Adding Sentiment to Multifactor Equity Strategies
Sentiment in news and social media is increasingly seen as a valuable addition to more traditional investment signals and is gaining popularity in the investment community as a source of alternative data. The data can be used to seek additional return as well as limit the risk of investment portfolios.
In this paper, we explore sentiment in the framework of multifactor investment strategies. We investigate the relationships between sentiment-based and more traditional factor strategies and show that sentiment offers an additional signal that can contain information that is not incorporated in traditional investment factors.
We show that an equity investment strategy based solely on sentiment performs on par with a traditional multifactor model that incorporates several well-known factors such as size, growth, momentum and volatility. If sentiment is added to these factors, the strategy shows further additional gains in terms of return, without increasing risk.
A more detailed investigation of adding sentiment as a risk overlay – essentially avoiding stocks with declining sentiment – demonstrates that strategies’ returns are enhanced if sentiment is used in an aggressive way to screen the stocks. Avoiding entire sectors with declining sentiment further improves performance.
Finally, we conclude that combining the use of sentiment in several ways – such as risk screening for stocks as well as for sectors and adding sentiment as a factor to multifactor strategies – achieves the greatest benefits in terms of portfolio performance. It can lead to over 4% yearly return enhancement, even in periods of bullish markets when beating the benchmark is particularly difficult.