| Summary: | I add a sentiment index to the classical asset pricing factors to determine stock returns. Previous studies found that stocks that have a bundle of characteristics which make them more difficult to value and to arbitrage are those more sensitive to market sentiment fluctuations. I analyse the period from June 2002 to July 2019 and find that market sentiment still plays a role in affecting stock return, but the salient characteristics which make some stocks more sensitive to sentiment fluctuation are only six instead of ten, as it was in less recent sample periods. I find evidence that the difference in results is a consequence of the advent of new forms of arbitrage strategies, such as HFT and quantitative investment strategies based on web search engine, and more sophisticated models to determine firm value. Moreover, learning needs to be considered in capital markets when we compare past and recent times.
|
|---|