GEI Newsletter Issue No. 1

Reforming the International Monetary System: Lessons from the Mexican Experience - by David Vines

Also in this issue:

Editorial
by David Vines
Report on 'The Future of Global Economic Institutions' A Workshop of the Global Economic Institutions Research Programme of the Economic and Social Research Council, London, 22/23 March 1995
by Ray Barrell

Reforming the International Monetary System: Lessons from the Mexican Experience
by David Vines

1. Emerging Markets, Volatile International Capital and New Capacities

Recent events have suggested the need for increased capacity to deal with the kinds of problems which engulfed Mexico, and may yet submerge Argentina. The problems envisaged seem most likely to appear in ‘emerging market’ economies such as Mexico, Indonesia, Argentina, China, but the issues discussed may be of relevance elsewhere, including Eastern Europe.

The representative ‘emerging market’ is a low-wage developing country in which sound macroeconomic polices have been put in place. These provide some guarantee of stability. Against the background of such policies, the returns to inward investment come to seem high.<$F The most obvious case of such investment opportunities is that of “product-cycle” manufacturing for export: relocation to the low-wage emerging economy of the production, for export back to advanced countries, of standardised manufactured goods.> The prospect of high returns on such activities leads to an inrush of funds – equity, bond, and loan. Suddenly the country involved is exposed to the volatility of international capital flows, and to the possibility that the new funds might be withdrawn.

This essay sketches out the details of the Mexican problem, and discusses the kinds of policies which might have been adopted, and might be adopted in related future circumstances. Many of the ideas were floated at the workshop of the furture of global institutions, reported earlier in this newsletter.

2. The Mexican Crisis

2.1 The Internal Logic of the Collapse of the Peso


The following is a sketch of the Mexican events concentrating on one aspect: the collapse of the fixed exchange rate vis-à-vis the dollar. Other aspects are mentioned subsequently.

The Basic Story

First of all there was a boom. The joining of NAFTA involved a very big reduction in the risk premium on capital: investments which had been affected by policy risk now had a guarantee of policy stability. The result was a surge in investment, causing a large shock to aggregate demand. The commitment to a fixed exchange rate led to an inability to choke off this boom by monetary tightness and exchange rate appreciation; there was not the experience or the institutional mechanism to deal with it by fiscal contraction. The running unchecked of the boom stimulated liquid capital inflow in excess of the investment surge, in search of financial and other assets appreciating in value, and speculating on the possibility of exchange rate appreciation; this further fuelled the boom. The result of the boom’s surge was rising costs and prices, which, most critically, reduced exports and promoted imports, reducing aggregate demand and thus output.

At the same time, the investment boom led to potential expansion in supply capacity. Such an enhanced supply capacity would work against the above mentioned tendency for rising costs and prices, but only gradually. As in real business cycle theory, such capacity expansion takes ‘time to build’.

One can in fact think of the initial reduction in risk premia as setting up a race between the demand boom and time to build: between the cost-raising consequences of the demand boom, which depressed net exports, and the cost-reducing consequences of the supply expansion, which expanded them. The essential claim of the analysis presented here is that the outcome of such a race is uncertain. If the supply expansion is not large enough or quick enough then macroeconomic outcomes and policies which look sustainable turn out not to be. This is essentially Hayek’s version of real business cycle theory. (See Hicks, 1967)

It appears ex post that time to build did badly: too little, too late. As a result, entering into what we might call the second period of our analysis, the economy had an uncompetitive cost structure, an emerging reduction in net exports, and low demand, and so was facing a slump.

To correct this slump at a fixed exchange rate would require a fall back in costs and prices – to mirror their racing up in the boom. But there is a basic asymmetry which makes such falls hard to achieve.

Another mechanism available to correct such price imbalance is currency depreciation. It is true that a government committed to a fixed exchange rate vis-à-vis the US dollar, as part of its policy reform, would want to resist depreciation, in order to avoid losing credibility. An initial cost-benefit calculation of the reform strategy might well have argued in favour of the credibility benefits of a fixed exchange rate, even if such a regime involved a tying-of-the-hands of monetary policy, preventing it being used to stabilize shocks such as the one just described. But like all such regime-choice questions this involves a cost-benefit calculation: beyond a certain point, soldiering on with a fixed exchange rate is too costly an option. This appears to be the best explanation of the underlying forces which were at work in Mexico last year: that the loss of export demand caused by uncompetitiveness caused maintenance of the fixed exchange rate to become too costly.

Notice that this is not a story about balance of payments problems, or about running out of reserves, as in the original Krugman (1979) model of balance of payments crises. ‘Mexico, like Sweden, was forced off of a pegged exchange rate, when the nominal exchange rate became overvalued and therefore inconsistent with internal balance’ (Sachs, 1995, pp17, italics added). We might add that the story is very like that of the demise of the UK’s membership of the ERM in 1992.

Further Details

The difficulty of sustaining the fixed exchange rate appears to have been compounded by a reversal of capital flows which, with hindsight, looks inevitable. With the rise in costs and prices, the uncompetitiveness of the export sector – and the collapse, with the slump, in the values of domestic assets – capital flowed out, further exacerbating the reduction in asset values and in aggregate demand. These problems were made worse by the assassination in March 1994; asset holders, fearing the possibility of abandonment of the policies of reform began to require a risk premium; depressing demand, making the slump worse and so making it more difficult for the government to soldier on. Furthermore, the rise in US interest rates increased the foreign rate on which the risk premium had to be applied, leading to further pressure. It is of course difficult to disentangle the relative strength of these three capital-flow effects on aggregate demand.

In models of collapsing exchange rates, such as that by Obstfeld (1994), wage setters, fearing the possibility of depreciation, begin to also bid up wages which creates a self-fulfillingness about the collapse; further weakening competitiveness, making the slump worse and making it yet more difficult for the government to soldier on. Inspection of the Mexican inflation data suggest that this cannot have been a central feature of developments in 1994.

In other models of collapsing exchange rates, e.g. that by Ozkan and Sutherland (1993), asset portfolio holders, fearing the possibility of depreciation, begin to require a further risk premium, simply because of fears of impending collapse, which creates a further degree of self-fulfillingness about the collapse; further reducing aggregate demand, making the slump worse and making it yet more difficult for the government to soldier on. The account by Sachs, Tornell and Valasco (1995) suggests that this happened in two stages in late 1994:

• when the reserve loss became clear in November
• when rumours of devaluation began to circulate in December.

Summary

The intrinsic features of the above-described process are:


(i) a large boom in aggregate demand causing an increase in output and prices, intrinsically leads to a subsequent reversal in a ‘second phase’, with this latter setting in train a temptation to allow the fixed exchange rate to collapse;

(ii) an expansion in aggregate supply of uncertain speed and size, which if it is not fast or large enough fails to prevent the slump in output in the second stage, and thereby fails to save the regime; and

(iii) a reversal in capital inflow following the reversal in aggregate demand, making collapse in the second phase more likely.


In addition:


(iv) additional unexpected negative shocks – in this case the assassination – can provoke a rise in the risk premium, making collapse more likely; and

(v) a further increase in the risk premium in the final stages of the process can worsen the slump and so make the collapse more likely still.

This account sees evidence from the peso collapse as an issue of policy choice for the government; abandoning a commitment to the fixed exchange rate in the face of (doubly) adverse external circumstances.

2.2 The Delay from March to December

Sachs et al (1995) argue persuasively that the Mexican government resisted a central part of the above process, by resisting the rise in interest rates which it suggests became necessary from after March 1994. That is, the authorities determined to both (i) hold on to the fixed exchange rate (for credibility reasons) and (ii) prevent the interest rate from rising (to ameliorate the slowdown, especially in an election year).

Elementary reasoning suggests that such a delay can only be temporary and that it can only be achieved by running down reserves, or by borrowing. In defence it can be argued that reserves are precisely for buffering in the face of a temporary worsening of circumstances, and that (i) the outcome of the race between time-to-build and the demand boom might have come right in due course, and (ii) the negative effects of the assassination shock might have gone away in due course. Ex post , these hopes appear to have been misplaced.

2.3 The Crisis

What the account in Section 2 explains is the collapse of the peso. What it does not explain is the panic and crisis post December 1993. As Sachs et al (pp12) argue the reason for this can be found in the delay just described. As a result of this delay:

• there was a depletion of gross international reserves in the second and fourth quarters of 1994 of approximately $20bn; and
• there was a conversion of more than $20bn or more of short-term debt into dollar denominated debt in the course of 1994.

As a result, the ratio of the dollar component of M3 to gross foreign reserves rose precipitously from 55% to 170% in October 1994 (ibid.). The government became at risk of default, not because it was insolvent but because it was illiquid. Creditors stopped buying government paper because they feared default; this moratorium made default more likely. Furthermore the risk of default then spread from the government to the banks (because an illiquid government was unable to play the role of lender of last resort) and from there to the corporate sector (who became unable to roll over short-term debt and thus subject to panics).

The bailout route used to resolve the crisis has essentially involved all creditors being repaid (although those in pesos have taken a large capital loss at the present exchange rate). In particular the holders of tesobonos have been able to roll over at face value in dollars as the obligations became due. An alternative would have been to suspend repayments on tesobonos and to convert short-term tesobonos into long-term claims on the Mexican government. The reasons which can be given for proceeding in this way (see Sachs et al) are:

• Mexico has a chance to re-establish its market creditworthiness, which suspension of payments on tesobonos would have quickly destroyed; and
• Mexico’s long-term fundamentals (debt ratios, fiscal position, etc) are manageable, indeed good.

Continued - 3. Reform Lessons

The Newsletter of the GEI programme is published three times annually to inform policy-makers and the academic community of research, meetings, conferences, and Working Papers of the GEI programme.

To receive regular issues of the Newsletter of the Global Economic Institutions programme or to receive copies of back issues send an email containing your address and contact numbers to gei@cepr.org.