Oil price assumptions for macroeconomic policy

Abstract

We evaluate the economic usefulness of oil price forecasts by means of conditional forecasting of five US macroeconomic indicators. First, we forecast oil prices using a mixed sampling frequency framework, where oil prices are driven by information available at high-frequency; and subsequently we proceed with our macroeconomic conditional forecasts. Overall, there is diminishing importance of oil price forecasts for inflation projections, whereas the reverse holds true for inflation expectations, industrial production and producers price index. An array of arguments is presented as to why this might be the case. Our findings remain robust to alternative forecasting frameworks and model specifications.