Study Group Report 2009: the Credit Crunch - a discussion
This is the final report on the Credit Crunch discussion which took place during the Study Group in the evening of 31 March 2009. Click on the link at the bottom to download the full report as a pdf document.

Report coordinator
David Allwright (Knowledge Transfer Network for Industrial Mathematics)

Summary
There was a wide-ranging discussion of various different aspects of mathematical modelling within the financial sector with particular emphasis on those aspects that relate strongly the current “credit crunch”. These discussions were prompted by the comments of the Guest Speaker Pat Hagen.

The main points were the following:

• General remarks
  There is a tendency for the coordinated mathematically based advice given concerning possible outcomes to be neglected because of the herd instinct of banks. It is envisaged that the current “crunch” will need to last more than just a year if mathematical and intellectual honesty is to return to the markets with the status it had in the mid 1990s. The current situation indicates that there may be a very bright future for behavioural finance.
• Mathematical tasks
  • The model suggested in the talk used the leverage λ and the cost of funds c, to predict a hysteresis loop. This should be examined more carefully. An ideal development of this model should have few parameters and try to identify the critical dimensionless factors. Models of baskets should be derived that are less dependent on fitting to observed local behaviour and which account for the larger dynamics of the system. Hence feedback of the strategies on the market must be accounted for. The central difficulty appears to be modelling of the default rate h(t) and its distribution.
  • The insurance paradox, where each insurance contract with the very large firms widens the insurance spread and hence requires additional insurance, should be modelled. Here feedback effects resulting from the size of the firms involved is critical. The different roles of segments of the market, such as hedgers and fundamental investors needs to be examined.
  • Can the value of the recent introduction of non-recourse loans be assessed? Banks now turn over many of these mortgages but people can default without going into bankruptcy. What is the value of the change to nonrecourse mortgages and does it still make sense for banks to value these assets on the assumption of a 2% default rate?
• Specific examples of mathematical models:
  • Stratified AIG baskets which lead to initial jumps.
  • In practice only a bank can really give a good estimate of the value of its assets and hence external regulation is very difficult to achieve. This lack of transparency is a major mechanism in triggering the hysteresis effect.
  • There is a large element of the market that solve inverse problems in an essentially ad hoc manner to fit models to data. This needs very careful consideration especially to ensure the inverse modelling is reasonable.

Click on the link below to view the full report.

Download 'Evening discussion - Credit crunch.pdf' (55 Kb).