It is not too late to ask why the NPA recovery experience of the first two is so much worse than that of the third. Floating an ARC will enable DFIs to get rid of their hard-core NPAs from their balance sheets. The hot potato will be passed on to the ARC in which bad debts common to the three DFIs will be pooled.

Ostensibly, this will help share data and facilitate recoveries. But there is no guarantee that the proposed ARC will be able to enforce recoveries or even discipline defaulters. Faced with pressure, defaulters can declare losses and run to the BIFR for cover. Unless this loophole is plugged, the environment for recoveries will remain vitiated. The government must legislate to give ARC coercive powers to ensure recoveries. Such powers must also be conferred on debt-recovery tribunals. So long as the loophole is not plugged and the required legislation is not in place, ARCs, whether floated by DFIs or the nationalised banks, will be perfunctory creations. Besides, at what price discounts will the hard-core NPAs be transferred to the ARCs? Obviously, the discounts, or losses, will have to be taken on the books of the original creditor. Are IFCI and IDBI ready to take the losses? A related issue is who will foot ARC's losses should recoveries fall short of the amount at which NPAs have been transferred?.

There are miles to go for ARCs to become effective. For now, the DFIs must pull their chestnuts out of the fire. ICICI has shown what can be done: a fifth of ICICI's profit before tax has come from bad loan recoveries.