In the real world, credit rating agencies were supposed to address part of the information asymmetry. As is well known, they failed. Post-crisis, there have been calls for better rating and better disclosure on bundling. But these are pragmatic solutions. Economists have to address the theory.

Now, it’s not as if economics hasn’t dealt with information asymmetries. Extraordinarily insightful work has been done on this. Nobel Prizes have been awarded for work on information asymmetry. But what theoretical economics is unable to answer for now is whether the conflict between information asymmetry and risk distribution via asset bundling is solvable.

The Math Question. This centres on modelling returns on investment. The mainstream theory elegantly proves that diversified portfolios give the best bang per buck. Financial traders routinely use the tools of this theory. And mostly it does seem to work; everyone is not losing money all the time. But, like in the previous case, there’s a key assumption: returns to investment have to be normally distributed; there should be no extremes.

If a statistician says the average IQ of Indian Express staffers is x with a small standard deviation of y and that this is a normal distribution, what he means is that a big majority of Express staffers will have their IQ in the x-y and x+y range. There will be a very few on either side of this range.

But investment returns in finance may not always be normally distributed; there may not be a meaningful average or a meaningful small deviation from it. In jargon, this is called power law distribution. Such distribution of investment returns can happen, especially at a time of quick change, in technology, in globalised investment opportunities, in financial sophistication, etc.

... contd.