The aim of this paper is to raise a few open questions and to bring to light some mismatches between existing theories and the evidence. (1) It is shown that many standard international debt models unwittingly require some agents to behave irrationally. A method using triadic interactions is developed here to explain the occurrence of lending with sovereign risk and fully rational agents. (2) The market structure underlying the existing models is often left unclear. It is shown that these models, contrary to what is widely believed, are often not competitive in the traditional sense - they require lenders to be locked into more severe competition than the borrowers. The real-life validity of this is questioned and a model is constructed in which lenders act monopolistically while borrowers compete with one another. (3) Though most existing models exhibit excess-demand for credit in equilibrium, there is considerable evidence of 'loan-pushing* having occurred in the international credit market, with Third World countries being coaxed to take more loans than they would on their own. A preliminary attempt is made to model equilibria with loan-pushing.