engleski

Revisiting herding investment behaviour on Zagreb Stock Exchange: a quantile regression approach

Herding investment behaviour on stock markets has consequences for practitioners, theorists and policy makers. Thus, empirical research on this topic in the last couple of years has grown exponentially. There exist only few papers dealing with Croatian stock market regarding herding behaviour. However, this study employs quantile regression approach of estimating several herding investor behaviour models for the first time in literature. Based upon daily data on 37 most liquid stocks on Zagreb Stock Exchange for the period September 22nd 2014 – May 8th 2018, several model specifications are observed via quantile regression. Since quantile regression approach deals with specific characteristics of financial data (stylized facts) better than OLS method, more robust results can be achieved in order to evaluate if herding behaviour is present on Croatian market. The results indicate that weak evidence on herding behaviour exists in bull markets. However, in extreme (negative) market movements, a much greater dispersion in the market occurs. Moreover, market volatility has a prominent effect on the herding as well. Finally, Agrokor concern crisis in 2017 was controlled for in the model and was found to lower herding behaviour on the market.

herding investor behaviour, quantile regression, Agrokor concern crisis, stock market.

nije evidentirano