Crises, market shocks, and herding behavior in stock price forecasts

Abstract

This study examines the (anti-) herding behaviors of stock price forecasters, focusing on whether their behaviors are time-varying. It studies stock price forecasts for the Nikkei 225 price index from the ESP Forecast in Japan based on nonparametric methods. Empirical results show that stock price forecasters are likely to anti-herd, and the uncertainty caused by financial crises and market shocks is related to the prevalence of (anti-) herding. This study finds that an increase in forecast uncertainty works in both directions, toward herding and anti-herding. Unprecedented shocks, including the financial crises, European sovereign debt crisis, and newly introduced policy packages by Abenomics, increase incentives to differentiate forecasts from others, possibly due to reputation or superstar effects. However, some market shocks, including the BNP Paribas shock and the China shock, intensified herding or lessened anti-herding.