Seminario CEDE - Jesús Otero

We test for weak- and strong-form efficiency of fixed-event individual forecast revisions of Brent and WTI prices using proprietary microdata from Energy & Metals Consensus Forecasts produced by Consensus Economics. We contribute to the literature on strong-form efficiency testing by assembling a comprehensive list of empirically relevant variables that are assumed to be in the information set of the forecasters. This set includes variables that aim to capture the effects of COVID-19 and the Russia-Ukraine conflict. We overcome the variable selection problem with fat data using a novel form of multiple testing and two popular methods from the penalised regression literature. Our results show that analysts' current revisions are independent of past revisions in most cases, supporting that Brent and WTI price forecast revisions are weakly efficient. However, there is little support for strong-form efficiency, conditional on passing weak efficiency. Indeed, we find that forecasters use past values of several variables to revise their forecasts, including measures of the extent of COVID-19 and the Russia-Ukraine conflict.