On May 2, 1998, the leaders of the countries of the European Union formally selected those countries deemed eligible to participate in a common currency union. This union, commonly referred to as the European Monetary Union (EMU)1, was the third and final stage of a process established under the Maastricht treaty of December, 1991. While various criteria were established in that treaty as pre-conditions for membership, the most severe constraints were with respect to the “convergence” criteria regarding acceptable levels of inflation rates, interest rates, budget deficits and national debt. An additional criterion of exchange rate stability became obsolete after the European currency crises of 1992 and 1993. 11 countries were selected as eligible; an additional 3 (Denmark, Sweden, and the U.K.) chose not to participate, while Greece failed to meet the convergence criteria.
Although the EMU now appears virtually certain to begin on January 1, 1999 with broad participation, that outcome was not expected over most of the 1992-98 period. Academics discussed a two-tiered Europe consisting of a few core countries that would meet the convergence criteria, and a broader group that would not qualify until later. Paralleling this uncertainty, various academic researchers and banks developed “EMU probability calculators” to infer from financial data the probability of various countries qualifying for admission. The best-known is probably the one developed by J.P. Morgan, the results of which were regularly reported in the Financial Times. Four such probability calculations for Italy are illustrated below in Figure 1.
This article primarily reviews the methodologies employed in constructing such calculators. While the initial group of participants has now been determined, many other countries are waiting in the wings and may join later. A retrospective examination of the methodologies may therefore be useful in providing a guide to constructing new calculators for prospective entrants. Section I discusses two alternative direct approaches (Arrow-Debreu contracts, options-based assessments) to creating EMU probability assessments, while Section II examines the more common approaches that use the term structures of European interest rates. Section III concludes with some discussion of what can be inferred from financial data regarding future policies of the new European Central Bank.
Non-standard EMU Probability Calculators
The most direct assessments of the probabilities of specific countries joining a European currency union are the prices of Arrow-Debreu securities that pay off contingent upon the countries being admitted. Two such contracts, on Italy and on Spain, began trading on the Iowa Electronic Markets on March 3, 1998. Each contract paid off $1 conditional upon Italy or Spain being selected by the European Council as eligible to participate in the currency union — which transpired at the summit meeting on May 2. Market prices and volumes for the two contracts are graphed below, in Figure 2.
All countries except Greece that were interested in participating in the currency union met the interest rate, inflation and budget deficit criteria of the Maastricht treaty, while exemptions on the public debt criterion for heavily indebted Italy and Belgium were widely expected on the grounds of substantial improvement. Consequently, the prices on Italian and Spanish admission rapidly gravitated to 95 cents on the dollar once trading began in earnest on the contracts. Further substantial trading volume and price movements were observed prior to the March 27 official reports of the European Monetary Institute and the European Commission regarding countries’ eligibility, at which time prices moved to 99 cents on the dollar.
Options-based EMU assessments
An alternate direct method of assessing EMU probabilities is to exploit the information from currency options’ implicit volatilities regarding future intra-European cross exchange rate volatility. Stochastic volatility option pricing models such as Scott (1987) and Hull and White (1987) indicate that the “implicit” variance that equates the Garman-Kohlhagen (1983) currency option pricing formula to observed option prices should roughly be the expected average variance of exchange rate percentage changes over the lifetime of the option: Electronic Payday Loans Online