The role that the price of oil plays in economic analysis in central banks as well as in financial markets has evolved over time. Oil is not seen anymore just as a input to production but also as a barometer of global economic activity as well as a financial asset. A high frequency structural decomposition of the price of oil can therefore inform on the state of the global business cycle as well as on global financial market sentiment. In this paper we develop a method to identify structural sources of oil price fluctuations at the daily frequency and in real time. The identification strategy blends sign, narrative restrictions and instrumental variable techniques. By using data on asset prices, oil production and global economic activity we account for the double nature of oil: a financial asset as well as a physical commodity. The model offers novel insights on the relationship between the price of oil and asset prices. We also illustrate how the model could have been used in real time to interpret oil price movements in periods of high geopolitical tensions between the US and Iran and to read the drop of crude prices due to fears related to the Corona virus.