Everyone is talking about data science. Here’s how JP Morgan puts it into practice.

Paul Quinsee, global head of equities at JP Morgan Asset Management, thought he knew the skills that made analysts stars. Like the talent scouts in Silver ball, Michael Lewis’s bestselling book How Data Science Changed Baseball, Quinsee had watched basic research analysts play their game – albeit in less dusty areas – for nearly four decades.

When Kristian West, head of JPMAM’s investment platform and former global head of equity trading and equity data science, came back to him with the results of a predictive analysis of years of Research notes collected, Quinsee was surprised. He was perhaps less stubborn than the scouts described in Silver ball when they heard what the data “thought” it took to win a game, but the features of the company’s “superforecasters” weren’t all intuitive.

In addition to analyst personality tests, West’s machine learning model “read” years of stored notes, which included, among other things, the analyst’s perspective, how often they encountered companies, the types of models they used and their complexity.

West and his team discovered that the best forecasters write shorter research notes, and more. They saw companies more often, their models were simpler, and the tone and language of their notes were extreme. And contrary to the stereotype of the former Wall Street footballer / analyst, these superforecasters weren’t in team sports.

While quants like DE Shaw and Renaissance Technologies live and breathe data and advanced computing techniques, many traditional companies are still in the early stages of developing proprietary tools for their portfolio managers, who make decisions based on fundamental factors such as the financial potential of a new product. , the visionary qualities of running a business or the ability of a business to survive a global pandemic. “You have either the quantitative or the fundamental. What we’re trying to do is bring the two together, ”West said. “We pride ourselves on our knowledge as fundamental analysts, but how do you see that? “

Managers are at different stages when it comes to working with and assessing the potential of artificial intelligence capabilities and data. As an example, Wellington Management’s Investment Science Group of 70 people focused on the application of data analytics to investment ideas and the development of professional investors, which includes discovering and mitigating the drawbacks of their behavioral biases.

Equipping fundamental portfolio managers with artificial intelligence and data science tools and platforms is also costly. Although the industry competes with Alphabet and other tech companies for talented developers and programmers, surveys show that most asset managers believe it is essential for performance and risk management. Earlier this year, consultancy firm Accenture found that asset managers who had “industrialized and centralized” artificial intelligence techniques on their investment platform were seeing a significant increase in risk-adjusted returns.

JPMAM created a data science team over three years ago, but West redirected the team to focus on four projects. The first concerns environmental, social and governance information, for which, despite the media hype, there is still little data. The team strives to “fill in the gaps” in corporate reporting on the environment, sustainability, employee satisfaction and other issues. The second project is to use linguistics and natural language processing to filter internal and external documents to create predictive insights, while the third aims to give portfolio managers access to aggregated retail and business data. company of parent bank JP Morgan. PMs can see the information on a dashboard in Spectrum, the company’s technology platform, and can create models and cohorts, like one that might examine the spending habits of millennials.

The fourth part is about applied data science and the money ball – essentially, investing and finding alternative data.

JPMAM has spent $ 400 million over the past three years on this massive project, although that figure includes technology development for the broker relationship management team, stock trading (including the trading analysis) and the derivatives team.

In some ways, JPMAM’s machine learning fund, called the US Applied Data Science Value Fund, gives a big picture of the business framework. Imagine a fund that uses a model that mimics what an investor would do. The model “reads” internal and external research, examines business and other data, and then makes recommendations on actions.

Hamilton Reiner, portfolio manager and head of equity derivatives in the United States, said of a data science fund: “You automatically think it’s quantitative. But no, it’s a fundamental investment. But we can consume hundreds of thousands of data points for [inform] this fundamental lens.

Eric Moreau, who left the data science team to become the fund’s portfolio manager, said when investors hear about data science, they often think the fund is analyzing data that no one else has. views to help identify titles. “And we’re always on the lookout for new datasets, but the way we think about data science is to identify hidden relationships, to help size our positions, [and] control of extreme risks in the construction of the portfolio. It is the use of data beyond the selection of the right stocks.

Data science has helped baseball general managers identify the range of skills, beyond big hitters and base runners, needed to win a game. While investing, Reiner said data science has helped him think beyond stock picking. “I have thought my whole career that they are great stock pickers. You Hear About Stock Pickers of the Year, But You Never Hear About Portfolio Building [person] of the year, ”he said.

The company has deliberately built the model to be completely transparent. Machine learning applications, which are essentially computer programs that learn and adapt as they process data, can be inexplicable. Applications learn as they encounter more data. However, many institutional investors need to explain the decision-making behind their investments to boards, directors, donors, and others. JPMAM has programmed its ML model to be able to explain at all levels why decisions were made, so that the company can explain its decisions to customers and regulators. “With ML models, they tend to be quite dark. We have consciously designed and structured [ours so that] it’s self-explanatory, ”West said.

But West said the product was less important to the business than the ML framework behind it. For example, everyone in the business can use the platform to receive alerts about changes that could affect holdings or fund themes. “This framework is important for all of us, especially in the fundamental investment space,” he said. “There may be a theme that concerns a particular PM or team. You can then model this theme in order to [that] if a security [appears] get lost in [it], you can be alerted. While quants do this all the time, this is especially important for fundamental managers, he said.

Much of West’s work relied on the work that George Gatch, CEO of the Asset Management division, did in 2018 to redefine JPMAM’s technology strategy and infrastructure. West had used AI and data science to overhaul stock trading, but the asset manager still built apps in a vertically integrated fashion, which meant developers could build an expensive app that boosted returns for the value team, while another tech group would find source data that helped generate profitable fixed income ideas. Sadly, both technologies were seated in their own world.

In early 2021, after Mike Camacho, head of the investment platform, became CEO of JP Morgan Wealth Management Solutions, Gatch called West and asked him to lead a technology infrastructure overhaul, which includes research, development of investment ideas, portfolio construction. (basically putting all the ideas in the portfolio) and trading. West had already revamped the business strategy, which included the integration of new techniques. Using a global order management system and global machine learning model, 53% of all transactions this year involved automated orders, the vast majority driven by a machine learning model. West recalls the reaction of Mary Erdoes, CEO of Asset and Wealth Management, including Private Banking, and Gatch, who asked, “Can we do this in the front office?

Comments are closed.