Oil-price uncertainty and the U.K. unemployment rate: A forecasting experiment with random forests using 150 years of data

We analyze the predictive role of oil-price uncertainty for changes in UK unemployment rate using more than a century monthly data covering period from 1859 (when drilling first oil well started at Titusville, Pennsylvania, United States) to 2020. To this end, we use machine-learning technique known as random forests. Random forests render it possible model potentially nonlinear link between and subsequent an entirely data-driven way, where is control impact several other macroeconomic variables financial uncertainties. estimate on rolling-estimation windows find evidence that predicts out-of-sample rate, especially longer (six twelve months) forecast horizons. Moreover, relative importance has undergone substantial swings during history modern petroleum industry . Relative was high 1970s 1980s, higher price itself most sample period. also value when Lasso estimator, have superior forecasting performance forecasts. • Analyze rate. Use Oil-price out-of- predictor changed substantially over time.