#5 The analyst wants to test the Fisher effect, the hypothesis suggested by Irving Fisher that nominal interest rates increase by 1 percentage point for every 1 percentage point increase in expected inflation. The Fisher effect assumes the following relation between nominal interest rates, real interest rates, and expected inflation: i=rto where i = the nominal interest rate r = the real interest rate (assumed constant) It = the expected rate of inflation To test the Fisher effect using time-series data, we could specify the following regression model for the nominal interest rate: i, = bot bilt , + Es Specify a 0.05 significance level for the test. The Survey of Professional Forecasters (SPF) has compiled data on the quarterly inflation expectations of professional forecasters. We use those data as our measure of expected inflation. We use three-month Treasury bill returns as our measure of the (risk-free) nominal interest rate. We use quarterly data from the fourth quarter of 1968 to the fourth quarter of 2002 to estimate the Equation. Results from Regressing T-Bill Returns on Predicted Inflation Coefficient Standard Error *-Statistic Intercept 0.0304 0.0040 7.6887 Inflation prediction 0.8774 0.0812 10.8096 Residual standard error 0.0220 Multiple R-squared 0.4640 Observations 137 Durbin-Watson statistic 0.4673 A. Does the data support the Fisher Effect? B. Is there conditional heteroskedasticity at .05 and .01 levels of significance? (R2 in the squared residuals regression is 0.1651.) C. What can we conclude about the error term in the Fisher effect regression, the standard errors and the t- test