Sensitivity Analysis of Taylor Curve Estimation: A Comparison of GARCH and Stochastic Volatility Models

The Taylor curve is a concept in macroeconomic policy analysis that represents the efficiency frontier of monetary policy. Previous studies have used generalized autoregressive conditional heteroskedasticity (GARCH) models to estimate the conditional volatility of inflation and output gap, which are key inputs for estimating the Taylor curve. However, the sensitivity of these results to the choice of the volatility model can be substantial. The purpose of this study is to compare the performance of the GARCH and stochastic volatility models for estimating the conditional volatilities of output gap and inflation, which are key inputs to the time-varying parameter model used to empirically test the Taylor curve. Sensitivity analysis is conducted by estimating the conditional volatilities using both models and comparing the resulting estimates of the Taylor curve. The results show that the choice of the volatility model can have an impact on the estimated efficiency frontier for the evaluation of pre-financial crisis period in the United States.