Journal of Risk

Farid AitSahlia
Warrington College of Business, University of Florida

In the first paper in this issue, “The informativeness of risk factor disclosures: estimating the covariance matrix of stock returns using similarity measures”, Lukas Tilmann and Martin Walther use textual analysis to extract information from regulatory disclosures such as 10-K and 10-Q filings to compute cosine similarity measures. These measures are then used to estimate the covariance matrix of stock returns. The authors show that combining their approach with shrinkage estimators is superior to relying solely on returns data.

In “Uncovering the hidden impact: noninvestor disagreement and its role in asset pricing”, the issue’s second paper, Tingli Liu, JiaNing Liu, Junjun Ma and Yafei Tai make use of social media data generated by individuals who do not trade securities but who do express views on products and services provided by firms. Liu et al particularly capitalize on disagreements, showing empirically that these can capture an additional factor that affects securities returns.

Our third paper, “An approach to capital allocation based on mean conditional value-at-risk” by Yuecai Han, Fengtong Zhang and Xinyu Liu, proposes an alternative to the standard Euler approach for capital allocation. This approach is nonparametric and is based on adjusting the associated probability level in order to set the total capital to the quantile-based level. Using bootstrapped estimates, Han et al

are better than those based on value-at-risk alone. In the final paper in this issue, “Using a skewed exponential power mixture for value-at-risk and conditional value-at-risk forecasts to comply with market risk regulation”, Samir Saissi Hassani and Georges Dionne propose the use of a mixture of skewed exponential densities to estimate regulatory value-at-risk and conditional value-at-risk levels. Through the implementation of their procedure, which they show to be efficient and robust, they highlight the importance of scoring functions for backtesting in the progressive selection of models in order to address the regulatory requirement of ex post suitability.