Jun 26

Concentration limits

by David Harper, CFA, FRM, CIPM

FRM |

iStock_000003818325XSmall

The analytical treatment of concentrations in credit portfolio has followed an evolution, from basic to more advanced. The more advanced approaches tend to treat correlations in the context of portfolio theory; e.g., Moody's KMV Portfolio Manager via asset correlations, CreditMetrics via asset correlations and a transition matrix. Basel II's (foundation or advanced) internal ratings based (IRB) approach is somewhere in between simple and complex: it gives a nod to portfolio unexpected losses by assigning each credit a correlation to a single shared "systemic factor."

The use of concentration limits is very traditional (this term was assigned in the 2007 FRM but seems to have dropped as a foundational AIM in 2008). Concentration limits are simple and intuitive, and apparently, are still used by most financial conglomerates. Common sense appears to be making a comeback in the wake of the credit crunch and the elevation of model risk.

A concentration limit can apply to an industry/sector, a country, or to collateral type; e.g., in commercial real estate, a bank may limit its exposure to office buildings. The FRM assigned reading (Studies in credit risk concentration) is concerned with two types:

  • Name concentration: too much exposure to specific names (idiosyncratic)
  • Sector concentration: too much exposure to industry/sector/economic class (systemic)

Classic sector concentration limit

The simple example is given by Saunders (see EditGrid below). If you want to limit losses to 10% and you expect to recover 60 cents on the dollar (i.e., LGD = 40%), then your concentration limit = 25%.

Comments

  1. Be the first to leave a comment!

Leave a Comment