Project II Troy Dewitt Emelia Bragadottir Christopher Wilderman Qun Luo Dane Louvier
Scope of analysis Does monetary policy affect the U.S. Economy? What is going to be the status of U.S. economy upon graduation – will we be able to get a job? Will inflationary pressures force the Federal Reserve to raise the interest rate? Use univariate model to forecast the federal fund rates Use univariate model to forecast the unemployment rate If causality exists between the two variables use a VAR model to forecast the data
Trace Unemployment Rate and Federal Funds Rate - It appears that changes in the unemployment rate lag changes in the federal funds rate
Correlogram for FFRCorrelogram for Unemployment
Both time series have a unit root – Meaning we have to difference the data to eliminate the data’s dependence on time ADF Test Statistic % Critical Value* % Critical Value % Critical Value ADF Test Statistic % Critical Value* % Critical Value % Critical Value FFR - only at 1% level Unemployment rate
Differenced Time Series
Correlogram of differenced FFR Correlogram of differenced unemployment rate
Unit Root Test dFFR dUnemployment rate ADF Test Statistic % Critical Value* % Critical Value % Critical Value ADF Test Statistic % Critical Value* % Critical Value % Critical Value
Univariate ARMA Output
Correlogram of Residuals dFederal Funds RatedUnemployment Rate
Correlogram of Residauls Squared dFederal Funds RatedUnemployment Rate
ARCH LM Test dFederal Funds RatedUnemployment Rate
ARCH GARCH Model dFederal Funds RatedUnemployment Rate
Actual Fitted Residual Graphs dFederal Funds RatedUnemployment Rate
Correlogram of Residuals dFederal Funds RatedUnemployment Rate
Correlogram of Residuals Squared dFederal Funds RatedUnemployment Rate
Forecasts of Univariate Models
Causality Does the unemployment rate Granger cause the federal funds rate? Does the federal funds rate Granger cause the unemployment rate? Is there two-way causality between these two variables?
Cross Correlogram: Differenced Unemployment Rate and Differenced Federal Funds Rate (Evidence of Two-Way Causality)
Granger Causality Test Pairwise Granger Causality Tests Date: 05/31/07 Time: 15:10 Sample: 1954: :12 Lags: 24 Null Hypothesis:ObsF-StatisticProbability DFEDFUNDS does not Granger Cause DUNEMPLOYMENT DUNEMPLOYMENT does not Granger Cause DFEDFUNDS E-06 Both variables are significant, however the unemployment rate is more significant
Summary of Significant Lags of Vector Autoregressive Estimates
Impulse Response VAR Estimates (Unemployment appears to have significant effects on the Federal Funds rate for five months)
Impulse Response VAR Estimates (Federal Funds rate has an insignificant effect on the unemployment rate)
Test of the models ability to accurately forecast Apr 06 – Apr 07
Test of the models ability to accurately forecast Apr 06 – April 07
VAR Forecast May 07 – Dec 07
VAR forecast of May 07 – Dec 07
Forecast Comparison
Questions?