I am working on a project involving stock price estimation, so I want investor sentiments or behavioral bias to be an independent factor, impacting stock price movement. So, are there any means by which I can quantify this variable ?
Hi, in the US market the most popular and professional approach is to use the data of Baker & Wurgler at the bottom of this page: http://people.stern.nyu.edu/jwurgler/.
Alternatively, you can take a look also at this paper: https://ssrn.com/abstract=2778060.
Have a look at this paper as well https://scholar.google.com/citations?view_op=view_citation&hl=en&user=fnFoz8gAAAAJ&cstart=160&sortby=pubdate&citation_for_view=fnFoz8gAAAAJ:kNdYIx-mwKoC and previous work by Misako Takayasu. Hope it can help.
Correlate investor confidence or lack of confidence with actual investment or non-investment, in both particular equities and equities in general. When investors invest in an equity they are confident in it. When they do not invest in an equity they are not confident in it. If there are many investors in many equities, they have confidence in the equity market; If there are few investors in the equity market, those investors do not have confidence in the equity market. Don't guess or speculate! Survey the market and correlate the results.
Good answers above to which I would only add that investor sentiment can vary from sector to sector and over time. In the current UK and US markets, for instance, bond-like asset prices (such as utilities) are depressed by the 'sentiment' of investors who see rising interest rates as a threat to those assets, or otherwise benefit when interest rates are falling. Another sector bias can arise from exchange rates depending on those sectors or companies that are more (or less) exposed to currency risks. So using mathematical approaches based on whole-of-market data may overlook distinctions between sectors.
Hi Aayush, I know exactly what you are looking for. I used to subscribe to a data service when I was a professional trader that was quite famous in the 1980's-90's called "The Consensus Bullish Sentiment Index". It always worked very well for me. For example, I also used it to see how well correlated commodity specific ETF's tracked and could be anticipated by sentiment (attached). Also attached is the last set of data I have.