Edinburgh Napier University

  We use a novel database that identifies allegedly Donald Trump’s fake news during his presidency. We find that the number of daily fake news is positively related to contemporaneous US stock market returns. Fake news is typically positively biased in our context, increasing stock returns in the short term. We invalidate alternate explanations of the main relation, such as the notion that newly arrived information drives the relation. The mechanism of the relationship is the source used and the reliability of the fake news. Fake news matters to the extent that participants believe it is true. This positive relation reverses over the following days, indicating some evidence of correction. Overall, we find that a politician’s fake news influences financial markets temporarily