Journal of Operational Risk

Risk.net

Modeling very large losses. II

Henryk Gzyl

  • An interesting model for estimating the probability of a very large loss is proposed.
  • The probability of a very large loss can also be estimated as that of a rare event.
  • Outliers in the standard losses may overlap with the very large losses.

This paper is a follow-up to the author’s 2018 paper in The Journal of Operational Risk (Volume 13, Issue 2, pages 83–91), the underlying idea of which was that very large operational risk losses should be aggregated to standard losses as a model extension. The latter are the losses described by risk type and business line in the Bank for International Settlements (BIS) classification, observed and duly recorded. A very large loss, when observed, might fit into one of the cells of the BIS classification, but we would not consider that it is caused by the same risk factors as the other losses in that cell. A pending detail in that paper is how to estimate the frequency of very large losses. Here we propose an approach to estimate very large losses similar to that used by Fermi and Drake to estimate the existence of extraterrestrial life. It consists of supposing the event of interest is the result of a concatenation of independent factors and estimating the probability of each factor. The problem is that the events in the causal chain might be events that have never been observed, which ties our subject to that of the estimation of probabilities of rare events. The other issue we address is how to take into account the existence of outliers in the standard losses. These outliers might be comparable with very large losses and have frequencies of occurrence similar to those of rare events. This must be taken into account when computing the value-at-risk of the full loss distribution.

Sorry, our subscription options are not loading right now

Please try again later. Get in touch with our customer services team if this issue persists.

New to Risk.net? View our subscription options

If you already have an account, please sign in here.

You need to sign in to use this feature. If you don’t have a Risk.net account, please register for a trial.

Sign in
You are currently on corporate access.

To use this feature you will need an individual account. If you have one already please sign in.

Sign in.

Alternatively you can request an individual account here: