Systematic investment strategies are benefiting from the growing “everyday” use of AI, particularly large language models such as ChatGPT. Partly for this reason, these strategies can now more easily find their way into the wealth segment, observed Adam Riley of Blackrock: “When people use the technology themselves, they better understand its power.”
Quant investing was long primarily an “institutional playground,” said Adam Riley, head of wealth for the Systematic team at Blackrock. “Such an investor would always have a specialist on staff who understood the benefits of systematic investing, who saw how the diversification it creates improves the risk-return profile of the portfolio. In the wealth segment, that was not a given. The complexity of quant was partly to blame.”
AI has changed that, in two ways, said Riley. “AI is increasingly becoming part of daily life. As a result, there is growing interest in what data analysis and quantitative methods can add. In addition, AI itself now provides strong tools to demonstrate that added value: where quant investing could previously be a black box for non-specialists, we now have much better ways to explain how the products work, which signals they use, and how returns are generated. AI writes those explanations, significantly increasing the transparency of quantitative strategies.”
According to Riley, this enables these strategies to better reach independent asset managers and private banks. “Before an investment approach can be included in a portfolio, you want the investor to understand it. Thanks to AI, quant investors have become much better at providing that explanation; the language models add context to their product. That benefits both the advisor and the client. Everyone better understands how the strategy fits within the portfolio.”
Half engineers
At Blackrock Systematic, this broadening of the market for quant strategies has led to significant growth in recent years. The team now consists of about 220 researchers, investors, and “technologists.” “Roughly half have a background in finance, while the other half come more from an engineering background,” said Riley. “We cultivate that diversity. Investment ideas often start intuitively and economically, but only become concrete when combined with a technological perspective.”
Large-scale data analysis forms the most important fundamental contribution of technology. Riley explained: “We use a wide range of alternative datasets, licensed, aggregated, and anonymized. For example, data on payments and the use of retail apps, labor market indicators, corporate reporting, and so on. Of course, all under strict controls regarding data governance. Each year we test 100 new datasets, of which we end up using about ten. We track 15,000 stocks and read 6,500 analyst reports every day. Without AI, that would not be possible at this scale. I see it this way: we always had broad coverage, say a mile wide, but only an inch deep. Now we have a mile of width and also a mile of depth.”
Consistent
What does it deliver? Alpha, said Riley, quickly adding that at first glance it is not about spectacular outperformance: “These are mainly active ETFs, with low active risk and low tracking error, for example 1 percent. But if you do that consistently, year after year…”
Riley emphasized that the quality of the investment idea ultimately determines whether outperformance is achieved. “Others have the same data. What matters is the perspective you apply to that data. For example, we found that companies that cite many numbers during analyst calls tend, on average, to have better prospects. We also discovered that companies hiring climate specialists typically have a lower CO2 footprint twelve months later, and that companies located in buildings with certain energy certifications indeed have lower energy costs. Once you can confirm such hypotheses, they contribute those small increments of additional return that you are looking for as a quantitative investor.”