INTERVIEW

"We're at an inflection point," says Peter Zangari, Partner & Head of Americas at MDOTM Ltd. "The capabilities we're developing now can take a model portfolio and produce thousands of individualized portfolios based on specific client preferences. These tools can execute what used to be nearly impossible."

Zangari would know. Before joining MDOTM, he led research and product development at MSCI, where he managed a team of 400 professionals and witnessed firsthand the convergence of massive computational power, vast amounts of data, and evolved AI algorithms. Prior to that, he led quantitative risk strategies for Goldman Sachs Asset Management.

The London-based company, whose clients include Zurich, Amundi, Momentum Investments and Lazard Asset Management, has built an AI platform to efficiently manage thousands of portfolios while maintaining personalization for each client.

MDOTM's AI platform - called Sphere - combines three AI capabilities: AI-driven investment insights that support strategic and tactical asset allocation decisions using market, fundamental and macro data (rather than social media sentiment); a portfolio management system that creates and rebalances customized portfolios; and a solution to streamline the creation of portfolio commentaries and reporting with analytical and generative AI. 

In this interview, Zangari shares his thoughts on AI's transformation of portfolio management, the impact of unstructured data analysis, and why some firms are still hesitant to embrace the technology. With a PhD in Economics and Econometrics from Rutgers University, he combines deep quantitative expertise with practical business experience in implementing AI solutions.

Key Takeaways:

• AI enables portfolio customization at scale: seconds vs weeks

• AI interconnects traditionally separate portfolio management functions

• AI unlocks insights from unstructured financial data

• AI adoption succeeds by solving specific problems, not leading with tech

This interview has been edited for clarity and length.

Tell me about yourself 

I joined JPMorgan back in 1994, and as it turned out, I was part of a pioneering team building metrics for market risk assessment. We laid down the foundations that are still widely used today, particularly in how people understand and manage market risk. 

I spent about four years there, during which many interesting conversations began to spin out around opportunities in quantitative investment, and then I joined Goldman Sachs. This was in the 2000s when quantitative investment strategies were really gaining momentum. But by 2011, after about 13 years or so, I had an opportunity to pivot toward the growing field of business analytics at MSCI. 

Over the years there, I eventually led research and product development, which only deepened my interest in technology and artificial intelligence. I've always been interested in technology, but around that time, I started to see the transformative potential of AI beyond the experimental stage. I began to realize the concrete ways it could be applied to real-world problems, especially in finance. 

My growing interest eventually led me to leave MSCI and enter the venture investing space, focusing primarily on tech. By chance, I came across MDOTM, I reached out to a former colleague of mine, and he introduced me to Tommaso Migliore, the CEO and co-founder of MDOTM. We met a few times, talked business—bottom line is, I was very impressed with the leadership team, and I invested in the company, and they offered me the opportunity to lead and develop the U.S. business. It was the perfect blend of technology and finance. 

What was it about AI that really captured your attention?

For me, it was the real-world applications. When I led research at MSCI, I was very interested in how machine learning could enhance our understanding of data. We had a new generation of researchers coming in, people already familiar with technology, and it made sense to tap into that. At the time, we were starting to see the convergence of three key things: massive computational power, vast amounts of digital data, and the evolution of AI algorithms, many of which had been sitting unused for years because the infrastructure simply wasn't there. 

But now, with the availability of both data and computational power, we could dust off these algorithms and begin exploring their real-life use cases. It was a realization of the tremendous potential AI held for creating value, which is why we're here today discussing it.

At MDOTM, we're doing a lot of things, and one is this new capability called “StoryFolio.” Often, clients don't just want numbers—they want to understand what's happening with their investments, particularly during volatile market periods. In the past, firms would spend enormous resources to create reports that explained these fluctuations to clients. I know firsthand how costly that was, both in terms of time and effort. With StoryFolio, we combine Analytical AI with Generative AI to produce insightful narratives around portfolio performance. Clients get a well-rounded view, with analytics feeding into the generative component to produce tailored, client-friendly explanations of what's driving portfolio changes. It's a transformative product that didn't exist before, and I think the industry truly needs it.

How important is AI's ability to examine unstructured data? 

AI's ability to process unstructured data is one of the biggest advancements in the field. Ten years ago, no one was talking about unstructured data—meaning data that isn't organized in a pre-defined way, like written text or images. Today, we can feed such data directly into AI algorithms, which can then interpret it without us needing to create a structured table or spreadsheet. The AI essentially takes this raw data, understands it, and applies analytics to extract meaningful insights. This is just the beginning of what's possible. We used to manually process such data to analyze it, but now the technology does it automatically. It's incredible progress, and I think we're only scratching the surface of what it will eventually be able to do in finance.

My son is in college, and he's taken an interest in being an equity analyst. I told him, "Think about what ChatGPT and other tools mean for equity research." Equity research won't go away, but it'll change fundamentally. 

How is AI affecting portfolio management? 

AI can read information, analyze its impact, and suggest portfolio adjustments autonomously. We have algorithms that are trained to do this and, depending on the specific asset class or market, AI can make these adjustments with considerable accuracy. So, instead of needing an analyst to oversee every decision, we have algorithms that can make those calls independently, with analysts overseeing them to ensure the algorithms are functioning properly. This is particularly valuable in high-frequency trading, where speed and precision are essential. 

In portfolio management, AI has helped bridge the gap between the portfolio management and trading functions, which were traditionally separate. Now, with AI's help, these functions are becoming more interconnected. Information can flow back and forth between the portfolio management and trading systems more seamlessly, enabling real-time adjustments based on market data.

In the wealth management space, AI enables large-scale, personalized model portfolios. For example, a wealth manager could have an investment committee that sets specific market views—whether on sectors, countries, or industries. In the past, translating these views into client portfolios took considerable time and effort. But now, AI can take these views and create a customized model portfolio that aligns with each client's unique preferences. It's an incredible level of customization at scale. What used to take weeks can now be done in seconds, all while maintaining alignment with the core investment strategy. This is especially valuable for wealth management, where personalization is increasingly important.

What's your sense of AI adoption? 

There's definitely a learning curve. In my experience, the best approach is to focus on the problem or opportunity at hand and then introduce AI as the solution, rather than leading with a discussion on AI itself. When we speak to clients, we show them how AI can solve a specific problem, and only then do we explain that it's powered by artificial intelligence. This way, clients see the results first, and they're more receptive to the idea. People use a GPS every day without thinking about how satellites work. They just know it helps them navigate. It's the same with AI in finance—clients don't necessarily need to understand the technicalities; they just need to see how it benefits them.

How do you see MDOTM fitting into this landscape?

MDOTM has a unique position because we've been working in this space for a while, and we're serious about solving real-world problems with AI. There's a lot of noise in the AI field, and at MDOTM, we're focused on cutting through that by delivering tangible results. Our products are designed with clear use cases, whether it's Storyfolio for client communications or portfolio optimization tools that leverage AI to achieve more precision. We're building AI-powered solutions that are tailored to what the industry needs right now, not just chasing the latest trends.

We're at an inflection point. The capabilities we're developing now are just the start. For instance, MDOTM has tools that can take a model portfolio and produce thousands of individualized portfolios based on specific client preferences, whether that's excluding certain sectors or adding ESG elements. These tools can execute what used to be nearly impossible or prohibitively time-consuming. And it's not just about the technology; it's about the impact it has on client experience and efficiency across the board. The industry is evolving rapidly, and we're thrilled to be at the forefront of that change.

ICYMI
Recent Interviews:

• Arta’s Chirag Yagnik on AI-powered wealth management

• Finster’s Sid Jayakumar on AI agents for Wall Street

• Sov.ai's Dr. Derek Snow on AI for fundamental Investors

• Bain’s Richard Lichtenstein on AI adoption in private equity

• Snowflake’s Jonathan Regenstein on AI building novel datasets

Thanks for reading!

Drop me a line if you have story ideas, research, or upcoming conferences to share. [email protected]