Enhancing Price Momentum With News Sentiment

New research shows that sentiment derived from financial news can improve price momentum strategies. In a recent report from Macquarie Equity Research, they examined 150 Australian stocks as covered by RavenPack over the period 2005 through September 2010. As indicated by previous research, they find that factors based on news sentiment hold relatively low correlation to more traditional quant factors with 12m and 6m price momentum showing the highest correlation (42% and 37%, respectively). In addition, they find that news sentiment has outperformed more traditional quant factors capturing sentiment over the last five years.



While the theoretical returns to the news sentiment signal are impressive, the portfolio has a high turnover. On average stocks in the top quintile are only 50% likely to stay in the top quintile in the following month. This makes the cost of trading the strategy rather high. An alternative, lower cost way to use this information is to use it as an overlay to an existing strategy, for instance as knockout filter. More specifically, Macquarie considers a standard momentum strategy and combines it with news sentiment by doing the following:

• In Quintile 1 of Momentum they knock out stocks with negative news sentiment
• In Quintile 5 of Momentum they knock out stocks with positive news sentiment

From Figure 9, it can be observed that the return spread achieved from a standard momentum strategy is improved significantly by applying a news sentiment overlay (31.8% rather than 20.1%), with only a modest increase in turnover.

Based on the Macquarie study, it can be concluded that news sentiment has the potential to add value to a multifactor model - in this case, as part of an enhanced momentum factor.