The arguments put forward by Bulow and Rogoff [Brookings Papers on Economic Activity, 2 (1988) 675; Quarterly Journal of Economics, 106 (1991) 1219] against sovereign debt buybacks are re-examined in a willingness-to-pay framework. This paper argues that the Bulow–Rogoff framework treats default by a debtor as an event with no dead-weight loss, and, as such, underestimates the potential gains from a buyback. The willingness-to-pay framework allows dead-weight costs of default to be introduced in a consistent and simple fashion into the buybacks calculus. Two versions of this framework are considered. First, a model in which the default costs induce an all-or-nothing default decision is analyzed. In this case, an ambiguous result is derived in which the variability of the debtor's income determines whether (small) buybacks are beneficial to the debtor, even though expected total transfers to the creditor increase, consistent with Bulow–Rogoff. In a second version, default costs are modeled so as to induce at most a partial default. This model corresponds most closely, in terms of the repayment behavior of the sovereign debtor, to the models used by Bulow and Rogoff. It is shown that small buybacks are always beneficial to the debtor in this case. The second version is extended to include an investment opportunity. Only if the country has sufficiently scarce resources when the investment can be made will a buyback be harmful to the interests of the debtor.