Banking blues

Rational Expectations

I don't have the number of cases where the vigilance commission has actually detected fraud in banks, though I suspect it can't be too great considering every time there's a major scam you hear of some collusion with a bank manager, collusion that has completely escaped the attention of the vigilance authorities.

Given this mixed record, as well as for other reasons I'll come to in a moment, it may be time to take a hard look at the entire business of getting vigilance officials to be examining each bank loan/proposal, with the implicit assumption that if a loan has gone bad, it is due to collusion with the bank manager.

Sure, there must be collusion in several cases, but poor government policy is also responsible for a large part of bank loans turning bad (loans used to be sanctioned to units which were healthy when the market was protected, but weren't viable when imports opened up). And, in cases of siphoning off of funds, which is why a lot of loans have become NPAs, in any case, it is not the bank's job to be looking at this, it is that of the Department of Company Affairs.

But, to get to the point of this piece, three MIT economists (Abhijit Banerjee, Shawn Cole, and Esther Duflo) have analysed the loan portfolio of a well-run public sector bank, and found that whenever there's any type of vigilance activity (clearly that's long before the actual case gets registered), there's a 3-5 per cent fall in lending by that branch, and this lasts for around two years. That's clearly a lot of lending not taking place, with the attendant lack of credit for producers.

If this isn't frightening enough, consider another effect on overall lending. For one, with bank managers getting risk averse, they generally prefer to invest their money in government securities "" as against the required norm of 25 per cent, banks invested 42 per cent of their deposits in government securities in 2003, up from 27 per cent in 1998-99.

Banks also try to lend to bigger firms who very often don't want their money, even though they would be better off lending to smaller firms. Figuring this out is actually quite easy, using elementary probability theory kind of analysis "" easy, I must hasten to add, if you're a maths-whiz like Pronab Sen who's the advisor to the Perspective Planning Division of the Planning Commission, is the man who really writes the plans, and who did the maths for me.

Since any bank will try to invest in the activity that gives the highest risk-adjusted return, the default-adjusted return on a loan to a blue-chip firm has to be equal to the return on a zero-risk government bond. Given the interest on government bonds today, this means the risk of default of a blue-chip firm is around 15 per cent. That's Sen's calculation, by the way, not mine.

Now, extend the argument to a situation in which a bank lends to several small units in the same industry. Again, the risk-adjusted return from lending to these units has to be equal to the risk-adjusted return on the loan to the blue-chip firm.

Get the maths done by someone who knows how to do it, and you'll see that even if the risk of these smaller firms not repaying their loan is a whopping 60 per cent, the banks will get a higher risk-adjusted return by lending to them instead of to the bigger corporate.

Yet, as we all know, banks don't want to lend to smaller firms, given a chance. The reason is simple: even if one of those loans goes bad, the vigilance official steps in. So, the perceived risk of lending to a smaller firm gets increased even further.

In fact, it's interesting, but the only way to keep the vigilance official off the bank manager's back, is through the concept of 'priority sector' loans "" since lending to small units is mandated under 'priority sector' lending, bank officials are relatively free from vigilance cases since they can always take protection under this umbrella. While most pro-reformists argue against priority sector lending, Banerjee, Cole and Duflo, in fact, come out in favour of priority sector lending, partly for this reason.

But, while lending to groups of smaller firms, like say in the textiles trade, may be something a bank's board may look at, why should an individual loan officer look at smaller firms that have a higher default-risk? Clearly, the interest rate he can charge has to be higher "" Sen calculates the interest rate has to be around 15 per cent, as compared to 7 for the blue-chip, for the loan officer to be interested. But socialist India does not allow this, for we don't want the small- scale unit to be fleeced, right?

Well, guess what? The Grameen Bank of Bangladesh, the bank that everyone tom toms as the saviour of the poor, charges a 16 per cent interest today "" it began with around 21 per cent. And the reason for this was simple: the Bank argued that a firm earned around 50 per cent returns from extra capital, and so didn't mind parting with a third or so of it.

Because if it didn't, it lost that return also. Banerjee, Cole and Duflo, interestingly, find that the marginal returns to be made from an extra unit of loan is 100 per cent. Clearly then small units won't mind parting with higher interest rates, provided they get the money.

Catching crooked bankers who extend loans for favours is certainly important, but the country's rulers need to take a call on whether choking off credit to large parts of industry is too high a price to pay for this.

suniljain@business-standard.com