Jan Szilagyi Podcast Transcript
Jan Szilagyi joins host Brian Thomas on The Digital Executive Podcast.
Welcome to Coruzant Technologies, Home of The Digital Executive Podcast.
Brian Thomas: Welcome to The Digital Executive. Today’s guest is Jan Szilagyi. Jan Szilagyi has a record for fastest ever Harvard economics, PhD, two and a half years studying under Ken Rogoff. He spent most of his career with Stanley Drunkenmiller at Dukesne Capital, and now the co-founder of Reflexivity, formerly known as Toggle AI, a 39-person startup that uses machine learning to turn institutional grade data.
In the insights for investors of all stripes, he was co CIO of global macro at Lombard ODA and also managed portfolios at fortress. In addition to his economics PhD, he holds a BA and MA degrees in mathematics and economics from Yale.
Well, good afternoon, Jan. Welcome to the show!
Jan Szilagyi: Good afternoon. Thanks for having me!
Brian Thomas: Glad to be here. Absolutely. I appreciate you making the time and I’m early morning Kansas City, but you’re obviously in London today and I appreciate you making the time, Jan. Love doing international podcasts. Obviously, that’s one of my favorite things to do, but we’re going to jump right into your first question, Jan.
You hold the record for the fastest ever completion of a PhD in economics at Harvard. How did your experience with mentors like Ken Rogoff and Andre Schleifer shape your approach to both academia and the financial industry?
Jan Szilagyi: Good question. I think, you know, sometimes there’s not an obvious direct link between an experience in academia and then particularly investment management or fintech.
But I do think that it gives you a way to think about the world in certain frameworks to try to think about a model of the world that That can explain certain implications and so on. So, what I would say it taught me rigor. It definitely encouraged my curiosity and ultimately it introduced me to a variety of quantitative concepts that we do now leverage quite heavily that we are building a fintech platform that is serving hedge funds and other more quantitatively minded clients.
I was very motivated to finish, and I always had other things that I wanted to get to. So that was also a part of it and having supportive advisors was a huge, huge contributing factor.
Brian Thomas: That’s awesome. Again, I think that’s amazing that you were able to complete a PhD in record time, but also at a very prestigious school.
So, I appreciate that. And I know that has basically carried over into your entrepreneurship. So, I appreciate that. Jan, Toggle AI was created to make sense of the overwhelming amount of financial data available. Thank you. Can you walk us through how Toggle AI’s machine learning algorithms help investors of all levels transform institutional grade data into actionable insights?
Jan Szilagyi: Yeah, so, you know, my experience that actually a lot of my team at Toggle also comes from a similar background, which is investment management. Our shared experience with data in that profession was that there was a lot available, but the infrastructure To use and extract insights from that huge volume of data, just haven’t really kept up with the growth that we’d seen in the data itself.
And so you often found yourself, I would analogize this to being in a dark room only with a flashlight and we wish somebody could just turn on the lights and we could see where different data series could be that we could use to answer any number of questions as you’re trying to think about how to reallocate your assets from maybe this investment to another one and so on.
So, Toggle was created out of that frustration, really about the fragmentation of data and the inability to extract all the value that we felt was locked in it. And the way we approached it was to really leverage both the advances in machine learning in and outside of large language models, as well as the fact that over time, computing had gotten through cloud compute much, much less expensive, which really enabled us to process.
Ever larger amounts of data ever faster and through large language models, make the interaction between the machine that was doing the calculation and the processing and the user that was asking the question a lot more straightforward. Again, I think by now, probably most people and of course, everybody who’s listening to your podcast must have played around with the likes of chat and so on.
But the thing that is most striking for us as builders of data analytics software is that we’re able to offer communication with a very sophisticated calculating machine instead of forcing you to press buttons and go to the drop-down menus by simply requesting what you would like to get done and the system understanding this like an analyst might and then in the background, get the data, do the calculation and then answer your question.
Brian Thomas: It’s amazing. And with entrepreneurs, it seems like on every podcast, the story of trying to solve that problem that users are experienced. How do we become more efficient? How do we get the data that we need and leveraging LLMs in your case? So, I appreciate the backstory. And that helps us understand exactly what you were trying to solve.
Jon Toggle AI is now rebranding to reflexivity. What inspired the name change? And how does it reflect the evolution of the platforms capabilities or vision moving forward?
Jan Szilagyi: Yeah, so obviously this is a big milestone for us changing the name. It’s always something that is. A little bit emotional, but I think it ultimately was a name that aligned more closely with how our platform helps with the investment process.
So, in brief, for those of you who may not be familiar with this. Reflexivity. Is the act of examining your own assumptions, beliefs and judgment systems, and then thinking carefully and critically about how these influence the investigation or the analysis process. This is exactly what we try by bringing together all these data analytics tools and data sets.
What we try to do with investors, make sure that they don’t stay with the initial beliefs, make sure that they test them through what we have, because markets. Are an incredibly complex web of interrelationship between both micro forces, for example, how a company runs its business and also macro forces, which is the geopolitical and macroeconomic context in which the company has to operate.
So, inflation interest rates, geopolitical conflicts and so on all of these ultimately combined and are reflected in asset prices. So, Toggle AI soon to be reflectivity really helps investors dig deep into the data to understand the truly important forces that at any one point in time drive stock prices, right?
Trying to sort of pull away the curtain and really understand the key drivers that at any one time are important. For both the upside and downside performance of an investor portfolio, which often results in confronting their initial beliefs about what really matters for a particular asset. So, I would say this, the description of reflexivity, or what it means is the description of the process that we assume takes place when users.
Like portfolio managers and analysts and investment bankers interact with our platform are ultimately engaging and there’s an interesting side note to this, if I may, which is that the notion of reflexivity was actually 1st popularized in the market by a very well-known investor, George, who basically recognize the feedback loop nature of the investing process, meaning that stock market prices, of course, reflect The company’s business decisions as theory would have it, but they also in turn affect those decisions and that creates this constant feedback loop that results in price dislocations that ultimately offer both the risks and opportunities for market investors.
Brian Thomas: Thank you. I appreciate you stepping through that. I know, you know, rebranding is a hard thing to do, especially when you’re founding a company, and you had a name and all that. But I do like how you shared kind of like an analogy of how you can help investors understand what are the driving forces behind investments, inflation, different aspects of economics.
Really do appreciate that. And Jan, last question of the day, as someone who has worked in both traditional hedge fund environments and now leads a cutting edge fintech startup, how do you think the roles of traders and fund managers will change as AI continues to evolve and integrate into financial strategies?
Jan Szilagyi: Yeah, it’s a good question. So, if you look at a platform like ours, our ultimate ambition is to build a completely autonomous AI investment analyst. However, along the way, much like what you see, for example, when people talk about self-driving cars, we have developed features and capabilities that really translate into giving portfolio managers and analysts superpowers, right?
All of a sudden, We are letting them spend a lot less time in trying to figure out how do I answer this question? Where do I find the data to answer this question in allocating that time instead to thinking about other questions? They should be asking if I’m able to answer using data, a question that you’re asking very, very quickly.
And I don’t worry about obtaining the data, running the analysis and so on. I’m really freeing up. That’s most exciting and most imaginative part of your job where you’re trying to think about different hypotheses, different events, different things that could impact the market and therefore help you very quickly eliminate potentially ideas that might be a dead end.
So, let’s say that you’ve come upon a particular stock that you really like, you think the management is very strong, you think the revenue growth is particularly impressive. And yet it’s a stock that does very badly in the environment that you expect will prevail over the next 6 to 9 months. Let’s say high inflation or high interest rates and so on.
If I’m able to show you that very quickly through analyzing past cycles, when this has happened in this stock underperformed, I will free you to pursue other ideas because you know that this is not the moment to pick this one. And so. What I believe increasingly we will see the role of an analyst and a portfolio manager become will be the kind of the creative brain behind a lot of the investment processes and less of a, what I think many of them consider to be drudgery anyway, of, you know, downloading data, cleaning it, aligning it in columns in an Excel spreadsheet and so on.
And I think ultimately, I think that’s going to be very, very exciting. So, when we typically share what we have built with a lot of portfolio managers. The first reaction is, oh, wow, this is magical. And also, it’s going to save me a huge amount of time on things that I don’t particularly enjoy doing.
Brian Thomas: Thank you. Yeah. And that’s awesome. We are seeing the power of a I power of LLMs, big data. Come together and really make everybody’s lives a lot easier. You know, we’re solving a lot of the world’s problems now using the technology. Granted, we need to keep ethics in place and check, but we are seeing some amazing things happening.
So, I appreciate you breaking that down for us. And Jan, it was such a pleasure having you on today. And I look forward to speaking with you real soon.
Jan Szilagyi: Yeah, likewise. I really appreciate having you on a pleasure to talk to you and looking forward to another time.
Brian Thomas: Bye for now.
Jan Szilagyi Podcast Transcript. Listen to the audio on the guest’s podcast page.
