Whilst the world is flooded with a vast amount of new data, the need to structure and make sense of it grows – particularly in the financial industry. The fusion of public data with alternative data such as sentiment or attention can deliver new insights and the ability to get ahead of the markets. In 2015, Goldman Sachs Asset Management outlined: “We believe a world of possibility would open up for businesses which succeed in quantifying the qualitative.” (Source).
We at Sentifi believe that the data we provide offers many benefits across the investment decision-making process, but most importantly, it helps investors to generate alpha. To verify this belief we conducted a backtest based on a portfolio of S&P 500 stocks. We checked if, and how, investment decisions driven by sentiment can generate risk-adjusted returns. The portfolio consisted of the ten stocks that registered the highest sentiment change over an observation period of two weeks. Every two weeks we rescreened the sentiment change for all 500 stocks and reallocated the portfolio accordingly. The portfolio consisted of long positions only. Dividend payments, transaction costs and taxes were ignored.
The strategy generated returns of 27.6% vs. 14.1% for the S&P 500. There were twice as many winning than losing trades while the average gain was significantly higher than the average loss.