Technical paper/Risk management
Correlation stress testing for value-at-risk
The correlation matrix is of vital importance for value-at-risk (VAR) modelsin the financial industry. Risk managers are often interested in stressing a subsetof market factors within large-scale risk systems containing hundreds ofmarket variables…
Evaluating credit risk models using loss density forecasts
The evaluation of credit portfolio risk models is an important issue for both banks and regulators. It is impeded by the scarcity of credit events, long forecasthorizons, and data limitations. To make efficient use of available information, the…
VAR: history or simulation?
Greg Lambadiaris, Louiza Papadopoulou, George Skiadopoulos and Yiannis Zoulis assess theperformance of historical and Monte Carlo simulation in calculating VAR, using data from theGreek stock and bond market. They find that while historical simulation…
Unexpected recovery risk
For credit portfolio managers, the priority is to properly incorporate recovery rates into existingmodels. Here, Michael Pykhtin improves upon earlier approaches, allowing recovery rates todepend on the idiosyncratic part of a borrower's asset return, in…
Ultimate recoveries
Measuring recovery using the ultimate rate observed at emergence from bankruptcy may be conceptually desirable, but modelling it is difficult.
A false sense of security
Credit portfolio models often assume that recovery rates are independent of default probabilities. Here, Jon Frye presents empirical evidence showing that such assumptions are wrong. Using US historical default data, he shows that not only are recovery…
Ultimate recoveries
Measuring recovery using the ultimate rate observed at emergence from bankruptcy may be conceptually desirable, but modelling it is difficult. Craig Friedman and Sven Sandow tackle the problem by maximising the creditor’s utility function, constructed…
A false sense of security
Credit portfolio models often assume that recovery rates are independent of defaultprobabilities. Here, Jon Frye presents empirical evidence showing that such assumptions arewrong. Using US historical default data, he shows that not only are recovery…
Unexpected recovery risk
For credit portfolio managers, the priority is to properly incorporate recovery rates into existing models. Here, Michael Pykhtin improves upon earlier approaches, allowing recovery rates to depend on the idiosyncratic part of a borrower’s asset return,…
Market-implied ratings
There has been much debate over the respective merits of credit ratings and market-based indicators. Ludovic Breger, Lisa Goldberg and Oren Cheyette present a new approach that tries to incorporate the benefits of both approaches.
Correlation evidence
Like ratings, default correlation is an area of fierce industry debate. But any fundamental, long-terminvestor searching for fair value in credit correlation will want to understand what the historical dataactually says.
Operational risk modelling: aggregating loss distributions using copulas
Capturing the dependence structure between business line/risk event types is an extremely important step for any serious attempt to model operational risk. In this article we show how this can be achieved by using a powerful statistical technique known…
Market-implied ratings
There has been much debate over the respective merits of credit ratings and market-based indicators. Ludovic Breger, Lisa Goldberg and Oren Cheyette present a new approach that tries to incorporate the benefits of both approaches. Starting with agency…
Correlation evidence
Like ratings, default correlation is an area of fierce industry debate. But any fundamental, long-term investor searching for fair value in credit correlation will want to understand what the historical data actually says. Here, Arnaud de Servigny and…
Operational risk: a practitioner's view
The Basel Committee on Banking Supervision ("the Committee") released aconsultative document that included a regulatory capital charge for operationalrisk. Since the release of the document, the complexity of the concept of "operationalrisk" has led to…
The road to partition
Applying the ensemble approach developed in these pages last month, Kevin Thompson and Roland Ordovas calculate risk contributions and show how to measure higher-order default dependence using the method of partitions. The results provide tools allowing…
A better approach to operational risk aggregation
Professor Carol Alexander proposes an aggregation methodology that takes account of dependencies between op risk losses that have some common risk drivers.
Credit ensemble
Kevin Thompson and Roland Ordovas address the question of how individual counterparties contribute to the total credit risk of a portfolio. They provide an analytic method, new to credit modelling, to estimate all joint default statistics conditional…