• Selected
• Entire List

Abstract:
Even though stock returns are not highly autocorrelated, there is a spurious regression bias in predictive regressions for stock returns related to the classic studies of Yule (1926) and Granger and Newbold (1974). Data mining for predictor variables interacts with spurious regression bias. The two effects reinforce each other, because more highly persistent series are more likely to be found significant in the search for predictor variables. Our simulations suggest that many of the regressions in the literature, based on individual predictor variables, may be spurious

Institutional subscribers to the NBER working paper series, and residents of developing countries may download this paper without additional charge at www.nber.org.

Abstract:
Even though stock returns are not highly autocorrelated, there is a spurious regression bias in predictive regressions for stock returns related to the classic studies of Yule (1926) and Granger and Newbold (1974). Data mining for predictor variables interacts with spurious regression bias. The two effects reinforce each other, because more highly persistent series are more likely to be found significant in the search for predictor variables. Our simulations suggest that many of the regressions in the literature, based on individual predictor variables, may be spurious.

Abstract:
Previous studies have identified predetermined variables that have some power to explain the time series of stock and bond returns. This paper shows that loadings on the same variables also provide significant cross-sectional explanatory power for stock portfolio returns. These loadings are important, over and the above the variables advocated by Fama and French (1993) in their three factor model,' and also the four factors of Elton, Gruber and Blake (1995). The explanatory power of the loadings on lagged variables is robust to various portfolio grouping procedures and other considerations. The lagged variables reveal information about the cross-section of expected returns that is not captured by popular asset pricing factors. These results carry implications for risk analysis, performance measurement, cost-of-capital calculations and other applications.

Abstract:
Previous studies have identified predetermined variables that have some power to explain the time series of stock and bond returns. This paper shows that loadings on the same variables also provide significant cross-sectional explanatory power for stock portfolio returns. These loadings are important, over and the above the variables advocated by Fama and French (1993) in their three factor "model," and also the four factors of Elton, Gruber and Blake (1995). The explanatory power of the loadings on lagged variables is robust to various portfolio grouping procedures and other considerations. The lagged variables reveal information about the cross-section of expected returns that is not captured by popular asset pricing factors. These results carry implications for risk analysis, performance measurement, cost-of-capital calculations and other applications.

Abstract:
This paper combines the use of portfolio holdings data and conditioning information to create a new performance measure. Our conditional weight-based measure has several advantages. Using conditioning information avoids biases in weight-based measures as discussed by Grinblatt and Titman (1993). When conditioning information is used, returns-based measures face a bias if managers can trade between observation dates. The new measures avoid this interim trading bias. We use the new measures to provide fresh insights about performance in a sample of U.S. equity pension fund managers.

Abstract:
This paper combines the use of portfolio holdings data and conditioning information to create a new performance measure. Our conditional weight-based measure has several advantages. Using conditioning information avoids biases in weight-based measures as discussed by Grinblatt and Titman (1993). When conditioning information is used, returns-based measures face a bias if managers can trade between observation dates. The new measures avoid this interim trading bias. We use the new measures to provide fresh insights about performance in a sample of U.S. equity pension fund managers.

Performance Evaluation, Pension Funds, Conditioning, Portfolio Weights

Abstract:
Recent studies of portfolio performance have separately introduced the use of portfolio holdings and conditioning information. This paper combines these innovations to create a new performance measure. Our conditional weight-based measure has several advantages. By using conditioning information, it can avoid biases in weight-based measures as discussed by Grinblatt and Titman (1993). Using lagged instruments and portfolio weight data, we obtain more precision. Further, when there is conditioning information, returns-based measures (even conditional measures) face a bias if managers can trade between observation dates. The new measures avoid this "interim trading bias". We use the new measure to provide fresh insights about performance in a sample of U.S. equity pension fund managers.

Abstract:
Recent empirical studies use the returns of attribute-sorted portfolios of common stocks as if they represent risk factors in an asset pricing model. If the attributes are chosen following an empirically observed relation to the cross-section of stock returns, such portfolios will appear to be useful risk factors, even when the attributes are completely unrelated to risk. We illustrate this result using a parable and argue that the moral of the story is important in practice.

Asset pricing, Factor models, Cross-sectional regressions, Arbitrage portfolios, Anomalies

Abstract:
We evaluate the performance of fixed income mutual funds using stochastic discount factors from continuous-time term structure models. Time-aggregation of the models for discrete returns generates additional empirical "factors," and these factors contribute significant explanatory power to empirical the models. We provide the first conditional performance evaluation for US fixed income mutual funds, conditioning on a variety of
discrete ex ante characterizations of the state of the term structure and the economy. During 1985-1999 fixed income funds returned less on average than passive benchmarks that don't pay expenses, but not in all economic states. Fixed income funds
typically do poorly when short term interest rates or industrial capacity utilization rates are high, and offer higher returns when quality-related credit spreads are high. We find more heterogeneity across fund styles than across characteristics-based fund groups. Mortgage funds under perform a GNMA index in all economic states. These excess returns are reduced, and typically become insignificant, when we adjust for risk using the stochastic discount factors.

Abstract:
We study stochastic discount factor (SDF) models for evaluating investment performance. Constructing artificial funds with known levels of ability, we find that the measures of performance are not highly sensitive to the SDF model. Most of the models have a mild negative bias when performance is neutral. We evaluate a sample of U.S. equity mutual funds. Adjusting for the observed bias, the average mutual fund has enough ability to cover transactions costs. Extreme funds are more likely to have good rather than poor risk-adjusted performance. Our analysis reveals a number of implementation issues relevant to other applications.

Abstract:
We study the use of stochastic discount factor (SDF) models in evaluating the investment performance of portfolio managers. By constructing artificial mutual funds with known levels of investment ability, we evaluate a large set of SDF models. We find that the measures of performance are not highly sensitive to the SDF model, and that most of the models have a mild negative bias when performance is neutral. We use the models to evaluate a sample of U.S. equity mutual funds. Adjusting for the observed bias, we find that the average mutual fund has enough ability to cover its transactions costs. Extreme funds are more likely to have good rather than poor risk adjusted performance. Our analysis also reveals a number of implementation issues relevant to other applications of SDF models.

Abstract:
We study the use of stochastic discount factor (SDF) models in evaluating the investment performance of portfolio managers. By constructing artificial mutual funds with known levels of investment ability, we evaluate a large set of SDF models. We find that the measures of performance are not highly sensitive to the SDF model, and that most of the models have a mild negative bias when performance is neutral. We use the models to evaluate a sample of U.S. equity mutual funds. Adjusting for the observed bias, we find that the average mutual fund has enough ability to cover its transactions costs. Extreme funds are more likely to have good rather than poor risk adjusted performance. Our analysis also reveals a number of implementation issues relevant to other applications of SDF models.

Performance evaluation, stochastic discount factors, mutual funds, conditional asset pricing

Abstract:
This article shows how to evaluate the performance of managed portfolios using stochastic discount factors (SDFs) from continuous-time term structure models. These models imply empirical factors that include time averages of the underlying state variables. The approach addresses a performance measurement bias, described by Goetzmann, Ingersoll, and Ivkovic (2000) and Ferson and Khang (2002), arising because fund managers may trade within the return measurement interval or hold positions in replicable options. The empirical factors contribute explanatory power in factor model regressions and reduce model pricing errors. We illustrate the approach on US government bond funds during 1986-2000.

Abstract:
This paper evaluates the performance of government bond mutual funds with stochastic discount factors from continuous-time term structure models. The approach addresses the interim trading bias described by Goetzmann, Ingersoll and Ivkovic (2000) and Ferson and Khang (2002). It replaces the ad hoc selection of empirical factors and instruments with variables prescribed by theory. Time-aggregation of the models for discrete returns generates empirical factors that contribute explanatory power to the models, and may be useful in other settings. We provide the first conditional performance evaluation for US fixed income mutual funds. During 1986-2000 most government bond funds returned less on average than passive benchmarks that don't pay expenses, but not for all states of the term structure. The abnormal returns are reduced, and with a few interesting exceptions become insignificant, when we adjust for risk using the stochastic discount factors from term structure models.

Abstract:
We explore the different factors that drive expected returns in world markets. Our research offers two innovations. First, the introduction of the Euro currency unit greatly reduces the complexity of including foreign exchange risk in asset pricing models. We use a synthetic Euro excess return along with a Yen excess return to assess country equity sensitivities to currency risk factors. Second, when combining the currency factors with a group of economic factors, we measure the incremental information in the factor proposed in Fama and French (1998). We find that a global price-to-book factor offers little additional explanatory power over and above a model that includes economic risk factors.

Abstract:
This paper evaluates the ability of US fixed income mutual funds to time common factors related to bond markets. We control for potential non-timing-related sources of nonlinearity in the relation between fund returns and common factors. Such nonlinearities may arise from dynamic trading strategies or derivatives, funds' responses to public information, they may appear in the underlying assets held by the fund, or they may be related to systematic patterns in stale pricing. Controlling for these effects we find that funds returns are more typically concave than convex, in relation to benchmark returns, which suggests negative timing ability. Thus, the evidence is similar to what previous studies have found for equity funds.

Fixed income, mutual funds

Abstract:
We develop asset pricing models’ implications for portfolio efficiency with conditioning information in the form of lagged instruments. A model identifies a portfolio that should be minimum-variance efficient with respect to the conditioning information. Our framework refines tests of portfolio efficiency by using the given conditioning information optimally. The optimal use of the lagged variables is economically important; by using the instruments optimally, we reject several efficiency hypotheses that are not otherwise rejected. The Sharpe ratios of a sample of hedge fund indexes appear consistent with the optimal use of conditioning information.

G11, G12, G23

Abstract:
We develop asset pricing models' implications for portfolio efficiency when there is conditioning information in the form of a set of lagged instruments. A model of expected returns identifies a portfolio that should be minimum variance efficient with respect to the conditioning information. Our tests refine previous tests of portfolio efficiency, using the conditioning information optimally. We reject the efficiency of all static or time-varying combinations of the three Fama-French (1996) factors with respect to the conditioning information and also the conditional efficiency of time-varying combinations of the factors, given standard lagged instruments.

Institutional subscribers to the NBER working paper series, and residents of developing countries may download this paper without additional charge at www.nber.org.

Abstract:
We develop tests of stochastic discount factor models and portfolio efficiency when there is conditioning information, in the form of a set of lagged instruments. In this setting a model identifies a portfolio that should be efficient with respect to the conditioning information. Our tests refine previous tests of portfolio efficiency, and appear substantially more powerful than tests that do not incorporate conditioning information. The tests reject the efficiency of the three Fama-French factors in any fixed-weight combination.

Abstract:
Mimicking portfolios have long been useful in asset pricing research. In most empirical applications, the portfolio weights are assumed to be fixed over time, while in theory they may be functions of the economic state. This paper derives and characterizes mimicking portfolios in the presence of predetermined state variables, or conditioning information. The results generalize and integrate multifactor minimum variance efficiency (Fama, 1996) with conditional and unconditional mean variance efficiency (Hansen and Richard (1987), Ferson and Siegel, 2001). Empirical examples illustrate the potential importance of time-varying mimicking portfolio weights and highlight challenges in their application.

Abstract:
Mimicking portfolios have long been useful in asset pricing research. In most empirical applications, the portfolio weights are assumed to be fixed over time, while in theory they may be functions of the economic state. This paper derives and characterizes mimicking portfolios in the presence of predetermined state variables, or conditioning information. The results generalize and integrate multifactor minimum variance efficiency (Fama, 1996) with conditional and unconditional mean variance efficiency (Hansen and Richard (1987), Ferson and Siegel, 2001).

Abstract:
This paper provides a global asset pricing perspective on the debate over the relation between predetermined attributes of common stocks, such as ratios of price-to-book-value, cash-flow, earnings, and other variables to the future returns. Some argue that such variables may be used to find securities that are systematically undervalued by the market, while others argue that the measures are proxies for exposure to underlying economic risk factors. It is not possible to distinguish between these views without explicitly modelling the relation between such attributes and risk factors. We present an empirical framework for attacking the problem at a global level, assuming integrated markets. Our perspective pulls together the traditional academic and practitioner viewpoints on lagged attributes. We present new evidence on the relative importance of risk and mispricing effects, using monthly data for 21 national equity markets. We find that the cross-sectional explanatory power of the lagged attributes is related to both risk and mispricing in the two-factor model, but the risk effects explain more of the variance than mispricing.

This paper is an unabridged version of our 1997 publication in the Journal of Banking and Finance.

International asset pricing, fundamental ratios, country risk, time-varying risk, valuation

Abstract:
Hansen and Jagannathan (HJ, 1991) describe restrictions on the volatility of stochastic discount factors (SDFs) that price a given set of asset returns. This paper compares the sampling properties of different versions of HJ bounds that use conditioning information in the form of a given set of lagged instruments. HJ describe one way to use conditioning information. Their approach is to multiply the original returns by the lagged variables, and much of the asset pricing literature to date has followed this ihmultiplicativel. approach. We also study two versions of optimized HJ bounds with conditioning information. One is from Gallant, Hansen and Tauchen (1990) and the second is based on the unconditionally-efficient portfolios derived in Ferson and Siegel (2000). We document finite-sample biases in the HJ bounds, where the biased bounds reject asset-pricing models too often. We provide useful correction factors for the bias. We also evaluate the asymptotic standard errors for the HJ bounds, from Hansen, Heaton and Luttmer (1995).

Abstract:
Hansen and Jagannathan (HJ, 1991) describe restrictions on the volatility of stochastic discount factors (SDFs) that price a given set of asset returns. This paper compares the sampling properties of different versions of HJ bounds that use conditioning information in the form of a given set of lagged instruments. HJ describe one way to use conditioning information. Their approach is to multiply the original returns by the lagged variables, and much of the asset pricing literature to date has followed this "multiplicative" approach. We also study two versions of optimized HJ bounds with conditioning information. One is from Gallant, Hansen and Tauchen (1990) and the second is based on the unconditionally-efficient portfolios derived in Ferson and Siegel (2000). We document finite-sample biases in the HJ bounds, where the biased bounds reject asset-pricing models too often. We provide useful correction factors for the bias. We also evaluate the asymptotic standard errors for the HJ bounds, from Hansen, Heaton and Luttmer (1995).

volatility bounds, stochastic discount factors, finite sample bias, asset pricing, return

Abstract:
This paper evaluates the ability of bond funds to "market time" nine common factors related to bond markets. Timing ability generates nonlinearity in fund returns as a function of common factors, but there are several non-timing-related sources of nonlinearity. Controlling for the non-timing-related nonlinearity is important. Funds' returns are more concave than benchmark returns, and this would appear as poor timing ability in naive models. With controls, the timing coefficients appear neutral to weakly positive. Adjusting for nonlinearity the performance of many bond funds is significantly negative on an after-cost basis, but significantly positive on a before-cost basis.

Mutual funds, Market timing, Bond funds, Investment performance evaluation

Abstract:
This paper evaluates the ability of bond funds to market time nine common factors related to bond markets. Timing ability generates nonlinearity in fund returns as a function of common factors, but there are several non-timing-related sources of nonlinearity. Controlling for the non-timing-related nonlinearity is important. Funds' returns are more concave than benchmark returns, and this would appear as poor timing ability in naive models. With controls, the timing coefficients appear neutral to weakly positive. Adjusting for nonlinearity the performance of many bond funds is significantly negative on an after-cost basis, but significantly positive on a before-cost basis.

Institutional subscribers to the NBER working paper series, and residents of developing countries may download this paper without additional charge at www.nber.org.

Abstract:
Three concepts: stochastic discount factors, multi-beta pricing and mean variance efficiency, are at the core of modern empirical asset pricing. This paper reviews these paradigms and the relations among them, concentrating on conditional asset pricing models where lagged variables serve as instruments for publicly available information. The different paradigms are associated with different empirical methods. We review the variance bounds of Hansen and Jagannathan (1991), concentrating on extensions for conditioning information. Hansen's (1982) Generalized Method of Moments (GMM) is briefly reviewed as an organizing principle. Then, cross-sectional regression approaches as developed by Fama and MacBeth (1973) are reviewed and used to interpret empirical factors, such as those advocated by Fama and French (1993, 1996). Finally, we review the multivariate regression approach, popularized in the finance literature by Gibbons (1982) and others. A regression approach, with a beta pricing formulation, and a GMM approach with a stochastic discount factor formulation, may be considered competing paradigms for empirical work in asset pricing. This discussion clarifies the relations between the various approaches. Finally, we bring the models and methods together, with a review of the recent conditional performance evaluation literature, concentrating on mutual funds and pension funds.

Institutional subscribers to the NBER working paper series, and residents of developing countries may download this paper without additional charge at www.nber.org.

Abstract:
This paper empirically examines multifactor asset pricing models for the returns and expected returns on eighteen national equity markets. The factors are chosen to measure global economic risks. Although previous studies do not reject the unconditional mean- variance efficiency of a world market portfolio, our evidence indicates that the tests are low in power, and the world market betas do not provide a good explanation of cross-sectional differences in average returns. Multiple beta models provide an improved explanation of the equity returns.

Abstract:
This paper makes indirect inference about the time-variation in expected stock returns by comparing unconditional sample variances to estimates of expected conditional variances. The evidence reveals more predictability as more information is used, and no evidence that predictability has diminished in recent years. Semi-strong form evidence suggests that time-variation in expected returns remains economically important.

Institutional subscribers to the NBER working paper series, and residents of developing countries may download this paper without additional charge at www.nber.org.

Abstract:
This paper makes indirect inference about the time-variation in expected stock returns by comparing unconditional sample variances to estimates of expected conditional variances. The evidence reveals more predictability as more information is used, and no evidence that predictability has diminished in recent years. Semi-strong form evidence suggests that time-variation in expected returns remains economically important.

Institutional subscribers to the NBER working paper series, and residents of developing countries may download this paper without additional charge at www.nber.org.

Abstract:
Habit persistence in consumption preferences and durability of consumption goods are two hypotheses which imply time-nonseparability in the derived utility for consumption expenditures. We study a simple model with both effect, in which lagged consumption expenditures enter the Euler equation. Habit persistence implies that the coefficients on the lagged expenditures are negative, while durability implies positive coefficients. If both effects are present, then estimating the sign of the coefficients addresses the question as to which of the two effects is dominant. Earlier empirical work on monthly data supported the durability of consumption expenditures. We estimate and test the Euler equation using monthly, quarterly and annual data and find evidence that habit persistence dominates the effect of durability.

Abstract:
This paper tests models of mutual fund market timing that (1) allow the manager's utility function to depend on returns in excess of a benchmark; (2) distinguish timing based on lagged, publicly available information variables from timing based on finer information; and (3) simultaneously estimate the parameters which describe the public information environment, the risk aversion and the precision of the fund's market timing signal. Using a sample of more than 400 U.S. mutual funds for 1976-94, the estimates imply that mutual funds behave as risk averse, benchmark investors. Conditioning on public information variables improves the model specification, and after controlling for the public information we find no evidence that funds have significant market timing ability.

Abstract:
This paper studies the ability of long-run risk models, following Bansal and Yaron (2004) and others, to explain out-of-sample asset returns associated with the equity premium puzzle, size and book/market effects, momentum, reversals, and bond returns of different maturity and credit ratings. We examine stationary and nonstationary versions of the models using annual data for 1931-2006. Unlike the in-sample and calibration evidence of previous studies, we find that the models perform no better overall than the simple CAPM. For some effects the long-run risk models deliver smaller average pricing errors, but they have larger error variances than the CAPM, and the mean squared prediction errors are similar. One exception is the Momentum effect, where the long run risk models perform relatively well.

Long-run risk models, Out-of-sample, Equity premium puzzle, Size effect, Book to market effect, Momentum, Reversals, Term premium, Default premium

Abstract:
This paper evaluates persistence in the performance of institutional equity managers. We build on recent work on conditional performance evaluation, using time-varying conditional expected returns and risk measures. We find evidence that the investment performance of pension fund managers persists over time. A conditional approach is able to better detect this persistence and to predict the future performance of the funds than are traditional methods. The performance persistence is especially concentrated in the managers with negative prior-period conditional alphas. "

Abstract:
This paper presents evidence on persistence in the relative investment performance of large, institutional equity managers. Similar to existing evidence for mutual funds, we find persistent performance, concentrated in the managers with poor prior-period performance measures. A conditional approach, using time-varying measures of risk and abnormal performance is better able to detect this persistence and to predict the future performance of the funds than are traditional methods.

Abstract:
This paper studies the estimation of asset pricing model regressions with conditional alphas and betas, focusing on the joint effects of data snooping and spurious regression. We find that the regressions are reasonably well specified for conditional betas, even in settings where simple predictive regressions are severely biased. However, there are biases in estimates of the conditional alphas. When time-varying alphas are suppressed and only time-varying betas are considered, the betas become baised. Previous studies overstate the significance of time-varying alphas.

Abstract:
We explore the different factors that drive expected returns in world markets. Our research offers two innovations. First, the introduction of the Euro currency unit greatly reduces the complexity of including foreign exchange risk in asset pricing models. We use a synthetic Euro excess return along with a Yen excess return to assess country equity sensitivities to currency risk factors. Second, when combining the currency factors with a group of economic factors, we measure the incremental information in the factor proposed in Fama and French (1998). We find that a global price-to-book factor offers little additional explanatory power over and above a model that includes economic risk factors.

Abstract:
This paper provides a global asset pricing perspective on the debate over the relation between predetermined attributes of common stocks, such as ratios of price-to-book-value, cash-flow, earnings, and other variables to the future returns. Some argue that such variables may be used to find securities that are systematically undervalued by the market, while others argue that the measures are proxies for exposure to underlying economic risk factors. It is not possible to distinguish between these views without explicitly modeling the relation between such attributes and risk factors. We present an empirical framework for attacking the problem at a global level, assuming integrated markets. Our perspective pulls together the traditional academic and practitioner viewpoints on lagged attributes. We present new evidence on the relative importance of risk and mispricing effects, using monthly data for 21 national equity markets. We find that the cross-sectional explanatory power of the lagged attributes is related to both risk and mispricing in the two-factor model than mispricing.

Abstract:
This paper studies average and conditional expected returns in national equity markets, and their relation to a number of fundamental country attributes. The attributes are organized into three groups. The first is relative valuation ratios, such as price-to-book-value, cash-flow, earnings and dividends. The second group measures relative economic performance and the third measures industry structure. We find that average returns across countries are related to the volatility of their price-to-book ratios. Predictable variation in returns is also related to relative gross domestic product, interest rate levels and dividend-price ratios. We explore the hypothesis that cross-sectional variation in the country attributes proxy for variation in the sensitivity of national markets to global measures of economic risks. We test single-factor and two-factor models in which countries' conditional betas are assumed to be functions of the more important fundamental attributes.

Abstract:
We study consumption-based asset pricing models which allow for both habit persistence and durability of consumption goods, using quarterly consumption and asset return data for six countries. We estimate the parameters representing habit persistence or durability, risk aversion and time preference for each of the countries. We find that time-nonseparable preferences improve the fit of the model. When the nonseparability parameter is statistically significant, its magnitude indicates that the effect of habit persistence dominates the effect of durability in consumption expenditures. However, the international evidence for habit persistence is weaker than it is for the United States. The results indicate that the simple model of time nonseparability does not provide a satisfactory explanation of consumption and asset returns.

Abstract:
Two problems, spurious regression bias and naive data mining, conspire to mislead analysts about predictive models for stock returns. This article demonstrates the two problems, how they interact, and makes suggestions for what to do about it.

Dividend yield, valuation ratios, time series, yield spreads, predicting stock returns, asset allocation, market timing, active portfolio management

Abstract:
This article studies predictability in U.S. stock returns for multiple investment horizons. We measure to what extent predictability is driven by premiums for economy-wide risk factors, comparing two standard methods for factor selection. We study single-beta models and multiple-beta models. We show how to estimate the fraction of the predictability in returns captured by the model, simultaneously with the other parameters. Our analysis indicates that the models capture a large fraction of the predictability for all of the investment horizons. The performances of the principal components and the prespecified-factor approaches are broadly similar.

Abstract:
The use of predetermined variables to represent public information and time-variation has produced new insights about asset pricing models, but the literature on mutual fund performance has not exploited these insights. This paper advocates conditional performance evaluation, in which the relevant expectations are conditioned on public information variables. We modify several classical performance measures to this end, and find that the predetermined variables are both statistically and economically significant. Conditioning on public information controls for biases in traditional market timing models, and makes the average performance of the mutual funds in our sample look better.