Between 1982 and 1989, the International Monetary Fund (IMF) acted as a third party in a total of 210 debt restructurings between 41 debtor states and their creditors - official lenders and hundreds of commercial banks. A detailed reading of the Executive Board Minutes and the Staff contributions offers a unique view of how this stable though entirely informal restructuring regime was gradually assembled, operated, and eventually abandoned. On the one hand, the principle that adjustment costs should be shared equitably between the parties shadowed the logic of a bankruptcy procedure, though in practice the later was substituted by a rule of mutual veto. On the other hand, the stark divergence of interests between debtors and creditors could be endogenized only as long as the IMF and its core member states had enough leverage over them. From this derive three defining though problematic features of this regime: (i) national banking regulators were prepared to exert considerable pressure on private lenders; (ii) the Fund voluntarily committed itself to not lending to its own member states unless these banks had agreed to follow suit; and (iii) IMF conditionality was explicitly mobilized in support of the new commitments that these banks had received. Therefore operational stability came together with a remarkable mix of quasi-judicial patterns, open brokering, and arm-twisting.