A New Predictive Measure Using Agent-Based Behavioral Finance

Authored by Todd Feldman, Shuming Liu

Date Published: 2018

DOI: 10.1007/s10614-017-9652-1

Sponsors: No sponsors listed

Platforms: No platforms listed

Model Documentation: Other Narrative Mathematical description

Model Code URLs: Model code not found

Abstract

We calibrate Friedman and Abraham's (J Econ Dyn Control 33:922-937, 2009) agent-based model using actual financial data in the US stock market. The evidence shows that the estimated price series from the model is similar to real S\&P price series and the model does match return moments at the second and higher order. In addition, we develop a new measure of investor heterogeneity based on the variability in the estimated position sizes across all mutual fund managers. Our results show that the volatility in individual fund manager positions is able to predict future returns in various time horizons. Moreover, increased variability in position sizes positively affects the contemporaneous change in the CBOE Volatility Index and also leads to greater probability of recession.
Tags
calibration bubbles behavioral finance agent-based finance Stock returns Markets Inflation