Recent studies of LDC debt have argued against `buy-backs' by debtor countries, since the payments involved correspond not only to the dividends due to their creditors but also to the increase in the nominal property right that any one outside investor in the debt would achieve vis-à-vis those of others. Moreover, for large buy-backs, debtors must also pay a capital gain to creditors by reducing their overall debt.

In Discussion Paper No. 462, Research Fellow Daniel Cohen and Thierry Verdier argue that these results depend critically on the assumption that the buy-back is known to the creditors. If a debtor can hide its transactions behind the veil of a fictitious outside operator, it may be able to buy back all its debt at the ex ante price, since it is only when the country is identified as the buyer that its solvency is improved by the purchase of the debt, and the price of its debt rises accordingly. A country may benefit from a `secret' buy-back, since its overpayment for the `first' dollar may well be outweighed by the underpayment for the `last' dollar if the transaction is sufficiently large.

Cohen and Verdier focus first on the case where rational lenders have perfect information on the debtor country's resources and preferences, so that they may foresee its incentive to transact on the secondary market (and consequently raise the price at which they are ready to sell), even though they do not observe the actual transactions. Intuitively, some such buy-backs by utility-maximizing debtors must occur, since if none occurred and hence none were expected by rational investors then one large buy-back would enable a country to repurchase all its debt at a cheap price. On the other hand, if a large buy-back were fully anticipated by the creditors, it would be fully acknowledged in its price, and therefore not rewarding for the debtor. The equilibrium must therefore lie between these two polar cases.

Cohen and Verdier then consider the case where there is a continuum of richer and poorer debtors and creditors' information on the debtors' cash constraints is imperfect. They show that a positive fraction of the (richer) debtors will certainly undertake the buy-back and that the `secret' buy-backs raise both the welfare of the debtor undertaking them and the profits of the banks. They conclude that secret buy-backs appear to represent a means of alleviating the free-rider problem that prevents the banks from reaching an efficient agreement with the debtors. This result may also help to close the gap between the normative literature (which argues against buy-backs) and the positive literature, which shows that many countries like to undertake buy-backs and usually prefer to do so secretly.

Secret Buy-Backs of LDC Debt
Daniel Cohen and Thierry Verdier

Discussion Paper No. 462, September 1990 (IM)