Home / How ‘machine learning’ is changing funds management

Mike Rierson
Big data may change the way some managers invest. A lot has already been said about the predictive power of data analysis on stock prices – things such as chatter on social media. BlackRock has now applied its “data science” research to fixed income portfolios.
BlackRock recently published a research note on the development of “machine learning” and how it is being adapted to funds management. After first being used in the management of long/short equity funds, such techniques are now moving into fixed income management through credit analysis and in macro-focused strategies.
The research note, ‘By the Numbers: Perspectives on Capital Markets’,  says that the new algorithms and modeling techniques drawn from analysing unstructured data, including news articles and other indicators of sentiment, has been much slower and less pervasive in funds management than other industries, such as advertising, real estate, retailing and pharmaceuticals.
Mike Rierson, a BlackRock managing director and head of research for the model-based fixed income group in the US, says: “We think these new data science techniques have tremendous potential to identify and capture systematic investment opportunities for our clients, as my colleagues in our Scientific Active Equity group have convincingly argued in recent publications.
“More specifically, using these machine learning techniques, we can develop highly adaptive investment strategies that respond dynamically to evolving market conditions, can enhance the predictive power of our trading models, and can quantify what used to be purely subjective assessments of tone in an analyst’s report, or in a CEO’s sense of optimism on a conference call.”
The paper says that, despite the generally smaller breadth of fixed income asset class data, compared with equities, there are substantial opportunities to apply machine learning and big data techniques to fixed income datasets as well.
Credit investors, like those in equities, form views on the relative health of individual issuers in the marketplace, and stand to benefit from the long and short insights harvested from the growing masses of firm-specific unstructured data. More macro-focused investors stand to gain as well, as these techniques can be applied to measure the sentiment in bodies of text such as news articles about the general economy, economic strategy research, Federal Reserve governor speeches, and Fed minutes.
But Rierson warns that, as exciting as the research is, it is not “magic”. He says: “Understanding market dynamics and economic insights still matter a great deal, but now we can use that market knowledge and investment expertise to identify and cultivate valuable datasets, guide how we apply our machine learning techniques toward harvesting predictive insights from those data, and then use those insights to develop successful systematic investment strategies.”