The other day, I had the misfortune of travelling with one of the nastier specimens of taxi-drivers in Mumbai: he was extremely rash, spat perhaps 30/40 times on the street in a ride of less than an hour, and while overtaking a young lady on the street, his right hand hanging outside the window deliberately touched (squeezed?) her. When I admonished him, his answer was "yeh to Bambai hai, yahan subkuchh chalta hai (this is Bombay and anything goes here)". To what extent is the "chalta hai" attitude responsible for the number of cases of huge losses on currency derivatives, to the extremely liberal interpretation of concepts like hedging manifest in many of the cases?
While I have criticised the corporate risk management culture in an earlier article (see World Money, March 24, 2008), the banks also do not seem to come out very well in terms of their adherence to the RBI's regulations and guidelines. One has heard several bankers arguing that these practices have being going on for the last several years; that, during this period, banks have had several inspections conducted by the RBI which did not find the practices objectionable; and that, therefore, they have the blessings of the regulator. At least one major bank has a grouse that when it went to the supervisor with specific queries, it could not get clear answers; that its faithful adherence to regulations was resulting into loss of business to other banks not so squeamish; and that, therefore, it too decided to follow the rest rather than ceding competitive ground. (The doctrine of safety in numbers of course has strong credentials in financial markets. So long as performance is evaluated on comparative basis, the "herd instinct" of market players is completely rational. No wonder Keynes remarked a long time back that it is much better for one's reputation to be wrong in the company of others than be right and stand alone!) Whatever the validity of these arguments, if true, they also point to the need for more strict insistence on regulatory compliance if the "chalta hai" attitude is not to gather further momentum.
Turning to the losses being made in credit derivatives in the global markets, one of the more interesting reports I have come across is the "confessional" prepared by UBS for its shareholders, about the $40 bn it has lost in its trading book. It seems the top management had little idea of the bank's huge exposure to complex credit derivatives. Again, the board had less than adequate knowledge of the risky positions and the weaknesses of the risk management systems: both are points bank executives/boards in India may find it worth pondering over. To be sure, the amount of losses Indian banks may sustain through the currency derivatives transactions are nowhere near what some of the global banks have suffered. (But surely they should not wait for those kinds of losses to learn the lesson.) On the other hand, at least some of their clients' corporate health and life have been threatened by the losses they may sustain.
The global experience also points to the weaknesses of statistical models in calculating the risk in a portfolio with a given degree of confidence. Quite apart from the problem of "fat tails", that is, the much larger incidence of extreme events than forecasted by normal distribution, two other issues are coming increasingly to the fore: first, historical data is not an accurate guide to the future; second, in the measurement of the portfolio risks, correlations between different exposures factored in the Value at Risk numbers are subject to change without too much notice, particularly in less than liquid markets. In short, mathematical models cannot be mechanically relied upon, and need to be used along with qualitative judgments.
As far as the derivatives scene in India is concerned, one of the system weaknesses that has come to the fore is the weak management of the credit risk on derivatives. Price risk for the company becomes credit risk for the bank. For effective monitoring, banks dealing in a product should be able to value it by breaking it into its generic elements. Dealing in products one does not know how to value can be very risky, not only for the corporate clients but for the bank itself. This apart, perhaps the most effective risk control is to monitor the return on equity (in relation to the economic capital) one is earning in a particular business. When that is high, it is often a sign that the activity has risks which are probably not being properly factored in the measurement of the risk and the allocation of economic capital. This applies equally to the higher returns on AAA-rated CDOs