Measuring the output gap, potential output growth and natural interest rate from a semi-structural dynamic model for Peru

This paper uses a calibrated version of the Quarterly Projection Model (MPT, for its acronym in Spanish), a semi-structural dynamic model used by the Central Reserve Bank of Peru for forecasting and policy scenario analysis, to jointly estimate the output gap, potential output growth and natural interest rate of the Peruvian economy during the inflation targeting regime (between 2002 and 2017). The model is employed as a multivariate filter with a sophisticated economic structure that allows us to infer the dynamics of non-observable variables from the information provided by other variables defined ex-ante as observable. As the results from the Kalman filter are sensible to the variables declared as observable, we use five groups of variables to be defined as such to build probable ranges for our estimates. The results indicate that the estimated output gap is large in amplitude and highly persistent while potential output growth is very smooth. Therefore, most of the variation in Peruvian economic activity during the inflation targeting regime can be attributed to the former. The estimation of the output gap also proves that monetary policy has been extensively responsive to this leading indicator of inflation. Meanwhile, the real natural interest rate is estimated to be considerable stable, averaging 1,6 percent in the sample with only a sharp decline to 1,3 percent during the financial crisis. The main finding of the paper, however, is that there has been a steady deceleration of potential output growth since 2012. A growth-accounting exercise proves that this trend follows mostly a reduction in total factor productivity (TFP) growth during the same time frame (although the drop of capital and labour contributions jointly explain almost a third part of average potential output growth slowdown between 2010-2013 and 2014-2017).