Journal of Risk

This issue of The Journal of Risk contains papers dealing with the estimation of risk measures for portfolios with undiversified idiosyncratic risk; the assessment of credit risk models; an extension of the classic futures-based hedging strategy to account for stochastic volatility; trading price impact; and regulatory capital in the face of market risk.

Individual positions in large-sized portfolios tend to be very small. As a result, such portfolios are characterized by undiversified idiosyncratic risk that is identified through granularity adjustment (GA) techniques based on one-factor models. The focus has typically been on credit risk. In the first paper in this issue, “Multifactor granularity adjustments for market and counterparty risks”, Jean-David Fermanian and Clément Florentin extend this paradigm to account for default, recovery and market risks in a multifactor setting. They highlight the challenges that arise in this context and develop explicit closed-form formulas to accurately approximate common risk measures associated with a variety of portfolios.

Our second paper, “Forecasting corporate defaults in the German stock market” by Richard Lennart Mertens, Thorsten Poddig and Christian Fieberg, is an empirical study comparing common credit risk models. Based on their new and manually compiled data set on corporate defaults in the German stock market, their findings show that accounting-based models do not offer discriminatory power and that combining accounting information with forward-looking financial data leads to much-improved default forecasts. Consistent with earlier US-based studies, the authors’ results also show that default forecasts generated by structural models are substantially biased and are dominated by reduced-form models.

Portfolio hedging via stock index futures is a standard in the risk management toolkit. In our issue’s third paper, “Optimal hedge ratios based on Markov-switching dynamic copula models”, Jinzhi Li revisits this strategy, which traditionally assumes static correlation between the futures and the portfolio, by accounting for stochastic correlation through Markov-switching and copula models. This approach is then illustrated on CSI 300 index data, which exhibits skewness and excess kurtosis. The results show markedly improved hedging effectiveness.

In our fourth paper, “Equity market impact modeling: an empirical analysis for the Chinese market”, Shiyu Han, Lan Wu and Yuan Cheng present another empirical study based on Chinese market data. The authors consider nonlinear forms of both temporary and permanent price trading impact, and their analysis strongly suggests that there is price manipulation in the Chinese stock market.

Finally, in “A review of the fundamentals of the Fundamental Review of the Trading Book II: asymmetries, anomalies, and simple remedies”, the fifth paper in this issue, Hany M. Farag addresses the still-evolving new standard of minimum capital requirement for market risk. Farag highlights particular issues for which simple remedies may be contemplated. These apply to standard as well as internal model approaches and include issues associated with delta/vega hedging, backtesting based on reduced- and full-risk factors, and the ambiguity of certain proxies.