Journal of Risk

Farid AitSahlia
Warrington College of Business, University of Florida

Predicting stock returns is an exercise that is fraught with difficulty, so the search for more accurate methods continues. This issue of The Journal of Risk offers two papers that suggest new approaches to this topic. They are complemented by a paper addressing the prediction of credit risk deterioration for major firms and another exploiting prudence and temperance in investor behavior for better portfolio selection.

In “Future portfolio returns and the VIX term structure”, the first paper in this issue, David Yechiam Aharon and Thomas Dimpfl propose to augment the predictive power of the standard implied volatility gauge, namely the CBOE VIX, by introducing the long-minus-short (LMS) implied volatility measure, which captures the difference between implied volatilities for, respectively, long (six-month) and short (nine-day) maturities. Through an empirical study, they show that this measure is better suited to predicting broad market returns, such as the Standard & Poor’s 500 (S&P 500) index, 60 to 90 days out than 30 days out.

In the issue’s second paper, “Application of the moving Lyapunov exponent to the S&P 500 index to predict major declines”, Stefanos Tsakonas, Michael Hanias, Lykourgos Magafas and Loukas Zachilas rely on physics concepts used to study equilibrium states in complex dynamic systems, where small perturbations can lead to extreme outcomes, to predict significant downturns in financial markets. In particular, they make use of the Lyapunov exponent for chaotic dynamics to detect phase transitions. In their empirical illustration, they show that major spikes in the moving average of the Lyapunov exponent tend to precede significant drops in the S&P 500 index.

In our third paper, “A new approach to detecting change in credit quality”, Rusudan Kevkhishvili proposes a relatively manageable model to monitor the creditworthiness of a well-established firm based on a mean-reverting (Ornstein–Uhlenbeck) process capturing the evolution of the (log) ratio of the firm’s debt to capital. In contrast with models focused on credit deterioration associated with default events, this approach is centered on identifying shifts in the long-run mean of this leverage process endogenously. With an empirical study focused on three well-known firms, Kevkhishvili addresses the practical implementation of this approach, especially with respect to the estimation of the parameters of the model.

The fourth paper, “Detecting prudence and temperance in risk exposure: the hybrid variance framework” by Jun Gao, Xiang Gao, Xiaoli Liu and Zhan Wang, completes the issue. The authors develop a portfolio selection approach that accounts formoments higher than the second moment through the concepts of prudence and temperance. The former concept captures an investor’s attitude regarding the impact of uncertain income on savings, and the latter captures the investor’s assessment of the portfolio’s risk in relation to that of its components. Gao et al introduce a risk measure called hybrid variance (HV), which incorporates the impact of both prudence and temperance, in order to compute an investor’s risk aversion. They show empirically that HV is negatively correlated with short-term equity returns and positively correlated with long-term equity returns. They also provide evidence that HV-based portfolios can outperform a market index while providing protection during market downturns.