In this paper, I propose a novel way to model sentiments in asset prices. Under this new representation, sentiments, or animal spirits, are sparked by exogenous shocks to beliefs, but feed on the uncertainty generated by imperfect information. Sentiments cause expectations to deviate from optimal, information-based estimates of fundamental values and their magnitude depends on the amount of uncertainty in such estimates. The higher the uncertainty, the larger the scope for psychological attitudes to affect expectations. Armed with this framework, I investigate the role of uncertainty on the transmission channel from sentiment shocks to prices in a market with imperfect information and Bayesian agents. The main result that emerges is that the source of noise generating uncertainty, whether fundamental or informational shocks, has important consequences for the effect of sentiments. Specifically, while more informational noise always amplifies the impact of psychological shocks on prices, more fundamental noise can actually reduce such impact, depending on the elasticity of sentiments to uncertainty. This result implies that, for example, noise traders in stock markets can actually reduce the relevance of animal spirits for asset prices.
|Number of pages||18|
|Publication status||Published - 27 Oct 2020|
|Name||Munich Personal RePEc Archive (MPRA)|
- Bayesian learning
- financial markets