Risk Management

Risk is nowadays a major issue for Banks and Financial Institutions. International Institutions, like the Basel Committee, are developing standards and guidelines to safely afford the various heads of this monster, namely financial (or market), credit and operational risk.

Risk is, by its nature, multidimensional. Market risk can be analysed, for instance, by Type (equity, interest rate, FX and commodity), Country, Maturity or Duration, Instrument type (options, forwards, futures, cash), Counterparty etc.

Other “dimensions” can be found for Credit Risk or Operational Risk. What makes Risk peculiar, and difficult to handle, is its non-additivity, due to the correlation of different risk components (the so called “diversification” effect).

But the experties of Daisylabs go well beyond the simple sum.

The last frontier of risk is, now, operational risk, whereby the mathematical models used to manage market risk, and the statistical models used to face up credit risk, are of little help. Bayesian techniques, which combine “hard” data that are thought to be more objective, with “soft” data, that can be purely subjective, are the tool of choice.

Daisylabs people started using Bayesian methods twenty years ago to help medical diagnosis, which presents similar problems.