At first glance, the sustainability and mortgage spaces are radically different, if not diametrically opposed. But to Hendrik Bartel, incoming CEO at Infima, an AI-powered mortgage-backed securities modeling platform, and the former head and founder of Truvalue Labs, an ESG-data AI firm, these two sectors are mutually constitutive and equally inspiring. Long-term thinking and wonky sets of data govern both; in conversation with one another, Bartel argues, sustainability can make for smarter mortgage policies, benefiting our changing planet and mitigating volatility in lending in the process. 

In an interview with The Financial Revolutionist, Bartel outlines his path to Infima, shares details of the company’s Seed round, peeks at the startup’s roadmap, and encourages a modest approach to startup leadership. 

This interview has been edited for length and clarity.

The Financial Revolutionist: How did you end up at Infima?

Hendrik Bartel: I was the founder and CEO of Truvalue Labs for a number of years. It was founded on the premise of looking at ESG data and understanding that there are better sources and ways to create ESG data in a faster manner, enabling the analyst or the investor to have better, faster-moving data from an ESG perspective at their fingertips. 

Kay Giesecke, who is a professor at Stanford, founded Infima technologies about two years ago with one of his PhD students here in the Silicon Valley, in San Mateo. Kay is well-regarded on the topic of mortgage-backed securities—how to calculate them, trade them, and use them as investment vehicles.

I got introduced to the founding team about a year ago. I sat on a couple of board calls, and then, towards the end of the year, we all decided that it would be very interesting to move the company from an academic spinoff at Stanford into the more commercial and product-oriented fintech world. 

What has that involved?

We made sure to have enough money to see through the current fintech environment, but also made sure that we have enough money to really grow the company. We went out there and raised a formal round of $5 million with Radical, a Toronto-based AI-specific VC, along with Franklin Templeton, ThirdStream Partners, and some others. 

With that funding supporting you, what market need is Infima meeting?

If you look at the mortgage-backed security market, and if you look at interest rates over the past 14 years, interest rates only went one way and that was down. We were hovering at the near-zero line in the US, and, in vast parts of Europe, we actually hit negative interest rates. 

Then, all of a sudden, look at what happened over the past eight or so months: We've doubled the mortgage rates over the past year. We're now looking at constantly rising interest rates, and we're still living in a world where mortgage-backed securities are living on manually created models that need a lot of manual input and a lot of manual labor. There aren’t many vendors who build mortgage prepayment models.

So Infima has built this machine-driven model that updates prepayment rates in real time. We're seeing upwards of an order of magnitude better outcomes in our prepayment predictions than we've ever seen in the market. So having this kind of data at your fingertips in this kind of environment—with the turmoil that we're living in, with constantly rising interest rates, constantly changing volatile environments, with possibly a looming housing bubble—helps investors, analysts, traders, and originators all access better data at their fingertips. 

Given your experience at Truvalue Labs, how are you as CEO taking your ESG mindset and applying it to a different field?

A lot of the problems facing the mortgage-backed securities space are parallel to what we'd seen in the ESG world. Very manual assessments, lots of manual data being contributed by analysts. Truevalue Labs uses modern-day computing technologies to make faster-moving models. At Infima, we're doing real-time building of models that constantly update themselves. We're constantly expanding our universe. That's sort of the underlying beauty and power of AI and fintech. 

Yet, there's also some stuff that's completely overlooked at this point, and I think that's where it's going to get really interesting in the second half of the year and going into the next year—peeking into our roadmap and peeking a little bit behind the curtain. The topics of climate as well as ESG are now on the forefront at Infima: seeing what you can actually do with our data set in predicting specifically social governance behaviors. Are there predatory loan environments out there? Are there companies or originators with bad governance? How does being in a flood zone affect a mortgage on a house in 15, 20 or 30 years? 

If you're looking at mortgages, mortgages have a 30-year lifespan; long-term investing is ESG. A lot of parallels are there. We're starting to look at all of these strands together and we're starting to weave them together and starting to see that we can enhance our models even further by looking into ESG, climate adversity, turning it around and being able to have very interesting ESG- and climate-related insights in our data sets. There’s a lot we can now uncover with the wealth of data that we're sitting on.

It seems part of the use case of having something powered by AI in the mortgage-backed security space is heightened when conditions are volatile. When foundational parameters change quickly, having AI as the mechanism that changes that data is favorable as compared to having human beings in a data center chipping away at massive data sets. That use case remains with climate-related volatility. But when mortgages were hovering around zero in the US and in negative territory across the pond, what was the pitch for switching to AI then? 

The data that we're using to train our models is loan-level data, so millions and millions of loans. Not just massive packages, not just securities, but the individual loans. You have a much more granular comb, if you will, to explain how these mortgages pay off and how these mortgages move. What are the external parameters when somebody feels compelled to pay off a mortgage? Maybe they go through a refinance, their financial environment has changed, the credit course has changed so that they now feel that they can actually go through some sort of change in their mortgage. Our models also include pre-2007 volatility, and then the economy that went into a pandemic, as well as our current environment. All of these factors are part of our model, and that's why we're seeing that our market model performs so well compared to anything we've seen in the market at this point.

As your client base has grown, have there been other demands or product requests that you've had to accommodate? And what do you think is next for Infima and the mortgage-backed securities space writ large?

We’re firming up our longer-term roadmap. If we take our binoculars out, we see very large applicability in the credit market. We’ve started looking at how corporate credit changes, and how the credit bureaus can be aided or investors can have access to better data before the upgrade or downgrade happens. 

What we've built truly allows going down the road of structured data of deep learning models and exhausting those on predicting corporate credit changes. And the advent of ESG climate needs to be on top of mind of every fintech, because there's only one planet. How can you force investors to make better investments—ones keep us around much longer and allow us to thrive on this planet? So, the wide world of fixed income is where we see our game being played over the next few years.

Any final advice for The FR’s readers?

When we built Truvalue Labs in 2012, that was the beginning of the fintech hype. Now there’s massive fintech hype, and valuations have been out of control and the raises have been out of control. (Until recently.) I'm actually very excited to be building again on this side of 2022. 

Valuations and those things are interesting, but they’re not the name of the game. I think what's the name of the game is really building a product that is interesting, building a financial model that works and is not built on ridiculous multiples, keeping your burn under control, and making sure you're building with a team that is highly motivated.