Whether European Monetary Union will have real economic effects, and of what sign, has been a matter of considerable debate. Bean (1992), for instance, argued that both the long-term costs and the long-term benefits are likely to be small, implying little divergence between actual and risk-neutral probabilities. On the other hand, the short-term consequences for economic policies and interest rates of failing to achieve membership in a currency union could in principle be substantial.
For instance, a failure by Italy to achieve either the interest rate or inflation criteria of the Maastricht treaty might well have caused yet tighter monetary policy, in order to qualify at a later date. Whether this matters for risk premia does, however, depend upon the degree to which Italian bond markets are segmented from world markets. It appears appropriate to interpret inferred probabilities as “risk-neutral” probabilities and, to a first approximation, as conditional probabilities as well. Getting fast easy payday loans online has never been easier. Visit us at review to get your money without waiting. No matter what you need – to pay off your credit card debt, buy something urgently or achieve any other financial goal: we can guarantee best interest rates and easy application.
A crude calibration of the potential difference between actual and “risk-neutral” probabilities, inspired by Hansen and Jagannathan (1991) and Cochrane and Saa-Requejo (1996), can be obtained by placing bounds on the maximal feasible Sharpe ratio from speculating in an Arrow-Debreu forward contract on EMU occurring. If f is the futures price on EMU and p is the true probability,
Thus, the bias depends upon the maturity of the forward rates used in assessing EMU prospects, and the divergence between the EMU and non-EMU contingencies. If, for instance, 1999 forward rates are used, then bond yields correspond to 1999 maturities and differ only insofar as the event of EMU occurring or not is associated with divergences in pre-1999 interest rate paths. For instance, Italy might be more likely to join the EMU under tighter pre-1999 monetary policies, implying yEMU > yNEMU and a downward bias in inferred EMU probabilities.
If forward rates from later maturities are used, as in Favero et al (1997) and as graphed above in Figure (3), alternate biases arise. Bond yields for longer maturities can be decomposed into the 1999 bond yield and post-1998 average forward rates:
If the difference between EMU occurring and not occurring implies a difference of 4% in the average forward rates, using instantaneous forward rates for the year 2002 would bias inferred probabilities upward by at most H(.04)(3) 9 .03. Thus, using forward rates instead of bond prices does not appear to significantly influence EMU probability computations.