In-sample vs. out-of-sample tests of stock return predictability in the context of data mining

We undertake an extensive analysis of in-sample and out-of-sample tests of stock return predictability in an effort to better understand the nature of the empirical evidence on return predictability. We find that a number of financial variables appearing in the literature display both in-sample and out-of-sample predictive ability with respect to stock returns in annual data covering most of the twentieth century. In contrast to the extant literature, we demonstrate that there is little discrepancy between in-sample and out-of-sample test results once we employ out-of-sample tests with good power. While conventional wisdom holds that out-of-sample tests help guard against data mining, Inoue and Kilian [Inoue, A., Kilian, L., 2004. In-sample or out-of-sample tests of predictability: which one should we use? Econometric Reviews 23, 371–402.] recently argue that in-sample and out-of-sample tests are equally susceptible to data mining biases. Using a bootstrap procedure that explicitly accounts for data mining, we still find that certain financial variables display significant in-sample and out-of-sample predictive ability with respect to stock returns.     

Do jumps contribute to the dynamics of the equity premium?

This paper investigates whether risks associated with time-varying arrival of jumps and their effect on the dynamics of higher moments of returns are priced in the conditional mean of daily market excess returns. We find that jumps and jump dynamics are significantly related to the market equity premium. The results from our time-series approach reinforce the importance of the skewness premium found in cross-sectional studies using lower-frequency data; and offer a potential resolution to sometimes conflicting results on the intertemporal risk-return relationship. We use a general utility specification, consistent with our pricing kernel, to evaluate the relative value of alternative risk premium models in an out-of-sample portfolio performance application.     

The Predictability of Equity REIT Returns: Time Variation and Economic Significance

This article presents evidence on predictability of excess returns for equity REITs relative to the aggregate stock market, small-capitalization stocks, and T-bills using best-fit models from prior time periods. We find that excess equity REIT returns are far less predictable out-of-sample than in-sample. This inability to forecast out-of-sample is particularly true in the 1990s. Nevertheless, in the absence of transaction costs, active-trading strategies based on out-of-sample predictions modestly outperform REIT buy-and-hold strategies. However, when transaction costs are introduced, profits from these active-trading strategies largely disappear.     

Exchange rate return predictability in times of geopolitical risk

We develop the hypothesis that geopolitical risk predicts exchange rate returns. Using data on 17 countries, we demonstrate that the information content embedded in geopolitical risk is economically useful and can improve the forecast accuracy of exchange rate returns. We show that geopolitical risk predicts 10 out of 17 (59%) exchange rate returns in in-sample tests while in out-of-sample tests predictability is found for 88% of currencies. Buy and sell signals generated from our model lead to higher returns compared to the historical average model. Our model delivers excess profits relative to the benchmark model in 11 out of 17 (65%) currencies.     

The predictability of aggregate returns on commodity futures

This paper provides evidence that aggregate returns on commodity futures (without the returns on collateral) are predictable, both in-sample and out-of-sample, by various lagged variables from the stock market, bond market, macroeconomics, and the commodity market. Out of the 32 candidate predictors we consider, we find that investor sentiment is the best in-sample predictor of short-horizon returns, whereas the level and slope of the yield curve have much in-sample predictive power for long-horizon returns. We find that it is possible to forecast aggregate returns on commodity futures out-of-sample through several combination forecasts (the out-of-sample return forecasting R² is up to 1.65% at the monthly frequency).     

Forecasting returns: New European evidence

This paper builds on the recent debate on the in-sample and out-of-sample predictability of US aggregate returns using a wide range of predictors by providing new evidence for smaller and less market-oriented European countries.We find evidence that macro and technical predictors can (statistically) improve forecast accuracy and (economically) generate gains to investors; in contrast to the US results, predictability in our sample of European countries exists in recent data. We also find that simple forecast combinations consistently yield substantial benefits both in forecast accuracy and economic gain. For example, the magnitude of the forecasting gains for our European countries is often larger than those found for the US and other G7 countries. We provide initial evidence on the link between country characteristics and out-of-sample forecast performance. Our empirical results indicate that market development is related to the forecast performance of macro variables. There is also some evidence that forecast performance is related to market size and liquidity.     

Developing a Trading Rule from the FTSE-100 Stock Index Futures Contract: Evidence in Support of the EMH

An extensive literature documents the predictability of both short and long horizon returns, over a wide range of sample periods, frequencies and markets. This predictability may represent weak form inefficiency, or it may be caused by a failure to account for a time-variation in risk. We develop statistically reliable ex ante models of the returns on the FTSE-100 stock index futures contract and test a simple trading rule based on the out-of-sample predictions from these models. We interpret the failure of our ex ante model to produce abnormal returns for a risk neutral investor as evidence in favour of the EMH. Our trading rule results clearly suggest that we should be careful in interpreting such ex ante models as evidence of financial market inefficiency.     

Reduction of estimation risk in optimal portfolio choice using redundant constraints

It is well known that when the moments of the distribution governing returns are estimated from sample data, the out-of-sample performance of the optimal solution of a mean–variance (MV) portfolio problem deteriorates as a consequence of the so-called “estimation risk”. In this document we provide a theoretical analysis of the effects caused by redundant constraints on the out-of-sample performance of optimal MV portfolios. In particular, we show that the out-of-sample performance of the plug-in estimator of the optimal MV portfolio can be improved by adding any set of redundant linear constraints. We also illustrate our findings when risky assets are equally correlated and identically distributed. In this specific case, we report an emerging trade-off between diversification and estimation risk and that the allocation of estimation risk across portfolios forming the optimal solution changes dramatically in terms of number of assets and correlations.     

Linear predictability vs. bull and bear market models in strategic asset allocation decisions: evidence from UK data

Most papers in the portfolio choice literature have examined linear predictability frameworks based on the idea that simple but flexible Vector Autoregressive (VAR) models can be expanded to produce portfolio allocations that hedge against the bull and bear dynamics typical of financial markets through careful selection of predictor variables that capture business cycles and market sentiment. Yet, a distinct literature exists that shows that non-linear econometric frameworks, such as Markov switching, are also natural tools to compute optimal portfolios arising from the existence of good and bad market states. This paper examines whether and how simple VARs can produce portfolio rules similar to those obtained under a simple Markov switching, by studying the effects of expanding both the order of the VAR and the number/selection of predictor variables included. In a typical stock-bond strategic asset allocation problem for UK data, we compute the out-of-sample certainty equivalent returns for a wide range of VARs and compare these measures of performance with those of non-linear models. We conclude that most VARs cannot produce portfolio rules, hedging demands or (net of transaction costs) out-of-sample performances that approximate those obtained from simple non-linear frameworks.     

Dynamic copula models and high frequency data

This paper proposes a new class of dynamic copula models for daily asset returns that exploits information from high frequency (intra-daily) data. We augment the generalized autoregressive score (GAS) model of Creal et al. (2013) with high frequency measures such as realized correlation to obtain a “GRAS” model. We find that the inclusion of realized measures significantly improves the in-sample fit of dynamic copula models across a range of U.S. equity returns. Moreover, we find that out-of-sample density forecasts from our GRAS models are superior to those from simpler models. Finally, we consider a simple portfolio choice problem to illustrate the economic gains from exploiting high frequency data for modeling dynamic dependence.     

Stock return forecasting: Some new evidence

This paper makes three contributions to the literature on forecasting stock returns. First, unlike the extant literature on oil price and stock returns, we focus on out-of-sample forecasting of returns. We show that the ability of the oil price to forecast stock returns depends not only on the data frequency used but also on the estimator. Second, out-of-sample forecasting of returns is sector-dependent, suggesting that oil price is relatively more important for some sectors than others. Third, we examine the determinants of out-of-sample predictability for each sector using industry characteristics and find strong evidence that return predictability has links to certain industry characteristics, such as book-to-market ratio, dividend yield, size, price earnings ratio, and trading volume.     

Forecasting forward defaults: a simple hazard model with competing risks

A forward default prediction method based on the discrete-time competing risk hazard model (DCRHM) is proposed. The proposed model is developed from the discrete-time hazard model (DHM) by replacing the binary response data in DHM with the multinomial response data, and thus allowing the firms exiting public markets for different causes to have different effects on forward default prediction. We show that DCRHM is a reliable and efficient model for forward default prediction through maximum likelihood analysis. We use actual panel data-sets to illustrate the proposed methodology. Using an expanding rolling window approach, our empirical results statistically confirm that DCRHM has better and more robust out-of-sample performance than DHM, in the sense of yielding more accurate predicted number of forward defaults. Thus, DCRHM is a useful alternative for studying forward default losses on portfolios.     

Cryptocurrency returns under empirical asset pricing

This study examines the predictability of cryptocurrency returns based on investors' risk premia. Prior studies that have examined the predictability of cryptocurrencies using various economic risk factors have reported mixed results. Our out-of-sample evidence identifies the existence of a significant return predictability of cryptocurrencies based on the cryptocurrency market risk premium. Consistent with capital asset pricing theory (CAPM), our results show that investors often require higher positive returns before taking on any additional risks, particularly in terms of riskier assets like cryptocurrencies. Tests involving the CAPM model demonstrates that the three largest cryptocurrencies have significant exposures to the proposed market factor with insignificant intercepts, demonstrating that the market factor explains average cryptocurrency returns very well.     

Stock return expectations in the credit market

In this paper we compute long-term stock return expectations (across the business cycle) for individual firms using information backed out from the credit derivatives market. Our methodology builds on previous theoretical results in the literature on stock return expectations and, empirically, we demonstrate a close relationship between credit-implied stock return expectations and future realized stock returns. We also find stock portfolios selected based on credit-implied stock return forecasts to beat equally- and value-weighted portfolios of the same stocks out-of-sample. Contrary to many other studies, our expectations/predictions are made at the individual stock level rather than at the portfolio level, and no parameter estimations using historical stock price- or credit spread observations are needed.     

Housing returns,precautionary savings and consumption: Micro evidence from China

We investigate how changes in home prices affect consumption in China via a wealth channel. Examining a panel of 7955 households via fixed effects and instrumental variable methods, we find a marginal propensity to consume out of housing wealth (home-price MPC) that is concentrated on goods consumed for pleasure rather than necessity. This trend is driven by the value of second homes rather than that of primary residences, suggesting a wealth channel. We further examine whether returns on housing investment, including rental income and home appreciation, fund the wealth channel; however, we find little supporting evidence. In contrast, a reduction in health risk increases the home-price MPC, but a reduction in income risk that also relieves precautionary saving motives does not. Our results are robust to alternative data, common-factor progress, expenditure shocks and bequest motives. We contribute by examining second homes, which carry little of the dual nature of housing that primary residences do, to identify a controversial wealth channel, and by studying the relative effects of health and income risks on the wealth channel.     

Exchange rates,interventions, and the predictability of stock returns in Japan

Using monthly Japanese data for the period 1991–2005, we examined the link between exchange rate movements and stock returns. We found that exchange rate movements per se do not help to explain stock returns. There is, however, evidence of in-sample predictability if one accounts for the interventions of the Japanese monetary authorities in the foreign exchange market. This evidence does not indicate a violation of market efficiency insofar as investors cannot use information on interventions to systematically improve the performance of simple trading rules based on out-of-sample forecasts of stock returns.     

Which past returns affect trading volume?

Anecdotal evidence suggests and recent theoretical models argue that past stock returns affect subsequent stock trading volume. We study 3,000 individual investors over a 51 month period to test this apparent link between past returns and volume using several different panel regression models (linear panel regressions, negative binomial panel regressions, Tobit panel regressions). We find that both past market returns as well as past portfolio returns affect trading activity of individual investors (as measured by stock portfolio turnover, the number of stock transactions, and the propensity to trade stocks in a given month). After high portfolio returns, investors buy high risk stocks and reduce the number of stocks in their portfolio. High past market returns do not lead to higher risk taking or underdiversification. We argue that the only explanations for our findings are overconfidence theories based on biased self-attribution and differences of opinion explanations for high levels of trading activity.     

A nesting framework for Markov-switching GARCH modelling with an application to the German stock market

In this paper, we establish a generalized two-regime Markov-switching GARCH model which enables us to specify complex (symmetric and asymmetric) GARCH equations that may differ considerably in their functional forms across the two Markov regimes. We show how previously proposed collapsing procedures for the Markov-switching GARCH model can be extended to estimate our general specification by means of classical maximum-likelihood methods. We estimate several variants of the generalized Markov-switching GARCH model using daily excess returns of the German stock market index DAX sampled during the last decade. Our empirical study has two major findings. First, our generalized model outperforms all nested specifications in terms of (a) statistical fit (when model selection is based on likelihood ratio tests) and (b) out-of-sample volatility forecasting performance. Second, we find significant Markov-switching structures in German stock market data, with substantially differing volatility equations across the regimes.     

Stock and bond market interactions with level and asymmetry dynamics: An out-of-sample application

We model the dynamic interaction between stock and bond returns using a multivariate model with level effects and asymmetries in conditional volatility. We examine the out-of-sample performance using daily returns on the S&P 500 index and 10 year Treasury bond. We find evidence for significant (cross-) asymmetries in the conditional volatility and level effects in bond returns. The out-of-sample covariance matrix forecasts of the model imply that an investor is willing to pay between 129 and 820 basis points per year for using a dynamic trading strategy instead of a passive strategy.     

The change in stock-selection risk and stock market returns

Following Jiang et al. (2021), who propose a stock-selection opportunity (SSO) measurement by the absolute average positive alpha of individual stocks to reflect stock-selection timing, we construct a stock-selection risk (SSR) measure from the perspective of negative alphas of individual stocks. Then, we investigate the predictive abilities of SSO, SSR, the change of SSO (CSSO), and the change of SSR (CSSR) on stock market returns. By using data from 2003 to 2020 in China, we find that only CSSR significantly predicts future one-month market returns. Moreover, considering other popular predictors, our in-sample and out-of-sample results and a series of robustness checks support the proposal that CSSR provides unique information for predicting market returns that reduces forecast errors and increases economic value for investors. Furthermore, our trading activity evidence shows that CSSR predicts stock market returns due to investors' underreaction to the information of CSSR.