Several important aspects of the 1994­95 Mexican economic crisis have been extensively discussed elsewhere.¹⁸ Accordingly, for present purposes, it is appropriate to focus on eight points that relate to broader concerns about the role of the IMF.

First, inadequate economic information can add to the uncertainty facing private asset holders and may also induce them to act on the basis of something other than a correct understanding of economic fundamentals (as discussed above under "Market Imperfections"). In the case of Mexico, many private asset holders appear to have misjudged the need for a correction in Mexico's balance of payments and based their investment decisions on the expectation that there would be no devaluation. Absence of, or delay in reporting, key economic data may have contributed to this situation. The rude awakening that came with the devaluation and subsequent sharp depreciation of the Mexican peso clearly caused many asset holders to re-evaluate both their portfolio positions and the credibility that they attached to the policies of the Mexican authorities. A sharp reversal of capital flows and a crisis of confidence ensued. As for the future, there is obviously no certain safeguard against sudden shifts in investor expectations or the economic effects of such shifts. Nevertheless, timely reporting of economic data and cogent analysis of economic prospects and policies can play a useful role. The IMF has responsibilities in both areas: it can encourage timely reporting of accurate and extensive economic data; and it can analyze and make known its general concerns about existing or emerging economic problems.¹⁹

Second, open financial markets and exposure to private capital flows put a premium on disciplined, continuous implementation of sound, consistent policies. The economic crisis that has beset Mexico since late December 1994 is, to a significant extent, the consequence of problems that accumulated during and before 1994. If timely policy measures had been taken to deal with the appreciation of the Mexican peso and the low national saving rate, Mexico might not have escaped all of its present economic difficulties, but the magnitude of these difficulties could have been significantly diminished. For the IMF, there are important lessons for both surveillance and financing. It is an essential task of IMF surveillance to recognize emerging balance of payments difficulties and to provide timely and relevant policy advice on how to deal with these difficulties, thereby limiting the dangers of major economic and financial disruptions. It is the responsibility of national authorities also to recognize emerging problems, to pay careful attention to relevant analysis and advice, and to act constructively to contain risks of major crises. The recent experience with the Mexican crisis suggests the potential and need for improvement on both sides of the surveillance process. However, it is also clear that IMF financing is necessary in some circumstances: it would be fanciful to suggest that better surveillance alone can satisfactorily resolve all problems, given the difficulties in reaching a domestic consensus on appropriate policies and the occurrence of unforeseen events.

Third, in the case of Mexico, weaknesses in the banking and financial sector appeared to have played a significant role both in developments leading up to the devaluation and in the severity of the crisis that followed the devaluation. For the IMF, the implication is that surveillance should focus particular attention on the soundness of the financial sector, both under normal circumstances and under a scenario in which the economy might be subjected to strong adjustment pressures. Early diagnosis and correction of weaknesses in the financial sector might do much to lessen the severity of losses associated with a country's necessary efforts to correct maladjustments in its balance of payments.

Fourth, a special feature of the Mexican case that contributed importantly to the severity of the crisis was the large volume of short-term, U.S. dollar-denominated debt issued by the Mexican Government--the tesobonos. For the IMF, which has a responsibility to assist countries experiencing balance of payments difficulties, the lesson is that surveillance should focus on policies, including debt management, that have potentially important implications for the use of IMF resources.

Fifth, a more favorable aspect of the situation in Mexico in late 1994, compared with that prevailing at the start of the debt crisis in the summer of 1982, was that a significant part of the capital flowing into Mexico during the 1990s had taken the form of direct or portfolio investment in the private sector.²⁰ When the crisis came, the market prices of these private assets absorbed the primary impact of the shift in investor sentiment. In contrast, the debt crisis of the 1980s had involved a drawn-out process of debt reschedulings and eventual write-downs of syndicated bank loans to the Mexican Government or loans guaranteed by the Mexican Government, during which Mexico had effectively been cut off from world capital markets. To preserve this advance in the structure of capital flows, it will be important to maintain the principle that the risks associated with private investments in private sector assets are not the responsibility of governments. Otherwise, if governments take on the responsibility for servicing the debts of private entities, the economic disruption and financing requirements associated with balance of payments adjustment problems could escalate significantly. The IMF can play a useful role in forestalling such adverse developments through surveillance that discourages the extensive provision of explicit or implicit government guarantees for private investments. Moreover, consideration could be given to sponsoring the adoption of a code of conduct, under which governments would agree not to press other governments to provide ex post guarantees benefiting foreign investors.²¹

Sixth, with respect to the effects of the crisis beyond Mexico, it is clear that financing flows and asset prices in other countries responded quickly, via contagion effects, to the onset of the crisis in Mexico. During the weeks immediately following the devaluation and floating of the Mexican peso, equity and bond prices across Latin America came under significant downward pressure, despite the different circumstances facing these countries.²² It seems likely that the relatively prompt availability of large-scale financing to support an appropriate adjustment program played an important role both in containing the crisis within Mexico and in limiting its broader contagion effects. For the future, the lesson is clear: in the modern international environment of highly mobile capital and sensitive financial markets, a considerable premium will be placed on prompt and forceful action to contain an economic crisis and its potential contagion effects, whether or not those contagion effects are rational from the perspective of individual investors. For the IMF and its members, this means both that countries should be prepared to act promptly and that adequate mechanisms of financial support should be provided for strong adjustment programs.

Seventh, while it is not possible to establish with a high degree of confidence the consequences of a hypothetical failure to establish a strong adjustment program for Mexico, it may reasonably be concluded that the results for Mexico (and other countries) would have been substantially worse than the painful adjustment process that is presently under way. As a practical matter, without the assurance of substantial external support (almost $40 billion of medium-term financing from the IMF and the U.S. Government), the Mexican Government might well have been forced to reschedule unilaterally its external and internal debt, most notably the tesobonos. Such an outcome would almost surely have led to a deeper and longer recession in Mexico, stronger and more persistent contagion effects for other emerging-market countries, and impairment of Mexico's medium-term economic growth prospects. By any reasonable standard of judgment, the benefits from avoiding these losses must be many times the economic cost of the financial support provided to Mexico.

Eighth, it has sometimes been suggested that the financial support provided to Mexico was a "bailout" that raises serious concerns about the "moral hazard" of inducing governments to pursue unduly risky policies. Such concerns, however, are exaggerated. Moral hazard would arise if the expectation of highly subsidized financial support induced governments to pursue otherwise unsound policies on the basis of the rational calculation that the generosity of potential support made such policies economically advantageous. As discussed above, however, the coinsurance mechanism limits such excessive risk taking, and it is clear that the economic pain of the present crisis in Mexico (even with large-scale financial support) is very substantial. No government would have rationally decided to risk the economic difficulties that currently beset Mexico because of the expectation of the financial support that Mexico has received in the current crisis. Indeed, several countries, taking heed of the Mexican example, have been motivated to correct policy deficiencies.

Another concern is that external support for Mexico undermined market discipline by shielding holders of Mexican assets from the size of the loss. Market discipline is important because, in principle, it helps to guard broadly against the general risks of unwise policies, not only against the narrower problem of "moral hazard" behavior on the part of the borrower. The weakening of market discipline would not appear to be a relevant issue for most assets, which dropped substantially in value in the Mexican crisis, but it is a cause for concern in the case of the tesobonos, whose holders suffered no losses.

In the case of Mexico, the use of reserves and the infusion of external financial support enabled the authorities to meet the rolling panic as investors sought to cash in their maturing tesobonos. Does this imply that market discipline will in the future be seriously undermined? That would be the case if market participants were to draw the conclusion from this episode that foreign-currency-denominated liabilities of sovereign governments are always risk free. However, this would not be the logical conclusion to draw, as holders of such securities have suffered capital losses in the past, for instance, during the debt crisis in the 1980s. It is also clear from the general external analysis of this case that most market participants see it as unique, and they are therefore unlikely to conclude that it is the relevant precedent when they consider future investments.

Looking to the future, one may hope that financing requirements on the scale of those in the Mexican crisis will not arise and that, when difficulties do arise, adjustment programs will be sufficiently strong that defaults on sovereign debts can usually be avoided. However, investors should not be guaranteed against the risk of default. With such guarantees, creditors would not have incentives for caution, and it would probably also be difficult to impose adequate disincentives to imprudence by borrowers. Thus, notwithstanding the fact that IMF-supported programs may in many instances assist in avoiding sovereign defaults (when circumstances so warrant), the general conception of the role of the IMF­as a provider of temporary balance of payments assistance under adequate safeguards­does not and should not include the provision of protection against such outcomes. Moreover, in view of the substantial costs and potential adverse spillover effects from sovereign defaults, and of the limits on financial support to forestall even true liquidity crises, the international community has reason to insist on prudent avoidance of policies that raise such default risks.