Data, Innovation, and the Future of Investment Technology: A Conversation with Neal Pawar

In a recent episode of the Investment Management Operations podcast, Enfusion COO Neal Pawar sat down with host Scott MacDonald to discuss the evolution of financial technology and data management. The following is an excerpt with key highlights of their conversation.

Key Takeaways:

Scott MacDonald: What drew you to the investment space as a technologist?

Neal Pawar: I studied computer science at Brown. After I graduated, I noticed finance firms were doing more interesting stuff with technology than most tech firms at the time. I ended up joining O'Connor Swiss Bank in Chicago, which merged with UBS in 1998.

The team I joined supported global market makers and traders across the world. The system we built ran off a single database in Chicago. Now, if you were a trader in Singapore, Hong Kong, or London, you had to take the network latency to connect to that system, which was pretty slow back then. One of the problems we had to solve was replicating data across all our locations so that traders and market makers could work with a local instance. Database replication technology didn't exist back then. We ended up building our own replication technology and many years later, we were able to retire it in favor of a commercial product. But the fact we were solving those types of problems before the big tech companies had solved them was exciting. It brought the technology to life and spurred a thirty-year career.

Scott MacDonald: Going through your journey, how did you handle change?

During the process when O'Connor eventually became UBS, we constantly combined technologies and brought in trades. I think the most important thing I learned from that time is you need to keep an open mind. Use it as a learning opportunity to grow, rather than hold on to a piece of technology because it's been your baby. After six years on the sell-side, I joined a firm called D.E. Shaw on the buy-side.

During my time at D.E. Shaw, the firm rapidly gained assets under management. When a firm grows that quickly, you're faced with two fundamental questions. Will you grow your business by building parallel verticals to handle the portfolio management, trading, and post-trade processes? Or are you going to combine them onto a single platform?

From a time-to-market perspective, it is often deemed quicker to buy something new, attach it to the side, and get the business off the ground. It takes a bit longer to integrate workflows into existing systems. Fortunately, some of my mentors believed we should not do bolt-ons. We took our time to scale internal platforms to support new asset classes, different flavors, and shapes of trading volume. After a decade-plus at D.E. Shaw, I returned to UBS and joined their wealth management division in Zurich as CIO.

Scott MacDonald: When you take your evolution on the buy-side and sell-side, your lens on the investment management world is probably deeply informed by that. These trading systems were rather nascent. How do you think about that today?

Neal Pawar: Back in the nineties—whether in portfolio construction, trading, post-trade operations, or research—we built nearly everything from scratch. There was very little off-the-shelf you could buy and still be agile. We even built HR, recruiting, and CRM systems before the likes of Salesforce. Now, it's no longer a strategic advantage to do that.

You migrate those engineers or analysts to work on things that will help differentiate you as a firm. For me, that’s the essence of the buy vs. build argument. I like building things that will drive an advantage and allow us to differentiate our offering from the competition. But I also don't want to make things I can buy off-the-shelf with 85, 90, 95% of the functionality straight out of the box.

Scott MacDonald: Tell me a little bit about what Enfusion is.

Neal Pawar: Enfusion is primarily for the buy-side. Asset managers, hedge funds, asset owners. If you're managing portfolios and you need to do portfolio construction, trade against those, and then run risk reports, P&L reporting, and accounting, Enfusion is a front-to-back platform that allows you to do all those things. What underpins all of that is a common investment book of record (IBOR). In the traditional setup, you often end up with different systems playing different roles in that lifecycle. One of the challenges of using different systems is keeping them consistent. You need to synchronize the data from one to the other. Having a common IBOR means that the entire investment lifecycle process is stored in a consistent database. That eventually saves on overhead costs.

The other thing about Enfusion is that it's designed as a SaaS platform from day one. We push out weekly releases for all clients. We're also multi-asset class. We've had clients who started as long-short equity shops and then grew into fixed-income strategies. They don't have to go out and find a new system, they expand with us.

Scott MacDonald: Let’s turn to trends and themes. What are your thoughts on the state of tech innovation?

Neal Pawar: Last year, I gave a talk at a TSAM conference in New York, where I referenced the Gartner Hype Cycle. It effectively tracks media hype of any innovation over time. The path roughly goes like this. The innovation launches, we start to see early use cases, which creates excitement. Often, we tend to overestimate the impact of that innovation in the near term.

The hype goes into full frenzy. Whether it's the internet boom where valuations went through the roof when, in reality, the world wasn't ready for Pets.com or even Amazon. There weren't enough of us on the internet to generate the kind of revenue people expected. People realized it wasn't delivering on those promises, so it went bust and it fell into what Gartner calls the "trough of disillusionment." Sometimes they slowly emerge from the trough of disillusionment. Those who make it are the ones who demonstrate an innovation’s real use cases and reach the "plateau of productivity," which is now where people can take practical advantage.

I would argue that the large language model (LLM) side of AI is currently going through peak hype. Many people see the large investments in building out LLMs. Roughly $90 billion in early-stage investment just over the last year. People are wondering whether companies are going to deliver.

Clients all the time ask me about how they should be positioning themselves. For me, the fundamental answer comes down to data. If you are a financial services firm and want to get into using LLMs to solve business problems but are unsure of which one to use—or whether now is even the time—there are many things you can do to prepare for this potential adoption. Ask yourself the following questions:

One other topic I’d like to bring up in regard to innovation is blockchain. When you look at the internet, which has gone through the full cycle and it’s obviously here to stay, and then you have AI, which is in a much earlier phase. I would argue that blockchain is somewhere in between.

Cryptocurrencies have gone through quite a journey. Now to the point where we have seen ETFs launch on crypto. But what I find fascinating about blockchain is the number of topics it’s spawned. I'll just name a few. Smart contracts is an intriguing concept. Tokenization, stablecoins, NFTs, each of these will take on a smaller hype cycle of its own.

What seems inevitable is that some of these concepts will stand the test of time and make their way into mainstream finance. To focus on tokenization, the notion of taking real world assets and representing them digitally, then storing them in an immutable ledger, such that you can now have either secondary markets or open up new liquidity pools. We're in the very early innings, but we're starting to see a lot of prototypes and ideas coming out of not just startups, but incumbent exchanges who are viewing this technology as something they need to wrap their heads around. Because in a decentralized world, some of their business models might be at risk.

Scott MacDonald: And what's a good use case that we might see coming up in the near term?

Neal Pawar: Going back to tokenization, when you think about a fund—such as a hedge fund or private equity fund—the process a client goes through in order to invest involves many subscription documents, paperwork, customer suitability checks. That involves a lot of back and forth if you're working with an investment advisor. Now there's an intermediary you've got to go through for all of those signatures. We’re seeing several funds adopting a tokenized structure that now takes a lot of that friction away but also potentially creates new liquidity pools, as well as the ability to trade in and out of those without necessarily the same friction we’ve seen historically. And when you see some of the companies starting to back these processes, you can see how this could take off in the next few years. 

Scott MacDonald: I'd love to get your thoughts on the concept of data as an asset class.

Neal Pawar: It's something I had been thinking about as I worked on a lecture series at Emory University's Business School. The course exposed MBA students to frontiers in finance and new asset classes that perhaps one wouldn't normally think about. It could be art, wine, or even data.

Preparing for the lecture series, I realized that data has a lifecycle similar to other commodities like gold. You start by searching for it, prospecting it. Then, if you find evidence, you mine it, extract it, and clean it. Eventually, you give it to a goldsmith who converts it into jewelry. When you follow that value chain and then replace gold with data, it works almost identically.

We've been mining precious metals and commodities for much longer than we've been mining data. What general concepts helped mature the gold industry that haven’t appeared in the data space yet? One thing is standards. There are clear standards around gold. But when we look at data, especially in the financial services industry, there's a dire lack of standards. A big part of that switching cost from one platform to another is because of the lack of standardization and therefore the amount of mapping and reconciliation you have to do.

Another more forward-thinking example is data recycling, which separates it from a commodity like gold. One of the punchlines I often use is that gold can never generate more gold. Data always generates more data.

Listen to the full interview on the Investment Management Operations podcast.