diversification of risk, the attainment of which requires healthy operational practices and efficient investment management. Now, systematic investment management involves a wide variety of activities among which selectivity plays the pivotal role in the return generation process. This study is an attempt to evaluate the investment management of Indian mutual funds in terms of selectivity skills of fund managers during May 31, 2000 to March 31, 2012. The results pertaining to the selectivity skills of fund managers, as found in the study, has revealed that although majority of the schemes have shown positive alpha they are not statistically significant. Only some of the fund managers (around twenty five percent) possess superior selectivity skills based on both unconditional and conditional Jensen model. Conditioning on public information however improves the coefficient of determination. JEL classification: G11; G23

Fixed-income fund performance: Role of luck and ability in tail membership

The risk-adjusted performance (alphas) of a comprehensive and survivorship-free sample of Canadian bond funds after (before) management-related costs is negative (positive) and is weakly sensitive to the choice of the return-generating process. A conditional multi-factor model that captures maturity differences and default risk best describes the return-generating process of these funds. Examination of funds in the tails of the performance distribution using the block-bootstrap method suggests that “bad luck” causes the before costs underperformance of extreme left-tail funds and no fund possesses truly superior management skills.

Country versus sector factors in equity returns: The roles of non-unit exposures

Abstract In this paper we disentangle, analytically and empirically, the roles of the unit-exposure restriction in Heston and Rouwenhorst (1994) . We show that if the purpose is to construct factors, the unit-exposure variance-analysis model can be viewed as just an algorithm that does not really assume a return-generating process; and in practice the effect of relaxing the restriction is immaterial. The restriction is more important if one wants to estimate whether, for a typical stock, the country factor generates more variance than the sector factor: exposure estimation becomes more important (i) the further the average exposures are from unity; or (ii) the higher the dispersion of the exposures. With respect to (i), the more important the corrections for sector (or geographical) structure in country (or sector) factors are, the more the average exposure falls below unity. Thus, the average exposure provides an alternative indicator of the importance of country versus sector effects. We empirically find that the average sector exposure is low (0.3) compared to the average country exposure (0.9). With respect to (ii) we correct the dispersion of exposures for estimation error in the exposures. We find that in our sample these estimation error corrections are more important for sector factors than for country factors, and that country factors are generating far more variance, in a typical stock's return, than do sector factors.

The predictive power of value-at-risk models in commodity futures markets

This paper examines the in- and out-of-sample performance of various value-at-risk (VaR) approaches for commodity futures investments: conventional VaR, the Cornish-Fisher (CF) VaR, GARCH-type VaR models, and semi-parametric conditional autoregressive value-at-risk (CAViaR) models, which do not depend on the assumption of normally distributed i.i.d. error terms. A model comparison reveals that determining the best VaR model depends strongly on the underlying return series. Our results suggest that the CAViaR and GARCH-type models generally outperform the other VaRs. These models can incorporate time-varying volatility adequately and are sensitive to changes in the return-generating process. This has important implications for the risk management of portfolios involving passive long-only commodity futures positions with heavy-tailed data-generating processes.

Differentiating Bullish From Bearish Factors In The Arbitrage Pricing Theory

This is a teaching note on a proposed approach that will correct a common flaw in the way the return-generating process within the APT framework is illustrated in textbooks. The problem can be resolved by dichotomizing the risk factors into two kinds. Based on this approach, the author eliminated the main source of confusion and developed an alternative way to teaching this important financial theory in a comprehensive and intuitive manner.

Ambiguity Aversion and the Arbitrage Pricing Theory

We consider a linear factor APT model and assume that agents are ambiguity averse with respect to payoffs of arbitrage portfolios. In contrast to the standard result, pricing errors need not converge to zero in the limit as the number of assets goes to infinity. Even in the case of exact factor structure, pricing errors may be nonzero and would not be eliminated by ambiguity averse arbitragers. This is because ambiguity about factor loadings and/or expected returns remains even in pure factor portfolios. Moreover, when there is ambiguity about factor sensitivities, strict arbitrage is impossible and pricing errors need not be bounded at all. Replacing a strict no-arbitrage requirement with one which imposes bounds on mean-standard deviation ratios for portfolio returns implies a bound on the pricing error which depends on the maximum admissible mean-standard deviation ratio, ambiguity about the coefficients in the return generating process (RGP), and factor volatilities. We also consider a case when agents can learn the RGP from the data using the framework of Epstein and Schneider (2007, REStud.). We show that pricing errors induced by ambiguity about RGP are quantitatively significant after accounting for the effects of learning. The results have empirical implications for tests of APT, market efficiency (anomalies) and pricing of idiosyncratic risk.

How “Informative” Is the Information Ratio for Evaluating Mutual Fund Managers?

This article aims to determine whether the information ratio (IR) is a useful and reliable performance measure to evaluate mutual fund managers. The authors use a dataset of nearly 10,000 mutual funds for the January 1998 to December 2008 period. The empirical results show that the IR varies over time and across different fund categories. The article finds that representing the true volatility in the return-generating process within a calendar year requires data with a higher frequency than monthly. This reference index or basket needs to capture a large proportion of the investment universe of the respective fund. Because the IR induces managers to hug the benchmark, it should be supplemented with, for example, the Active Share measure to control for the activity level in the portfolio. Finally, in order to separate lucky managers from skilled ones, the long-term track record is an important measure, because luck is generally not persistent over time.

Common factors in international securitized real estate markets

This study investigates the presence of common factors in the securitized real estate markets of the Untied States (US), United Kingdom (UK), Hong Kong (HK), and Singapore (SG). Using a combination of factor analysis and canonical correlation analysis on 10-year monthly return data for 142 real estate securities in the four markets, more common risk factors among real estate securities within a country than across countries are detected. In addition, there is at least one common securitized real estate market factor that is moderately correlated with the world real estate market, and to a lesser extent, with the world stock market. However, the degree of linkage across the four securitized real estate markets is much weaker than the strong linkages present across the four economies. It further appears that the extent to which correlations are found in international securitized real estate markets might largely be due to the increasing integrated nature of the world real economy, rather than a result of the globalization of financial markets. The results are preliminary, but indicative, and suggest that more studies exploring how common factors, together with the local market portfolio, could help explain the return-generating process of securitized real estate.

Measuring Financial Integration via Idiosyncratic Risk: What Effects Are We Really Picking Up?

We study the method proposed by Flood and Rose (FR, 2004, 2005) for checking for financial integration by estimating the risk-free rate using the idiosyncratic component of individual stock returns. Performing simulations with data with a known return generation process, we find that the FR methodology produces poor estimates of the risk-free rate, and hence the FR method fails to accept integration when true. We then show analytically that the FR method actually provides an estimate of the market return, and conclude the FR methodology would also falsely accept integration as long as the market returns in the two markets do not differ widely.

Skewness preference and the measurement of abnormal returns

In performing an empirical analysis of stock market returns there are certain conditions under which the quadratic characteristic line (QCL) will be the appropriate return-generating process compared to the linear characteristic line (LCL). These conditions are whether the parameter associated with the squared market term (the deviation from the mean) is significantly different from zero and whether the return on the market portfolio is asymmetrically distributed (or skewness is present). Examining abnormal returns surrounding stock splits, we find that these conditions hold for our data set. Having ascertained that the conditions for QCL hold, we find that the cumulative average returns (CARs) obtained using QCL dominate the CARs obtained using LCL in the event time and the CAR space for the dividend-increase sub-sample. Furthermore, the standardized abnormal returns for the QCL model are significantly different than those obtained using the LCL model. We find that neither the LCL nor the QCL paradigm reveals any statistically significant abnormal return for the dividend-decrease group. However, for the dividend-decrease group, the CARs for the LCL model dominate the CARs for the QCL model. The standardized abnormal returns for the QCL model are also significantly different than those of the LCL model. Using QCL, we find support for the signalling hypothesis of dividends. We also find that the extent of investor reaction obtained using QCL is statistically significantly different than that obtained using the LCL.

Country v. Industry Factors in Equity Returns: (When) Do Non-Unit Exposures Matter?

The Heston-Rouwenhorst (HR) estimates of country and industry factors have been criticised for assuming that each stock has unit exposures to its own country and industry factor. We address this issue analytically and empirically. Our position is that HR are not modeling and estimating a return generating process. Rather, they provide an algorithm for linearly combining pre-defined standard country and sector portfolio returns such that the effects of different sector weights in country indices (and vv) are taken into account. This is valid in itself, just like it makes sense to compute a valueweighted or equally-weighted market return even when we know that stocks' betas are probably different from unity. In addition, we find that a Fama-MacBeth variant of HR, which does allow for non-unit exposures, produces nearly identical factors as the original version. Allowing for non-unit exposures does make sense, though, if the purpose is to estimate a stock's return-generating process, for instance in an event study, or if one wishes to study the analog of 2 j var(rm), the variance generated by the market factor in an individual stock, rather than var(rm). If this is the purpose, the relative importance of country factors grows, and even more so if one corrects for the effects of estimation error in the exposures. In fact, exposures to industry factors are surprisingly low - about 0.30 instead of over 0.90 for world and country factors.

Event studies with a contaminated estimation period

Event studies are an important tool for empirical research in Finance. Since the seminal contribution of Fama et al. [Fama, E., Fisher, L., Jensen, M., Roll, R., 1969. The adjustment of stock prices to new information. International Economic Review 10, 1–21], there have been many enhancements to the classical test methodology. Somewhat surprisingly, the estimation period has attracted less interest. It is usually routinely determined as a fixed window prior to the event announcement day. In this study, we propose a test that reduces the impact of potentially unrelated events during the estimation period. Our proposition is based on a two-state version of the classical market model as a return-generating process. We present standard specification and power analyses. The results highlight the importance of explicitly controlling for unrelated events occurring during the estimation window, especially in the presence of event-induced increase in return volatility.

An Empirical Comparison of Price‐Limit Models*

ABSTRACTUsing futures traded on the Chicago Board of Trade, Chicago Mercantile Exchange and New York Board of Trade, we test six alternative models of the return‐generating process (RGP) in futures exchanges that adopt a price‐limit regime. We rank the six models according to their return‐prediction ability, based on the mean square error criterion, and we find that the near‐limit model performed best for both the estimation period and the prediction period. A reliable prediction of the expected return can have important implications for both traders and policy makers, concerning related issues such as the employment of long or short strategy, margin requirements and the effectiveness of the price limit mechanism.

Measuring Financial Integration via Idiosyncratic Risk: What Effects Are We Really Picking Up?

We study the method proposed by Flood and Rose (FR, 2004, 2005) for checking for financial integration by estimating the risk-free rate using the idiosyncratic component of individual stock returns. Performing simulations with data with a known return generation process, we find that the FR methodology produces poor estimates of the risk-free rate, and hence the FR method fails to accept integration when true. We then show analytically that the FR method actually provides an estimate of the market return, and conclude the FR methodology would also falsely accept integration as long as the market returns in the two markets do not differ widely.

Portfolio Choice and Estimation Risk. A Comparison of Bayesian to Heuristic Approaches

Estimation risk is known to have a huge impact on mean/variance optimized portfolios, which is one of the primary reasons to make standard Markowitz optimization unfeasible in practice. This issue has attracted new interest in the last years, and several approaches to incorporate estimation risk into portfolio selection have been developed only recently. In this article, we review these approaches as well as some older ones and compare them in an empirical out-of-sample study. The approaches can be classified along two criteria. First, we can differentiate heuristic approaches (restricting portfolio weights and employing simulation techniques) and those based on Bayesian statistics (shrinking the portfolios towards a pre-determined target). Second, the assumptions about the return-generating process differ, either assuming returns to be IID distributed or to be partly predictable. The central result of our empirical study is that all of the IID approaches, whether they account for estimation risk or not, are not superior to simple investment strategies like holding the market portfolio. A risk-adjusted outperformance is possible only if sample means are substituted with conditional expected return estimates. Furthermore, the Bayesian approaches reduce turnover and stabilize portfolio weights.

Portfolio Choice and Estimation Risk. A Comparison of Bayesian to Heuristic Approaches

Estimation risk is known to have a huge impact on mean/variance optimized portfolios, which is one of the primary reasons to make standard Markowitz optimization unfeasible in practice. This issue has attracted new interest in the last years, and several approaches to incorporate estimation risk into portfolio selection have been developed only recently. In this article, we review these approaches as well as some older ones and compare them in an empirical out-of-sample study. The approaches can be classified along two criteria. First, we can differentiate heuristic approaches (restricting portfolio weights and employing simulation techniques) and those based on Bayesian statistics (shrinking the portfolios towards a pre-determined target). Second, the assumptions about the return-generating process differ, either assuming returns to be IID distributed or to be partly predictable. The central result of our empirical study is that all of the IID approaches, whether they account for estimation risk or not, are not superior to simple investment strategies like holding the market portfolio. A risk-adjusted outperformance is possible only if sample means are substituted with conditional expected return estimates. Furthermore, the Bayesian approaches reduce turnover and stabilize portfolio weights.

Why Has Firm-Specific Risk Increased Over Time?

Firm-specific risk climbed steadily between 1962 and 1999, but fell sharply between 2000 and 2003. We hypothesize that changes in the composition of the market, rather than fundamental changes in the economy or return-generating process, drive these changes in aggregate firm-specific risk over time. Specifically, we posit that three factors are primarily responsible for observed changes in firm-specific risk: changes in the market weights of riskier industries, changes in the relative role of small stocks in the market, and measurement error associated with changes in within-industry concentration. Our empirical tests reveal that each of these factors contribute to changes in firm-specific risk. Moreover, our analysis demonstrates that, combined, these three factors largely explain changes in firm-specific risk over the past 40 years.

Macroeconomic announcements, volatility, and interrelationships: An examination of the UK interest rate and equity markets

This study investigates the intraday and daily pricing behavior of UK interest rate and equity index futures contracts. The paper initially examines the response of Short Sterling, Long Gilt, and FTSE100 to the release of scheduled macroeconomic announcements before employing dynamic time series techniques in order to reveal the nature of causal transmission patterns between these variables. In brief, short-term interest rates were found to be highly sensitive to indicators of prevailing economic conditions. However, the release of data important in the formation of inflationary expectations had a relatively subdued impact on long-term rates. Announcement effects appear somewhat ambiguous for the stock market. The analysis also reveals the bid–ask bounce and swift mean reversion in volatility to be important behavioral features of the return-generating process. Whilst the three variables appear to be bound by two cointegrating relationships, the tests for lead/lag relationships produce mixed results.