The Quantum Mechanics of Multiple Borrowing

In the strange world of quantum mechanics, a thing is everywhere until you look at it. I don't mean that it could be anywhere. It is everywhere. An electron in your fingernail possesses a probability distribution that represents how likely it is to be in any given place once observed. Nearly all the probability, but not quite all, is concentrated in a tiny volume around some atomic nucleus. Observing an electron with laboratory tools localizes it. Its probability distribution collapses to a singularity. In the universe as we understand it, observation does not merely remove uncertainty. It changes that which is observed.So it is with lending. Lenders carry outstanding loans on their books as assets. They discount the loans according to certain norms and rules to reflect the likelihood that not all will be repaid. But no ones really knows the value of an IOU until it comes time to collect. Things are particularly uncertain when several lenders are making loans to the same people at the same time (multiple borrowing). The borrowers may being using one creditor's loans to pay off another's, unable to ever climb out of debt. Yet each lender carries its loans at face value. Things are murkier still when lenders do not (or cannot) share information about borrowers, such as through a credit bureau. Then no lender knows how much each borrower owes.When someone falls behind on a loan, standard operating procedure in a bank's accounting department is to provision against the loan, adding a new liability to the balance sheet to reflect the probability of incomplete repayment. The loan stays on the books at face value, but is partially or fully offset by the new liability in calculations of the bank's worth. Then if the loan remains in default long enough, it is written off. Its value is permanently marked down, maybe to zero, and the provisioning against it is also erased. The lender gives up on the loan.A footnote in the Grameen Bank's monthly statements sets forth such rules:

On the last day of each month, 100 per cent provision is made against all overdue loans. Entire outstanding amount of overdue loans are written off one year after they become overdue.

Pondering comments on my post about rising delinquency at the Grameen Bank I realized that I should be able to calculate from public data how much debt the Bank has written off. After all,Total loans at end of month=Total loans at start of month+New disbursements--New repayments--Write-offs,so Write-offs=Total loans at start of month+New disbursements--New repayments--Total loans at end of month.The Bank reports everything to the right of the "=". So I applied the formula to all the data back to June 2002.How much debt did the Bank write off? If I did this right, the answer is: zero. It seems that Grameen's accountants hold out hope for every loan made since June 2002. But you'd think at least one loan had gone bad permanently by now. And this practice seems to contradict the footnote quoted above. Unless I am mistaken, the Grameen Bank is not following its own accounting rules.In general, a loan can leave the "overdue" category in two ways: the borrower catches up on payments, or the lender writes off the loan. If Grameen has blocked one of these exits, then the only way its bad loans have ever stopped being "overdue" is through refinancing: new loans are extended so borrowers can repay overdues. The new loans might be bigger to help with interest payments. That makes things look good. Refinancing can be what Rich Rosenberg calls smoke and mirrors. The temptation to refinance to cover problems is I'm sure as old as credit (or maybe a couple of months younger).Since Grameen's overdues have fallen at times, it appears that some bad loans are being refinanced. This would mean that the Grameen Bank's delinquency problems have been greater than reported. The red and green lines in the second graph of my earlier post are too low.Inevitably, we return to the speculation: is there a bubble in Bangladeshi microcredit? If so, what would prick it? One thing that could is a loss of fuel. But at the Grameen Bank at least, the fuel supply may be growing, in the form of savings deposits attracted by interest rates as high as 10--12%/year (see Graham A.N. Wright's comment). Last year Bangladesh's central bank cut interest rates in response to recession. Unless Grameen also reduces the rates it pays on deposits, money may flock its way. Then, seeking profits high enough to pay 10--12%, Grameen may pour the new money into its lending operation.But you can never be sure there even is a bubble unless it pops. Grameen and the other creditors won't know how much their loans are worth until they try to collect on them. And just as with the position of an electron, the true value of a loan is not a Platonic, fixed quantity awaiting discovery. It is fluid. It is localized by the actions of those who would observe it. Imagine that all 650+ Bangladeshi microcreditors other than Grameen forgive their loans tomorrow. Grameen might then easily recover all of its credits. Or Grameen might forgive first, raising the value of others' IOU's. More realistically, one can imagine a game of chicken whose winner is determined by luck, astute timing, and ruthlessness.Or maybe all Bangladesh's microcreditors could recover 100% today. I don't know. As quantum physics pioneer Niels Bohr said, Prediction is very difficult, especially about the future.