Central Bank Modelling and Variables Doing Random Walks

Abstract

In developed countries, independent central banks control monetary policy. The goals and instruments of monetary policy are taught in every economics course; however, at times theory fails to work perfectly in practice. In our paper, we review the system of inflation targeting applied by the Hungarian central bank in the past ten years, as well as the modelling framework supporting the making of monetary policy decisions. Many input variables of inflation forecast models are traded commodities. This is why we asked ourselves whether central bank analysts could have given more precise projections on these variables than market players. We used appropriate statistical (stationarity) tests for the time series of a number of variables, running approximately 50,000 tests. Based on the tests, we concluded that input variables dominantly follow a random walk process on the time horizon of inflation targeting, thereby their future values cannot be predicted. Our final conclusion is that based on erroneous forecasts, the central bank could even have made incorrect monetary policy decisions, thereby driving the Hungarian economy in an undesired direction.