Manuela Pedio

Regime Shifts in Excess Stock Return Predictability: An Out-of-Sample Portfolio Analysis

The authors analyze the out-of-sample performance of asset allocation decisions based on financial ratio predictability of aggregate stock market returns under linear and regime-switching models. The authors adopt both a statistical perspective to analyze whether models based on valuation ratios can forecast excess equity returns, and an economic approach that turns predictions into portfolio strategies. These consist of a portfolio switching approach, a mean-variance framework, and a long-run dynamic model. The authors find a disconnect between the statistical perspective, whereby the ratios yield a modest forecasting power, and a portfolio approach, by which a moderate predictability is often sufficient to yield significant portfolio outperformance, especially before transaction costs and when regimes are taken into account. However, also when regimes are considered, predictability gives high payoffs only to long horizon, highly risk-averse investors. Moreover, different strategies deliver different performance rankings across predictors.

Practical Applications of Regime Shifts in Excess Stock Return Predictability: An Out-of-Sample Portfolio Analysis

Practical Applications Summary In Regime Shifts in Excess Stock Return Predictability: An Out-of-Sample Portfolio Analysis, published in the Winter 2018 issue of The Journal of Portfolio Management, Giulia Dal Pra, Massimo Guidolin, Manuela Pedio, and Fabiola Vasile analyze whether one can outperform the stock market using asset allocation strategies based on certain financial ratios as signals. They construct linear regressions relating excess returns to selected ratios. They explore both ordinary linear regressions and ones that embody a Markov-switching (regime switching) feature. They assess performance of the strategies over several investment time horizons and include the effect of transaction costs. The authors also consider the efficacy of their strategies in an economic framework that accounts for an investor’s degree of risk aversion.

Forecasting Commodity Futures Returns: An Economic Value Analysis of Macroeconomic vs. Specific Factors

We test whether three well-known commodity-specific variables (basis, hedging pressure, and momentum) may improve the predictive power for commodity futures returns of models otherwise based on macroeconomic factors. We compute recursive, out-of-sample forecasts for fifteen monthly commodity futures return series, when estimation is based on a stepwise regression approach under a probability-weighted regime-switching regression that identifies different volatility regimes. Comparisons with an AR(1) benchmark show that the inclusion of commodity-specific factors does not improve the forecasting power. We perform a back-testing exercise of a meanvariance investment strategy that exploits any predictability of the conditional risk premium of commodities, stocks, and bond returns, also taking into account transaction costs caused by portfolio rebalancing. The risk-adjusted performance of this strategy does not allow us to conclude that any forecasting approach outperforms the others. However, there is evidence that investment strategies based on commodity-specific predictors outperform the remaining strategies in the high-volatility state.

Monetary Policy after the Crisis: Threat or Opportunity to Hedge Funds' Alphas?

We examine the effects of U.S. monetary policy announcements during and after the Great Financial Crisis on the average abnormal returns (the “alpha”) of the hedge fund industry as a whole and of a range of hedge strategy indices. We apply a variety of tests of increasing sophistication including simple event studies, formal tests for breaks, and Markov switching models. The event studies show that both the overall index and longshort equity and fixed income arbitrage hedge strategies were systematically affected by unexpected monetary policy announcements while other strategies appear to have been less impacted. Formal break point tests show that for all but one strategies as well as the overall index, there is evidence of five breakpoints. For the overall index and most of the sub-indices many of the endogenously determined breaks closely match a list of policy surprise dates that have been already singled out because they had strongly affected financial markets in general. Especially for the long-short equity, fixed income arbitrage, dedicated short-bias, and global macro hedge funds, there is a significant tendency for estimated alphas decline over time, following policy surprises.

Estimating stochastic discount factor models with hidden regimes: Applications to commodity pricing

We develop new likelihood-based methods to estimate factor-based Stochastic Discount Factors (SDF) that may accommodate Hidden Markov dynamics in the factor loadings. We use these methods to investigate whether it is possible to find a SDF that jointly prices the cross-section of eight U.S. portfolios of stocks, Treasuries, corporate bonds, and commodities. In particular, we test a range of possible different specification of the SDF, including single-state and Hidden Markov models and compare their statistical and pricing performances. In addition, we assess whether and to which extent a selection of these models replicates the observed moments of the return series, and especially correlations. We report that regime-switching models clearly outperform single-state ones both in term of statistical and pricing accuracy. However, while a four-state model is selected by the information criteria, a two-state three-factor full Vector Autoregression model outperforms the others as far as the pricing accuracy is concerned. Key words: Finance, Commodities, Stochastic Discount Factor, Hidden Markov model.

The Background: Channels of Contagion in the US Financial Crisis

In this introductory chapter, we provide the background material that will help a reader to understand the findings presented in the rest of the book. First, we provide a review of the key facts that characterized the onset and the unfolding of the 2007–09 US financial crisis. In addition, we explore different definitions offinancial contagion and discuss the key papers published on the topic because they provide benchmarks to our empirical analysis.

Estimating and Disentangling the Contagion Channels

In this chapter, we present the key empirical results of this book. First, we explain how our estimation exercises in Chapters 4 and 5 are used to identify different financial contagion channels. Second, we discuss the results and their economic as well as policy implications. We start by examining how and whether contagious patterns across different markets are characterized by regime dynamics. We then quantify the existence and extent of each contagion channel.

Comparing the US and European Contagion Experiences

In this chapter, we extend our work to European data, addressing two key questions. First, we investigate whether contagion channels similar to the ones observed in the US subprime crisis were also active during the European sovereign crisis. Second, we assess whether, in addition to cross-market contagion, there is evidence of cross-country contagion from US to Europe during the 2007–09 financial crisis, that is, whether shocks to low-grade ABS did spill over to European fixed income and equity markets, and through which channels.

Estimates of Single-State VAR Models

In this chapter, we estimate two single-state VAR models, one for the yield and one for the spread series described in Chapter 3. First, we describe our model specification search and identify two as the appropriate number of lags of the VAR for both the VAR-yield and the VAR-spread models. Second, we report the results of the estimation of the two VAR(2) models and comment on the significance and economic meaning of the coefficients.

Results from Markov Switching Models

In a symmetrical approach to what we do in Chapter 4, in this chapter, we estimate two MSVAR models, one for the yields and one for the spreads. First, we present our specification search, which in this case concerns not only the appropriate number of lags, but also the most adequate type of Markov switching model and the appropriate number of regimes. We identify the MSIH(3,1) model as the most adequate to fit the data. Furthermore, we try to provide an economic interpretation of the three regimes that we identify. Finally, we discuss the results of the estimation of these models and their economic interpretation.