FINANCIAL MARKETS’ ASSESSMENT: Term structure-based EMU 2

financial_market
The alternate EMU probability calculators diverge primarily in their method of estimating the no-EMU scenario for future interest differentials, and in the selection of the future time interval.

Figure 3 shows the forward spreads for 7 currencies and for the ECU currency basket. Instantaneous forward rates corresponding to fixed 1999, 2002, and 2005 future maturities were computed from 3-, 6-, and 12-month Eurocurrency rates and 1-10 year swap rates, using Svensson’s extension of the Nelson-Siegel (1987) methodology as described in Soderlind and Svensson (1997).8 Forward spreads over 1991-98 were predicting relatively narrow post-1998 interest rate spreads relative to Germany for Belgium, Denmark, the ECU, France, and the Netherlands, but substantially larger spreads for Italy, Sweden, and the U.K. We are an online lender offering quick instant cash loans to everyone interested. Once you have requested a loan from us at Reading here, you can leave the rest to us, and we will be reminding you about the payments and helping you consolidate any debt you may have, for as long as you need and to make sure you pay a fair price.

w6874-8
The forward spreads for most of the countries participating in the currency union (Belgium, France, and the Netherlands) had converged to zero by or before the fixing of official bilateral rates against the DM at the May 1998 summit. Interestingly, however, both the Italian and the ECU 1999 forward spreads were still slightly positive as of the summer of 1998. The latter can be attributed to exchange rate uncertainty regarding the rate of conversion of the ECU into euros, given the presence in the currency basket of non-EMU currencies such as the pound. The positive Italian spreads suggests that some uncertainty also remains as to whether the official lira/DM bilateral rate will in fact be achieved — an uncertainty reflected in an unsuccessful speculative attack on the lira in August.

Of the three countries (Denmark, Sweden, and the U.K.) that chose not to participate in the EMU in 1999, two (Denmark and Sweden) nevertheless exhibited substantial convergence in post-1998 forward rates relative to Germany over 1997 and 1998. The convergence was even more pronounced for Swedish 2002 and 2005 forward spreads, suggesting the possibility of delayed entry.
British 1999 forward rates by contrast diverged increasingly over 1995-98 from German rates, but a sharply inverted forward spread curve over 1997-98 also suggests the possibility of delayed entry.

An immediate problem for EMU probability calculators is that although Germany is assumed to be the low-interest rate country, as has historically been the case within the European Union over 1977-98, forward spreads relative to Germany have occasionally been negative by 0-40 basis points for Denmark, France, the Netherlands, the U.K., and the ECU. This implies that despite history, market participants are assigning some probability to a post-1998 non-EMU scenario in which Germany is no longer the low-interest country.

EMU probability calculators influenced by historical norms that calculate positive interest spreads conditional on EMU not occurring consequently estimate negative fNEMU values over these periods, and fEMU values greater than 1. De Grauwe’s and Morgan’s EMU calculators set such fEMU values to 1, under the interpretation that they indicate a strong probability of EMU. Theoretically, however, a negative forward rate spread can constitute as strong evidence against future interest rate convergence as a positive spread.

Three issues arise in the assessment of whether EMU probability calculators are likely to do a good job. First is the issue of whether European term structures of interest rates have in fact demonstrated any ability to forecast future intra-European interest rate differentials. Second, there are assorted convexity biases, risk adjustments, and currency adjustments that can potentially bias inferred probabilities of monetary union away from the true conditional probabilities. Finally, there is the key issue of identifying the non-EMU regime.