Private prison stocks and the 2020 presidential election

Objective Can we gain insight into the outcomes of presidential elections, and their determinants, other than through opinion polling or prediction markets? This matters because of recent misses in polling and the contestation of the 2020 election beyond Election Day. Methods Using alternative generalized autoregressive conditional heteroskedasticity models, we conduct an event study of two U.S. private prison companies, whose valuations have depended on their being awarded federal contracts, during the 2020 campaign and afterward. Results Comparison around Election Day of changes in prison company stock prices based on these models and in the predicted probability of President Trump being reelected based on a popular prediction market allows inference of the effects of the January 6 incident at the U.S. Capitol and the Biden inauguration on the subjective probability that Trump would retain power. Conclusion The probability of Trump retaining power that was reflected in asset markets remained positive up to the Biden inauguration—a real‐time indication of the fragility of American democracy.