Structures of dependence and covariation are central to financial risk assessment and management. This issue of The Journal of Risk offers insights into and advances on risk measures in a multivariate context, as well as a way of assessing transmission pathways in systemic risk. It concludes with a paper on growth-maximizing trading strategies with drawdown considerations.
The past decade has witnessed a flurry of activity in capturing joint probability distributions through sequential bivariate copulas such as so-called drawable vines (D-vines). In our issue’s first paper, “Impact of D-vine structure on risk estimation”, Catalina Bolancé, Ramon Alemany and Alemar E. Padilla Barreto conduct a sensitivity analysis that shows how copula choice and vine tree construction can impact estimates of risk measures such as value-at-risk and conditional value-at-risk.
The second paper in the issue, “Estimation window strategies for value-at-risk and expected shortfall forecasting” by Tobias Berens, Gregor N. F. Weiß and Daniel Ziggel, provides an illustration of the impact of both forecasting models and time windows on risk measure estimates. The authors show that combining forecasting models with different estimation windows yields the best results overall.
While the previous two papers are static in terms of their focus being on a single risk measure forecast, our third paper, “Monitoring transmission of systemic risk: application of partial least squares structural equation modeling in financial stress testing” by Necmi K. Avkiran, Christian M. Ringle and Rand Low, addresses risk propagation over time. Specifically, the authors evaluate empirically how systemic risk moves from the unregulated shadow banking sector to the regulated banking sector. They show that the former accounts for more than 75% of the variation in systemic risk associated with the latter. In effect, their study suggests that internal risk management in bank holding companies has a greater role to play in reducing the likelihood of systemic risk events.
In the issue’s fourth and final paper, “Risk averse fractional trading using the current drawdown”, Stanislaus Maier-Paape provides a new perspective on a classical problem: namely, the optimization of the Kelly criterion in devising portfolio strategies. As is well known, the standard resulting growth-maximizing strategies are too risky because of their large drawdowns. Maier-Paape shows how the Kelly criterion can be decomposed into two parts: one capturing the effect of chance and the other that of risk, which is measured by the current drawdown. As a result, he manages to derive allocation strategies that account for this risk and are less risky than the classically suggested approach.
Warrington College of Business, University of Florida
In this paper, a sensitivity analysis using pair–copula decomposition of multivariate dependency models is performed on estimates of value-at-risk (VaR) and conditional value-at-risk (CVaR).
This paper analyzes the impact of different estimation window strategies, including structural breaks and forecast combinations, on forecasting common risk measures such as VaR and ES.
Monitoring transmission of systemic risk: application of partial least squares structural equation modeling in financial stress testing
This paper illustrates how the transmission of systemic risk from shadow banking to the regulated banking sector can be modeled using partial least squares structural equation modeling in an effort to help regulators better monitor and manage contagion.
In this paper, the fractional trading ansatz of money management, also called growth optimal trading, is reconsidered. Special attention is paid to the chance and risk parts of the goal function for the related optimization problem.