In this paper, we first extract 8 factors from a monthly data set of 130 macroeconomic and financial variables. Then these extracted factors are used to construct a Factor-Augmented Qualitative… Click to show full abstract
In this paper, we first extract 8 factors from a monthly data set of 130 macroeconomic and financial variables. Then these extracted factors are used to construct a Factor-Augmented Qualitative VAR (FA-Qual VAR) model to forecast industrial production growth, inflation, the Federal funds rate and the term spread based on a pseudo real-time recursive forecasting exercise over an out-of-sample period of 1980:1-2014:12, using an in-sample period of 1960:1-1979:12. Short-, medium- and long-run horizons of one-, six, twelve- and twenty-four-month(s)-ahead are considered. The forecasts from the FA-Qual VAR is compared with that of a standard VAR model (comprising of output, prices, interest rate and the term spread), and that of a Qualitative VAR (Qual VAR) model (which includes the variables in the VAR and the latent business cycle index generated based on the information from the industrial production growth, inflation, the Federal Funds rate and the term spread). In general, we observe that the FA-QualVAR tends to perform significantly better than the VAR and Qual VAR for the one-month-ahead and six-months-ahead forecast horizons for the key US variables under consideration. In other words, adding information from a large data set (through the use of factors) tend to produce forecasting gains at short- to medium-run horizons.
               
Click one of the above tabs to view related content.