![]() There are multiple ways of comparing a model forecasts (a distribution quantile in the case of the VaR) with subsequent realizations (often a portfolio realized return). Sharing the same opinion as our panelists, in the last part of the conference, we had the chance to share our views on the latest developments in the field of back-testing and model benchmarking. This back-testing should be complemented with further quantitative and qualitative analyses to obtain meaningful conclusions. For instance, it was made clear that back-testing, especially in the context of model change, could not be limited to the counting of overshootings and standard statistical tests (e.g. Despite the diversity of management companies represented, the participants reached a consensus on several topics. One week ago, Deloitte Luxembourg held the investment funds risk management conference on Value-at-Risk.įollowing a turbulent autumn on the financial markets characterized by an increased number of VaR overshootings, the panelists provided key insights on VaR model assessment and validation.
0 Comments
Leave a Reply. |