top of page
  • Bill

Taking your Data Science Efforts from 0 to 60 with AlphaTech

Updated: Sep 8, 2020

Jordan Vinarub of T. Rowe Price gave the keynote addresses at AIR Summit 4.0 in 2018, and he made it clear to me, and I believe to our audience of money managers, that while technology is essential, it all starts with people. He said recruiting was a critical part of his role as Head of the NYC Technology Development Center and for any organization that embraces innovation. It’s having the right people and investing in them that is so important for success. That keynote led the AIR Summit to further explore what has been called “organizational alpha” and how it could become a more central component to our events and content.

Jordan’s comments resonated with me. A number of past participants at AIR Summit events have been quick to lament that while all the innovation and technology is exciting, their firms do not have the necessary talent to build out a proper data science or innovation effort. There are many ideas and tools, they concluded, but they didn’t have a large enough team with diverse skills to deliver new investment alpha. Or such a team did not exist at all.

Organizational Alpha Starts with Culture

With this insight, we added more content on organizational culture and talent-focused solutions that can help investment managers quickly experiment and implement more innovative ideas. At AIR Summit 5.0 in 2019, we highlighted Auquan, TalentStat, and System2, all companies focused on acquiring or managing talent in the investment process. Additionally, one of the keynote speakers was David Kidder, who delivered a compelling address centered on “Building a Strategy to Complement your Culture,” which pulled ideas from his latest book titled “New to Big.”

More and more, we are finding that there is a strong belief that technology plus human intelligence is the formula for improved performance, not just in investment management but in many other fields where decision making can be complex with many uncertain elements. A core challenge is integrating new skillsets and roles, such as data scientists, into an investment process that may have been successful for many years or even decades without such input. That is where culture plays a major role in whether these kinds of integrations and changes will enhance or cripple an incumbent process.

Using Auquan to Scale your use of Data

Introducing new ideas, data sets, analytics, etc. can be disruptive at the start of any transformation, and many firms prefer to try walking before running. One of the firms that I found interesting and helpful in this walk-to-run process was Auquan. They are a London-based “alphatech” firm that has created a community of 15,000+ data scientists, many with experience from the best academic universities and global technology giants. This community can quickly become an investment manager’s outsourced and highly scalable data science team working on a custom problem or thesis. Using Auquan’s platform, firms can generate models, output, and solutions much more quickly and economically than a typical internal team with limited capacity and diversity.

Since Auquan presented in 2019, they have developed a product called the Portfolio Activity Monitor (PAM). This is one of several products they have designed that will allow an analyst to ingest, navigate, and sift through mountains of data to find the highest value information. The PAM platform utilizes the latest artificial intelligence technologies, such as natural language processing (NLP), machine learning (ML), and knowledge graphs (KG), which can turn data into genuine insights.

In past columns, I have discussed NLP and ML, but knowledge graphs are interesting because they are an effective way to organize information into an ontology – a set of concepts or entities and the relationships between them. From this ontology, an analyst can perform interesting inference tasks through a visual display of relationship latticework.

Creating IP from a Knowledge Graph

These technologies are essential since PAM is digesting information from over 150,000 news sources globally. As an example, an analyst can use PAM’s knowledge graphs to monitor ACME Inc., a UK based automobile producer that has suppliers in China that imports a key raw material from Tanzania. PAM would develop an understanding of ACME’s web of relationships and be able to direct the analyst’s attention to news regarding increased export tariffs affecting the producer in Tanzania.

Piecing together this data across these relationships in a scalable way is virtually impossible without a knowledge graph. Ultimately, this would not replace an analyst’s role in evaluating ACME, but it would augment their ability to quickly connect the applicable dots in the company’s ecosystem and create internal IP from mountains of public data.

Auquan’s PAM is essentially turning data into insights across an entire portfolio and displaying results in an easily viewable interface for an analyst on a daily basis. From there, an analyst can quickly discern potential opportunities where additional research could be performed. I believe this manual deeper dive and decision making is where true alpha can be uncovered.

Winning by Bringing it all Together!

At the AIR Summit, we strive to find technologies that allow a portfolio manager, trader, or analyst to make a smarter or faster decision in the investment process. Clearly, we live in a world where there is information overload which makes alpha generation harder in one sense, but easier in another. Those with the skillsets, the data, the culture, and the tools to determine the value of relationships in the ecosystem of a company, will ultimately win.

65 views0 comments

Recent Posts

See All
bottom of page