Thomson Reuters launches model to generate alpha from short interest data

Source: Thomson Reuters

Thomson Reuters, the world's leading source of intelligent information for businesses and professionals, today announced that it has introduced a model to generate alpha by assessing the information contained in the activities of short sellers in the US equity markets.

The StarMine Short Interest model profitably ranks stocks based on the observation that stocks with a high number of shares shorted will underperform while those with low short interest will outperform. The Short Interest model accounts for the cost of shorting different stocks by examining the level of institutional ownership as a proxy for the number of shares available to be lent to short sellers, while adjusting for well-known arbitrage strategies such as M&A and dividend arbitrage that can affect levels of short interest.

The Thomson Reuters StarMine Short Interest model is the only such model that employs a short squeeze indicator, which ranks stocks based on their predicted likelihood of large upward spikes in the near future based on medium and long-term price volatility combined with the level of short interest. The launch of the Short Interest model supports Thomson Reuters vision to connect and enable the global financial community by helping professional investment managers generate more profitable investment ideas and mitigate risks, giving them an edge over their peers.

"Thomson Reuters goes beyond the basic shares shorted divided by shares outstanding signal that many managers use," said Dr. George Bonne, director of quantitative research at Thomson Reuters. "By taking into account the cost of borrowing, and really zeroing in on the shares shorted as a result of investors placing directional bets, we do significantly better than the basic signal. What we do is more difficult because it requires merging together multiple pieces of disparate information and data - from short interest, deals, institutional holdings, and fundamentals - but that's what is required to generate real alpha these days. The proof is in both the performance of live data and in backtests of historical data."

The Short Interest model is one component of the StarMine Smart Money suite of models, which leverage information about the acctions of various groups of informed investors, including short sellers, institutions, and corporate insiders. Other models in the Smart Money suite include the StarMine Smart Holdings model, which ranks stocks based on the predicted future change in institutional ownership.

StarMine Short Interest provides a different approach to short interest data, and can add orthogonal alpha (returns uncorrelated with and independent of the financial assets in the portfolio) to other factors and models. Not only has the performance of the StarMine Short Interest model been strong historically, it also has low correlation with most other quant signals. Both long-only and long-short investment managers can benefit from the unique perspective of the StarMine Short Interest model, either in a multi-factor model framework or as part of a screening process.

The model produces daily updated 1-100 percentile scores, where securities with high scores are expected to outperform those with low scores. The StarMine Short Interest model is provided as a daily data feed from Thomson Reuters, as well as through a range of its desktops, such as Thomson ONE for Investment Management and Datastream Professional.

Today Thomson Reuters provides equities and credit based alpha and risk factors for investment managers as well as corporate investor relations teams. Thomson Reuters has long been known for its enhanced sell-side forecasts. One example is StarMine SmartEstimates, which predict the direction of earnings surprises with a success rate of around 70% when the SmartEstimate significantly differs from the consensus estimate. 

Comments: (0)