I find some results in my research that in CSAD the herd behavior was not found and in CSSD was found. Any suggestions of similar results in other studies?
It is rather difficult to detect herding bevaiour with CSSD because of its dummy nature than the CSAD. However, your finding is rather the reverse . May consider Quantile Regression or Simple Bayesian Regression with normal or t distributed innovations.
It is rather difficult to detect herding bevaiour with CSSD because of its dummy nature than the CSAD. However, your finding is rather the reverse . May consider Quantile Regression or Simple Bayesian Regression with normal or t distributed innovations.
I found much more herding when I used CSAD instead of CSSD. Christie & Huang (1995) discuss that CSSD captures only extreme market movements. Many studies discuss that CSAD is much more useful than CSSD in regard to above reason.
It means that there is a presence of herd behavior during extreme market moments(upper 1% and lower 1% market returns observations). Whenever market returns reach at peak(or maximum level) or at lowest (or minimum level), investors start to follow each other and show herding. When investers observe highest return in market, they start to follow each other to take maximum returns on their investment. And when investors observe lowest level of returns in market then they follow each other to avoid maximum losses. It's natural phenomena.
In contrast, there is a absence of herding during up and down market(less extreme market moments). In normal up or down market conditions, they do not follow each other.
There is possibility of such type of results in some markets. There is nothing wrong in these results.