In the general discussion following the commentaries there was considerable agreement that we know much more about the factors affecting long-run trends in exchange rates than about the short-run behavior. To improve our knowledge of the latter, it was constantly emphasized that we need to distinguish clearly between anticipated and unanticipated changes in policies and economic conditions.
Considerable attention also was given to the question of the degree of diversiflability of exchange risk and nature of risk premiums in the foreign-exchange market. It was pointed out that these questions have important implications for investigating exchange rate behavior and evaluating the efficiency of the foreign exchange market. Although a good bit of interesting analysis has been done in this area in recent years, many of these questions remain unanswered. The floor discussion revealed that the accumulated empirical evidence was interpreted by most discussants as having been quite strong against the adequacy of the purchasing power parity as a guide to equilibrium exchange rates and against claims that the foreign-exchange markets have been persistently dominated by destabilizing speculation. Nevertheless, a number of discussants pointed out that, so far, we have neither conclusive evidence of a high degree of efficiency of the foreign-exchange markets nor objective criteria for identifying conditions under which official intervention might be desirable. In the absence of such evidence and criteria, not surprisingly, a wide range of views was expressed.
In spite of noticeable differences in views expressed by the discussants, a rather widespread agreement emerged that the predominant cause of exchange-rate fluctuations has been the instability in underlying economic and financial conditions, rather than irrational and exaggerated speculative behavior. Indeed, to the extent that there may have been serious deficiencies in the behavior of speculation under flexible rates, this is probably as much the result of too little stabilizing speculation as it is the result of actively destabilizing speculation. Opinions were expressed that the current state of research does seem to indicate strongly that most of the exchange rate variations we observed under