Efficient predictability of oil price: The role of number of IPOs and U.S. dollar index

We find the number of initial public offerings (NIPO) a simple but efficient oil price predictor, and provide new evidence that U.S. dollar index (USDX) also has significant out-of-sample forecasting power on oil price. Furthermore, we reveal the highly complementary relation between NIPO and USDX in forecasting oil price and explain it from the points of linear correlation and economic transmission mechanism. With the help of multivariate prediction methods, both the statistical and economic performance of NIPO and USDX can be tremendously boosted. The oil predictabilities of NIPO and USDX partially stem from the forecasting power on oil market sentiment and inflation, respectively, and they can be further enhanced using appropriate nonlinear model. •We find Number of IPOs and U.S. dollar index can effectively predict WTI and Brent spot prices.•Out-of-sample results indicate the Number of IPOs and U.S. dollar index can obtain superior prediction performance.•An investor can realize remarkable economic gains using the Number of IPOs and U.S. dollar index.•The prediction performance of the Number of IPOs and U.S. dollar index is robust to a series of extension test.