Blackrock’s Use of Machine Learning to Deliver Performance
Blackrock, world's largest asset manager, and use of machine learning
Blackrock is world’s largest asset management firm by asset under management total US$6.28 trillion across various asset classes including equities, fixed income, alternative investment, real estate, etc. It’s known for its investing capabilities across both active and passive strategies and its focus on delivering strong performance for its clients.
Despite its large size, the firm is facing increasing competition from passive funds such as Vanguard and other long-only investors and hedge funds. Given the competitive dynamics, Blackrock, like many other asset managers, are exploring potential AI solutions to leverage data and improve investment outcomes.
Blackrock’s use of machine learning
Black believes that evolving and adapting to new technology is important to keeping a competitive advantage in the asset management industry. Currently, BlackRock’s Systematic Active (SAE) group analyses Big Data using cutting-edge technology to trade and invest in various asset classes. The team’s goal is to combine human insights with innovative technologies to deliver consistent alpha-generating results rather than dramatic wins.
So, what exactly is Big Data in this context? It includes information applied to financial data, economic data, news, satellite images, consumer search patterns, and even social media data. Rather than simply employing a quantitative trading strategy, the SAE’s investment strategy is consistent with Blackrock’s overall firm strategy which is to focus on fundamentals and sentiment. In a typical due diligence process of a long-only investor and without the help of machine learning, a portfolio manager needs travel to company site to examine its operations. He or she also has to listen to earnings calls or read research reports, which tend to be very time-consuming. However, with the help of machine learning, the asset manager can use satellite images to examine companies’ on-the-ground activities. He or she can also save time by not having to go through so many texts or listen to management calls. Instead, the machines can read the text – in different languages such as English, Chinese, Japanese and Korean – and generate summary reports. The same logic applies to other important themes that may influence the market including consumer trends, healthcare and social governance. The below table highlights examples of how SAE leverages technology. 
Source: company website
Not only does Blackrock’s SAE arm improves operational efficiency, it has also proved to be effective in generating good performance. As indicated in the chart below, Blackrock’s SAE unit outperformed Blackrock’s traditional equity units in both the short run and long run.
Source: company website
In addition to using machine learning to gain superior insights that help the investment making process, Blackrock uses machine learning to reduce risks for portfolios. More specifically, it offers an operating system for investment managers called Aladdin Risk Platform, which uses machine learning to provide the portfolio managers risk analytics to enable good risk management and informed decision-making. Aladdin can monitor over 2,000 risk-related factors per day and test performance under various market or economic conditions.
Machine learning is an exciting opportunity for Blackrock, both in terms of improving performance and managing risks. To further enhance its edge in the market using machine learning, Blackrock recently announced the BlackRock Lab for Artificial Intelligence and centralized its data through creating a new team called Data Science Core. Blackrock’s long-term goal is to bring the benefits of its technologies to more employees across its business units and its customers. 
In the short run, Blackrock should continue to utilize machine learning to make more informed decisions (such as SAE) and manage risk (such as Aladdin). At the same time, it should constantly examine the output from these technologies and find ways to further improve. It’s very important to ensure the reliability of the data for machines to deliver quality output that can help with investment decisions.
In the medium term, Blackrock should adopt its machine-learning technologies to more business units (not just the SAE team) so that all portfolio managers within the organization can benefit. This will potentially further improve overall performance, firm operational efficiency and reduces costs, all of which will impact its bottom line.
Finally, Blackrock should stay true to its core value and strategy regardless of which technologies they use. It’s very easy for an asset manager to turn into a pure trading house if it solely relies on technology to make decisions. For Blackrock to stay as a long-term investor focused on fundamentals, there needs to be layers of checks and balances in place to manage risks, handle unexpected events and more importantly maintain the human elements of investing (including meetings with management, sentiment, behavioral biases, etc).
- How can Blackrock make sure that the data it gathers are reliable?
- What are other risks involved with using machine learning in making investment decisions?
Word count: 783 words
Student comments on Blackrock’s Use of Machine Learning to Deliver Performance
Do you think in some point of time they should merge Aladdin and SAE? It comes to me that these two products should go hand in hand as the ultimate result is to deliver alpha to the investors, and this require to be able to do both: make good decisions and manage risk. I’m not sure how technically advanced Aladdin is compared to SAE.
Interesting article! It seems that machine learning will have a significant impact on asset managers of all strategies and sizes over the next few decades. A question that this brought up for me is who will win this game in the long run? I wonder if those with the most assets and fees today will be the ones able to invest most in this sort of technology and talent, resulting in a winner-take-all industry. Alternatively, will the old school quant funds (e.g., Renaissance, D.E. Shaw, etc.) be able to leverage their lead in the space to ultimately retain and grow market share? My guess is that the funds with the most to invest in developing these capabilities and strategies will ultimately be the ones that win, and the industry will go through a period of consolidation.
Separately, I wonder if BlackRock is planning on taking this investment in machine learning to the extreme and pull expensive portfolio managers out of the equation. Your article suggests that this technology is an aid to their portfolio managers investment decisions versus a replacement, but is that the end goal? Given that BlackRock has such a leadership position in passive strategies, it seems to make sense that their ultimate goal may be to replace costly investment managers with either passive strategies or computer based strategies. It will be very interesting to watch how this plays out.