Source: Thomson Reuters
Thomson Reuters today announced that it has introduced a model to predict which stocks fund managers are likely to buy and sell from their portfolios.
StarMine's Smart Holdings model forecasts institutional investors' future buying and selling behaviour, identifying which companies would become increasingly or less attractive to these investors over the following quarter. This forms part of Thomson Reuters aims to help professional Investment managers generate more profitable investment ideas and mitigate risks, giving them an edge over their peers.
Smart Holdings combines several different Thomson Reuters content sets including ownership data, corporate financial data and a proprietary, predictive measure of analyst revisions, called SmartEstimatesR , to gauge how well a company is aligned with the current preferences of institutional investors.
Dr Stephen Malinak, global head of intelligent analytics at Thomson Reuters, said: "The main driver in our research was to determine what funds are likely to do next, rather than reacting to what they have just done. The basic idea is that we have reverse-engineered fund managers' stock selection processes to predict what changes they are likely to make in their portfolios based on a comparison of individual stocks to the fundamental characteristics they appear to care most about."
StarMine Smart Holdings is based on the underlying hypothesis that fund managers tend to buy companies with particular fundamental characteristics (such as price-to-earnings, estimate revisions, or price momentum). StarMine Smart Holdings determines which of 25 widely followed factors are currently important to a given fund owner, and with what relative importance, to create a purchasing profile for each fund.
This purchasing profile is then applied against 40,000 stocks globally to determine which of them may be attractive to a given fund. Results are then aggregated across funds to provide a gauge of a stock's attractiveness to the current market preferences. The result is a predictive model that ranks stocks on future increases or decreases in institutional ownership, indicatingicating which stocks will become increasingly or less attractive to investors. Eleven years of data reveal that the Smart Holdings model produces consistent positive returns globally.
The StarMine Smart Holdings 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 the newly-released, Datastream Professional.