Warrington College of Business, University of Florida
The procyclicality of risk-based margins reemerged during the Covid-19 pandemic and is addressed in this issue of The Journal of Risk. Other topics included in this issue regard the customization of risk measures, their nonparametric derivation on the basis of subjective estimates, and the determination of a common factor in correlated idiosyncratic volatilities.
In the first paper of this issue, “Performance measures adjusted for the risk situation (PARS)”, Christoph Peters and Roland C. Seydel propose an alternative risk measure to address the fact that different decision makers have different levels of risk aversion when considering risk. While PARS generalizes performance risk assessments such as tracking error relative to a benchmark, Sharpe ratios and several others, Peters and Seydel illustrate its application in a case study involving a sovereign debt manager (the German finance agency).
In our second paper, “Correlated idiosyncratic volatility shocks”, Xiao Qiao and Yongning Wang propose a variant of the classical GARCH model, which they call dynamic factor correlation (DFC), in order to capture a common factor for correlated time-varying idiosyncratic volatilities. They derive a closed-form likelihood function that helps with efficient parameter estimation. Through an empirical study based on standard sorted portfolios, they show that DFC is more accurate than existing alternatives that model time-varying volatilities.
The issue’s third paper, “A numerical approach to the risk capital allocation problem” by Henryk Gzyl and Silvia Mayoral, considers an approach to capital allocation that captures subjective risk estimates based on experts’ upper and lower bounds. Theirs is a model-free and nonparametric method that maximizes entropy in the mean in order to both determine capital allocation without an explicit risk measure and simultaneously infer one from a collection of prices.
In the last paper in this issue, “Procyclicality control in risk-based margin models”, Lauren W.Wong and Yang Zhang address the common use of rigid buffers to prevent risk-based margins from becoming too low in periods of market turmoil. They propose a more dynamic and efficient alternative that smooths the transition between cycles based on a scaling factor, and they illustrate its advantage over standard autoregressive and smoothing approaches.
This paper proposes the use of a new class of performance measures adjusted for the risk situation (PARS), as the perception of risk depends on the individual situation including risk preferences.
To capture the commonality in idiosyncratic volatility, the authors propose a novel multivariate generalized autoregressive conditional heteroscedasticity (GARCH) model called dynamic factor correlation (DFC).
The aim of this paper is to use a model-free, nonparametric approach based on the method of maximum entropy in the mean to solve the capital risk allocation problem.
This paper revisits the procyclicality issue in risk-based margin models and provides additional insight on procyclicality mitigation techniques.