Model risk

A risk management strategy is implemented on the assumption of a specific model. However, it is not certain whether the model used as a basis also reliably reflects reality. Model risk therefore always arises when the model’s assumptions do not correspond exactly to reality. – A frequently cited example of model risk is the computationally simple normal distribution for stock returns. However, experience has shown that extremely negative price swings in stocks occur much more frequently than would be the case if the normal distribution (Gaussian distribution: the theoretical frequency distribution for a set of variable data, commonly represented by a bell-shaped curve symmetrical about the mean) were valid. As a result, the risk situation can be severely underestimated. – See Worst Case Hedging. – Monthly Report of the Deutsche Bundesbank, December 2007, p. 65 (model risk in the context of risk calculation at banks).

Attention: The financial encyclopedia is protected by copyright and may only be used for private purposes without express consent!
University Professor Dr. Gerhard Merk, Dipl.rer.pol., Dipl.rer.oec.
Professor Dr. Eckehard Krah, Dipl.rer.pol.
E-mail address: info@ekrah.com
https://de.wikipedia.org/wiki/Gerhard_Ernst_Merk
https://www.jung-stilling-gesellschaft.de/merk/
https://www.gerhardmerk.de/