Frank Rotman is the Co-founder and CIO of QED Investors. Before that, he helped start Capital One. Frank and I go deep on both of those founding stories, plus one of QEDs first big winners, Nubank.
Frank also gives us a crash course on fintech, lending businesses, and crypto use cases; his hot takes on the venture asset class, with lots of advice for emerging managers; and a case study on how high valuations too early on are bad for a startup.
If you’re thinking you might have listened to this one already, you’re probably right - it was published two weeks ago. Just sending out the transcript now. I have a few more from the past year that I’ll send over the next few weeks to get fully caught up on the backlog.
Timestamps to jump in:
4:22 Starting Capital One in 1988
7:04 Spinning out as an IPO
10:42 Starting QED in 2008 before Fintech was a category
22:03 Investing early in Nubank
25:11 Fintech opportunities in India
27:38 De-risk investing in new markets
29:55 How financial services have changed over the past 30 years
31:33 Inside a new Capital One credit card in the 90’s
36:31 How most companies launched new cards in the 90’s
39:46 The most profitable types of credit card customers
42:00 Mistakes founders make building credit businesses
48:33 Frank’s “Three Body Framework” for VC
54:48 Losing strategies in VC
1:03:39 Unpacking why high valuations are bad for startups
1:16:20 Frank’s journey in and out of crypto
1:24:23 Actual use cases for stable coins and NFTs
1:34:09 Unpacking the lending supply chain
1:41:44 The difference between Fundamentalist and Revolutionary investors
Referenced:
Find Frank on Twitter and LinkedIn.
👉 Find on Apple, Spotify, and YouTube
Transcript
Find transcripts of all prior episodes here.
Turner Novak:
Frank, welcome to the show.
Frank Rotman:
Well, happy to be here.
Turner Novak:
I'm excited to have you on. You guys have a really interesting story, you and QED. Going back to the earliest days, it actually didn't start with QED, it started with another company. Can you tell us the story?
Frank Rotman:
The story of Capital One is basically the precursor to the story of QED. Nigel Morris was one of the co-founders of Capital One, and I was hired directly out of college to be one of the early guard to create what eventually became Capital One.
The origin story of Capital One actually backs up a little bit because Capital One was a spin-out from Signet Bank. Signet Bank was the 50th largest bank in the country at the time, and Rich and Nigel were basically hired to help Signet Bank figure out what to do with their credit card business.
Rich and Nigel, who were the two eventual co-founders of Capital One, were consultants at the time with SBA. They were consulting to the financial services industry and really had this amazing idea called the information-based strategy. Which was a way of analytically managing a financial services product, and treating everything using the scientific methodology of experimentation, until you figure out what works, and when you figure out what works, you roll it out.
Turner Novak:
How is that different from what people were doing at the time?
Frank Rotman:
Well, it's very different from what people did back then.
Turner Novak:
What year was this, by the way?
Frank Rotman:
Rich and Nigel moved over from SBA to Signet Bank to help them with this strategy, and I think it was 1988. This goes way back in the way back machine.
Back then, a lot of financial service products were basically just assembled. You had one version of it and you put it in the market and either people bought it or they didn't buy it. There was no differentiation, there was no customization. If a bank had a credit card, it had the same terms for everyone.
In fact, the entire credit card industry had basically the same terms across all credit cards at the time. It's 19.8% interest rate with a $20 annual fee, and credit limits were very similar.
Turner Novak:
Did you get points, and bonuses, and airport lounge access?
Frank Rotman:
No. Back then, it was a very simple transactional product where you could borrow if you wanted to, but it was very basic. This basic credit card product crossed the entire industry.
Rich and Nigel said, "Why don't you treat people as individuals? Based on the data about how that individual is going to perform on the credit card, you should be able to offer them differentiated terms." The idea of treating each person as its own unique thing and customizing the product to get the right product in the right channel at the right time to the right customer, was very, very new back then.
Rich and Nigel left SBA, they joined Signet Bank. I joined them while it was still Signet Bank and then eventually, Capital One, which was the credit card division spun out from Signet Bank in an IPO.
Turner Novak:
Why did you guys spin it out?
Frank Rotman:
It was so incredibly successful within the bank that it was consuming all of the company's resources and then more. It was growing fast, it needed capital, it needed people resources. It basically was consuming everything that Signet could produce.
At that point, spinning it out seemed to be the right answer to capitalize it properly, to create its own culture, to figure out how to move faster and unbundle itself from a bigger organization. In some ways, it was one of the first fintechs out there before there even was a word for fintech. It was the fintech of the 1990s.
Turner Novak:
What was it called? Was there a buzzword back then? Or, information-based lending basically was it?
Frank Rotman:
It was a monoline credit card company. It basically was a single product that was spun out from a bank that it was a credit card. and the entire company existed to basically manage credit cards, which was a very different thing than being part of a full-service bank.
Turner Novak:
Were credit cards not a big products line for financial services companies back then? Is that part of why?
Frank Rotman:
Credit cards were still very valuable. There were a couple of big players in the space, but the big growth in credit cards really took place during the 1990s, like a huge surge in credit card volume, people taking on more than one credit card. Part of it started with the customization of cards. The more that it was a generic card, the less you needed more than one.
When the cards were customized and all of a sudden, you had unique functionality for different cards or different cards for different purposes, the industry opened up. Capital One also was the pioneer in launching products like balance transfer, the subprime credit card. There were new incarnations of credit cards that opened up the market to new types of consumers and new behaviors.
We're at the forefront of doing that, but the key was using information to make decisions, putting testing out into the marketplace using the scientific methodology to say, "If I have a hypothesis and I can structure a test around it and put that test into the marketplace, see if I'm actually correct or incorrect with my thesis. If I'm correct, then I can take advantage of that and roll out the product in a bigger way into the market."
That was really the origin of going from Signet Bank to Capital One, and Capital One being a pioneer in fintech. Nigel and I did that for a number of years together, and Nigel left in 2004, I left in 2005.
I actually left to go build a student lending company and Nigel left to go explore a bunch of other things and figure out what else was out there that was interesting. We joined back up a few years later to form QED investors. There was another person as well, whose name is Kara Buhonic, who was also ex-Capital one, and the three of us got together and created what became QED.
The concept of QED was that we were really ex-operators that were looking at new businesses that would compete with or support the banking ecosystem. Either they would spin out products that would compete directly with banks, or they would be service providers to banks because they were offering products that the banks could consume.
Turner Novak:
Was this a common category at the time? Was there a buzzword for it? Or were you guys making this up or creating a category?
Frank Rotman:
There was no buzzword for it. We were really a hammer in search of a nail in a lot of ways. The only thing-
Turner Novak:
That's usually not a good thing, right?
Frank Rotman:
No, it's usually not. The one thing we knew in life was financial services. By the time we left Capital One, it had gone full circle and it had started buying banks and it had become a full-service provider of banking services. It had an auto lending business, it had a mortgage business. It had gone full circle from monoline back to a bank.
Turner Novak:
You basically used a credit card as a wedge into the market, and then launch a bunch of products, and then suddenly looked like everyone else at the end of the day, maybe a little different still.
Frank Rotman:
Yeah, I mean it became a full-service bank, but with the basis of an information-based strategy, an analytic layer underneath everything in a different way of making decisions.
When it came to QED, we saw ourselves as operators masquerading as investors. We knew one thing and we knew one thing well, which was financial service products. We thought that there would be an opportunity to help young entrepreneurs that were trying to innovate in this space, figure out how to avoid landmines and just see around corners.
We started at precisely the perfect time, 2008, the global financial crisis. Nobody wanted to fund anything in fintech, and there wasn't a word for it, by the way. Nobody wanted to fund anything in financial services when the banks were just incredibly distracted of figuring out capital adequacy issues and their balance sheet and dealing with the global financial crisis of 2008.
There really wasn't a lot of activity in this space, and of course, we hang our banner up and we say, "We're open for business," and there's no business. It was a very interesting time.
I think globally, the year that we first started QED, about a billion dollars of venture money globally was deployed into fintech. If you look at during peak madness when fintech was super, super hot, it was over $140 billion. Those are not the same ecosystems from when we started to when it hit peak.
It's come down a bit since then as have a lot of venture capital, but it's still running at about $40 billion a year worth of funding.
Turner Novak:
I think as of today, like Q2, it seems like it's truly back again, fintech specifically, from what I've seen, at least in the data.
Frank Rotman:
At the end of the day, the services that are provided by fintechs, by banking, by insurance companies, they're not going anywhere.
Turner Novak:
They're still pretty outdated too.
Frank Rotman:
These are necessary products. You've got all these major pillars of banking, which range from borrowing, to lending, to storing money, to investing money, to transference of risk through insurance products. These products are going to exist, which means as long as there are new innovations taking place and new S curves being built, fintech is going to be a thing.
Turner Novak:
Was there a specific insight or opportunity set at the time? Any specific examples of these? Anything that stands out back in 2007, 2008 when you were really getting things off the ground?
Frank Rotman:
If you think about the fact that the word fintech really started in the 2009, 2010 era, that tells you a lot about the maturity of the space when we first invested.
When I think of v1.0 of fintech in this last super cycle, which started around 2008, it really consisted of UX/UI reduction in friction for taking applications and applying for products, as well as APIs. All of a sudden, you could take an application that otherwise would be a paper application or be a long form application online with people calling you to gather information.
Instead of an application process would take a significant amount of time, it would feel magical because you could click a couple of buttons, it would automatically connect with data sets that were out there. You could actually do it on a mobile device instead of sitting in front of a computer or talking to a human being. You could reduce the friction of applying for a product and figuring out whether you qualify for it with a few clicks instead of a long-form application.
That innovation may sound very simple because now it's just normal to reduce the friction of applications. If you go all the way back to the early 90s when I was part of Capital One, the number one way to apply for a credit card was through filling out an application that came in the mail. You physically had to fill it out and you had to mail it back and then someone had to data enter all this information and then had to push it through some models, and then weeks later, you might end up with a credit card in your hand.
Turner Novak:
I remember when I opened my first brokerage account in college, this was, I don't know, this was like 2010, I think, if I'm remembering the date. I had to go into a branch of a, I think it was called Scottrade at the time. I think it's been acquired by a couple of different companies over the years. I had to actually go in, I couldn't even open my account online.
I had to go into a branch, and then you paid a fee too. I think you paid $7 with every trade or nine. I think I signed up for Scottrade because they had the lowest commission. Even then as a college kid, didn't have a lot of money.
Each time I bought a stock, I never bought anything for less than a thousand dollars because I just needed to make the economics work. I knew I'd have to sell it, so I knew it'd be about $14. I knew there'd be that fee drag, I'd lose about 1%. I need to get an extra 1% return to justify the purchase.
Then now you can go on Robinhood and spend 10 bucks and it's "free." It's interesting how it's all changed over really just over the last 15 years.
Frank Rotman:
Again, v1.0 was really about these UX/UI changes, mobile adoption, APIs into data sets, reduce the friction and make it feel magical.
I think fintech has continued to evolve since then. The core manufacturing of the product looks different. The core storage of money looks different. The movement of money is going through different rails than it did. There's a lot of innovation that's still taking place in the ecosystem. v1.0, the first wave of fintech innovation, a lot of it was in lending.
A lot of it was on the application. A lot of it was in customer origination. The core product behind the scenes was relatively similar to the products that they would get otherwise, and that's changing now.
Turner Novak:
Then you alluded to this a little bit earlier before we started recording, but the first fund, the LP-base, what did that look like?
Frank Rotman:
We had the privilege of being able to do this independently. Nigel, being the co-founder of Capital One, left with a fair amount of money. He was very comfortable deploying some of that money as the core LP in the fund to figure out whether this new model that we were creating of being operators masquerading as investors actually was wanted in the marketplace.
Two, were we any good at investing? We had to figure that out. We knew we were good ex-operators, we did not know if we were good investors. The capital-base was Nigel, Kara Bu, and mine, it was all of our money. By that, it's mostly Nigel's money through the proceeds of what came from Capital One. That gave us the luxury of being able to slowly prove out a track record and show that we were actually good at attracting the right types of businesses and that we could identify the ones that would end up being winners.
We invested in a lot of very interesting companies in the early years. Credit Karma was one of them, Klarna was another one, TransUnion was another one. Braintree was one of our early investments that turned into an $800 million sale to PayPal. We were able to find a lot of very interesting companies doing interesting things.
Part of it's because fintech wasn't a hot sector and there wasn't as much activity going on as there were in some other spaces. It was very easy to hang our shingle out and say to other venture capital and private equity firms, if you ever come across something in the banking space, in the financial services space, just give us a call and we'll take a look.
Turner Novak:
Might be too hard for you, you might not know anything about this, but we'll give you our take. We'll take a look at it.
Frank Rotman:
It was very good to be a second set of eyes for VCs that weren't experts. They were generalists and they weren't experts in financial services. We had good partnerships with firms like Excel and General Atlantic where we would see a lot of the interesting things that were coming through their pipelines. And a couple of hours of digging and talking to the founders and understanding what they're building, we could save the venture capital firm dozens of hours' worth of independent research and really cut through all of the clutter and figure out what mattered to the business and what advice we would end up giving them if we would invest.
Then the quid pro quo was if a round came together, we could tuck ourselves in and we could figure out how to be part of the story going forward. Our origin really was helping other VCs and private equity firms until we built a bit of a brand of our own. Then, backing winners ends up attracting other companies because they know you're in business and they know you're writing checks.
Eventually, what happened is fintech became a real thing. Volumes started to pick up, companies started calling us directly instead of going through other VCs because they recognized we were real and writing checks. The opportunity set just got bigger than our own capital base. Eventually, about six or seven years in, we crossed the Rubicon and started to take LP money from outside LPs.
Turner Novak:
Then what was the story then when you were pitching LPs? Were they interested in the returns? Were they interested in the sector? Was it sexy at the time? I'm trying to think maybe when this was, this was like mid-2010s?
Frank Rotman:
Yeah, mid-2010s. The story was giving access to a fund that had a differentiated strategy, a differentiated position in the marketplace, and a track record that was proving to provide outsized performance.
Our early funds did quite well. I won't talk about all the details, but let's just say we had three or four funds in a row that were top decile funds. When you have that type of track record in an emerging category where you have unique access because you have unique value to provide back to the founders, it became a strategy that some of the LPs looked at as a way of planting some money in an ecosystem that could provide outsized returns. We were very early to some geographies as well, so our first investment in Brazil happened to be Nubank.
Turner Novak:
Amazing. That's a good one.
Frank Rotman:
A good one where we had been talking to David Velez before he started the business, many, many years before he started the business about the idea. He was actually part of General Atlantic's investing team when I first met him and used to fly around South and Central America with him doing diligence on financial services companies in Latin America.
Turner Novak:
Well, and the story was he was trying to find companies to invest in, and he just couldn't find one that he really liked, and ended up starting Nubank because of that. It was something like that?
Frank Rotman:
Yeah, we saw a lot of companies together. I remember being in Mexico City looking at some microfinance companies with him. We were in Columbia looking at a credit bureau. We were in Brazil looking at an insurance company together. We looked at a bunch of businesses together.
There was a small General Atlantic team that was focused on Latin America. We spent time with them. The whole time, David was saying that the structure of the Brazilian market, really, it lent itself to a new entrance to the credit card market that actually would create a product that consumers wanted.
Turner Novak:
It was actually, now that you talked through the Capital One story and knowing the Nubank story, pretty similar, started with a credit card, was a wedge into a broader financial services company. Nubank is pretty broad now.
Frank Rotman:
A lot of our early conversations was mapping out what we did in the early days of Capital One to even get it off the ground.
The story with David, he eventually left General Atlantic to go back to get an MBA at Stanford. And then he left Stanford to join Sequoia. At Sequoia, he kept bugging Doug about saying, "We need to fund a credit card business in Brazil. If one shows up, we're going to fund it." Doug eventually just kicked him out and said, "Here's a million bucks. You keep talking about this thing, just go build it."
We were talking to him in those very early days about, I mean, I remember a whiteboard session in our offices where we were mapping out that first year of what the infrastructure would need to look like and what he would need to build, and we ended up investing very early on in Nubank.
From there, it led us into other investments in Latin America, and now we have 30+ investments in Latin America. To LPs, we were very early to the Brazilian and Mexican and Latin American ecosystem.
Similarly, today, we have a small but growing portfolio in India. We have a small but growing portfolio in Africa. From a geographic standpoint, we're not scared to dip our toe in the water.
We don't jump in and splash a bunch of capital around and just say, "We're going to make this thing work." We study an ecosystem, we make some investments, we understand how it works before we decide to really double down and go big. Again, that's the practitioner in us. We just don't declare success without seeing how things work, but we're early to some of the trends and aren't scared to actually take some risks.
Turner Novak:
I guess India would be an interesting example. I'm not super familiar with the Indian ecosystem. I've made two investments. One of them was pre-launch, one of them was as the product was launching. I usually just invest super early. It's mostly on the founders.
How did you look at the Indian ecosystem and decide, okay, there's a bunch of places we could be right now, why are we leaning into India?
Frank Rotman:
It's interesting, because of our track record in terms of backing notable companies, a lot of volume comes our way just proactively. It finds our way to us through other investors. It finds our way to us by founders directly trying to contact us.
It's not a surprise. We're fintech specialists, we talk about how we're global. If someone wanted to fund a new business, we're an obvious stop on the train station schedule to come talk to. Not a surprise that we get a lot of deal flow even in regions that we're not active.
We were seeing a lot of very interesting deal flow coming to us out of India and out of Africa and some other places. When you start to see a lot of deal flow, it means you can at least study the things that talented entrepreneurs are interested in, and you can get a feel for where the innovation is taking place.
Until you spend a lot of time with the founders and actually spend time on the ground, understanding an ecosystem and is it ready, in which areas is it ready? Is the ecosystem ready? Is the currency ready? Is the geopolitical risk ready? You can only learn these things by spending time and you can only learn so much by being a student as an outsider looking in.
There's some work you have to do by actually investing in the ecosystem and getting to know people. You don't need to make a splash and put a billion dollars to work overnight. You can make a few investments, get to learn an ecosystem, understand how the businesses are being built before you definitively say that a geography is really, really interesting.
The more we've been learning about India, the more we've been leaning in. Africa is still early. We're still learning quite a bit about Nigeria and Egypt and a bunch of the other major geos. That's kind of how we do things, we make some investments and we learn.
Turner Novak:
Thinking about dipping your toes into a new sector or theme or thesis, do you think about it as, is there a certain stage? Do you just skew earlier? Do you skew a certain size of the portfolio? Like, we'll put 3% of this fund in India, which might be a couple seed investments? I mean, what's the general process that QED uses?
Frank Rotman:
Yeah, I mean, we do have some internal caps on doing new things. It's just good discipline from a portfolio construction standpoint. But we can always revisit those things. If it ends up that something is more attractive than we thought, we could decide to put more money to work. But it's more guidepost for portfolio construction and pacing than anything else.
When we decide on a new geography, we spend a lot of time studying the geography as well as the investment. And you hope to make a good solid, usually safer investment if it's your first investment. So instead of swinging for the fence with a brand new thing, in a new geography with a new founder, like new, new, new, new, new, you try to at least normalize a few of the dimensions and say, "Look, maybe this is a business model that we know really well, but it's a new geography." And that might be one way of de-risking it a bit because we know a bit about the business model.
Turner Novak:
And you have a network of other founders that have solved similar problems, it's a little bit easier to say, "Hey, you want to help them out a little bit?"
Frank Rotman:
So it's not a surprise in India, there are two credit card companies that we're invested in. One is one of them, and the other is Jupiter. And that's not a surprise because we know credit cards quite well, right? Capital One story, the Nubank story, Mission Lane in the US we're building. We know credit cards.
There's an early wage access company in India that we invested in. We actually, from the ground up, funded an early wage access company in the UK. We're invested in one in the US. We have one that we're invested in Mexico. We know the business model incredibly well, which makes it more comfortable to invest in a different geography.
Turner Novak:
One question just in terms of FinTech more broadly. Over the past, I was going to actually say 15, 20 years, but I guess you've really been in it since the early '90s. How has it changed over time? Are there maybe phases to it? What are some of the big differences? Just thinking back early, early days, maybe 15 years ago and then today? Anything that stands out?
Frank Rotman:
Yeah, I mean the beauty of most financial service products are that they're really just a configuration of a bunch of numbers sitting in computers.
Turner Novak:
That's a good way to think about it.
Frank Rotman:
You're not actually manufacturing anything physical. I mean, yes, there's a card or there's something associated with it, but I mean the majority of the product are numbers sitting in computers.
Turner Novak:
Yeah, just a bunch of spreadsheets.
Frank Rotman:
So you have ledgers, right? I mean, most of banking is about ledgers. And those numbers are stored in a very specific way, knowing who is ownership over what. And you're moving things from ledger to ledger.
But I think the ability to actually manipulate those numbers, configure them in different ways, it's only improved and gotten much easier over time, right?
I mean, the old systems of record that we were using back in the early days of Capital One, things were just very inflexible. If you wanted to launch a new product, it could take you months, if not a year, to assemble just the product in the system to make the changes that you needed in order to get it into the market.
And Capital One was renowned for doing a lot of testing, but the preparation for testing and then the channels that you would use to deploy it could take you three or four months to get a direct mail campaign ready to launch and get something in someone's hands.
Turner Novak:
How did you run a direct mail campaign back in the '90s? Because today there's APIs, just digital interfaces, right? You can upload an image and set all the parameters and go, right?
Frank Rotman:
Yeah. I mean, I remember when I was building this new credit card business at Capital. And that was one of the babies of mine in the early days building that. We had a very large mail drop, like this was a major mail drop for the entire year. We would look at things in quarterly mail drops. And even though things would go out monthly, you had these big campaigns that you were planning for.
I remember flying up to Minneapolis to one of our print vendors to actually watch the process because this was a really important one that had to go right. And watching a day in the life of an application actually get printed with all of the mail pieces around it and get stuffed into envelopes and put into the different test cells, and then eventually it turned into seven tractor trailers worth of mail that ended up getting shipped out from that mail campaign to go out to consumer's mailboxes.
And the whole process from probably ideation to execution was probably four months, five months. Making changes is very hard. So if you decided at the last minute that you wanted to change a test cell or you wanted to test a different price, a lot of your ideas were four months old or five months old. And then by the time you execute it, you're hoping that the market hadn't changed so dramatically that your product wasn't relevant anymore.
So it was a very, very different in actually building product and planning and getting things out the door. The speed of learning was a bit slower on the acquisition side.
For risk-based products, the speed of learning is identical because you can't accelerate time. If I'm making a three-year loan or a five-year loan or an open-ended credit loan, if I need 12 statements of data to understand if a customer is going to perform, I need 12 statements of data. I need a year worth of performance in order to see if things are matching our models. And you can't accelerate that even today.
So for a lot of businesses, you can learn a lot, you can launch them easier, you can configure them in much better ways, you can reduce the friction for application. But if you're annuities, you need time to figure out how the customer is going to interact with your product and perform, and there's no way to accelerate that.
Turner Novak:
So thinking back, when Capital One was in its heyday, did you have more of an advantage because you were data-driven, because you could make smarter decisions that didn't need to be changed? Or am I thinking about that right?
Frank Rotman:
Yeah. We thought about it as having this cheat code. So what we would do is we would do tests on 1/10 of 1% of your target audience into a market. And you would then get a read, you would get a response read to see if they were interested in the product. You would get early utilization to see if they were going to use the product. You would get early risk reads to see what type of size payments they were making with what frequency.
And if they were performing like other vintages have in the past, you would map them to old vintage curves. But six months in or nine months in, you might actually have an idea about the profitability of this strategy in the market.
And guess what? If you randomly tested it on 0.1% of the population and now you like what you're seeing, you have a thousand times that opportunity to basically go and drop direct mail and put it into the market where you already know the answer.
Turner Novak:
So you go from 10,000 mailers sent out to 10 million, or whatever the sample set is?
Frank Rotman:
That's right. I mean, we would usually start with tens of thousands or hundreds of thousands depending on what you're looking to learn. And then you would roll out and it would be in the millions or tens of millions. And at that point you kind of sort of knew the answer, right?
Because even though there is variance in markets change, in a test and learn scientific environment, you can control for enough of the dimensions that you have a pretty good idea of what the answer is going to look like when I roll out that mail.
So again, the more scientific you are about testing and structuring, the more that you can learn on small samples and then roll out when you already know the answer. And we try to help our companies at QED, we try to help them think through their learning agendas and do this appropriately, given the differences today and the new channels and how you would structure tests just very different today than it was back then.
Turner Novak:
So just to make sure I get this right, people used to not test? They would just say, "All right, we're launching 10 million of these things and we hope it works"?
Frank Rotman:
You basically had a three-headed monster for just about every product. You had the marketing group that was trying to create the lowest price product possible because that was the way that they could get the highest response, or they would give as much as possible to the customer.
Turner Novak:
Like kickback sign-up bonuses?
Frank Rotman:
Whether it was bonuses or credit limit or the annual fee or the APR, they just wanted the most favorable customer terms possible. So you had the marketing group, and they were responsible for the cost to acquire.
You then had the risk group. And the risk group, their goal was to say, "Let's eliminate as much risk as possible." So they want to say no to as many customers as they can, and they want to really target only the lowest risk customers, which by the way, drives up acquisition costs if you're saying no to a lot of the people who are applying, right? So the marketing person and the risk person would typically fight at a traditional financial institution.
And then you had the treasury department. And the treasury department said, "Let me tell you what it's going to cost us to manufacture this product, what is our internal cost of funds and our funds transfer price." And by the way, the treasury department at most banks was a profit center. So they would bake in a profit to the bank and charge the business unit for doing that. So even if they were paying out 1% in deposits, they might internally charge 3% and charge the P&L owner of this business whatever their funds transfer price is.
So the treasury department is trying to run a profit center. The marketing department is trying to get the cheapest product out there with the most favorable customer terms. And the risk department is trying to cut out as many of those customers as possible because they were responsible for just making sure losses stayed below a certain level.
And at Capital One, one of the big innovations is we said all of this is like an integrated holistic decision, because higher pricing actually drives the nature of the customers.
If you end up raising price because your funds transfer price ends up being too higher, the treasury department wants to make a certain amount of money, that might attract actually high-risk customers, and that then drives risk, which then drives your acquisition costs.
If you actually drop price, you might actually have better customers that are applying, which means you can approve more of them, and it might actually lower your acquisition costs. They might actually stick around longer.
So we operated using NPV models. Rather than these individual dimensions that different people controlled, we literally made integrated annuity-based NPV decisions at the customer level, and that was a big innovation at the time.
Turner Novak:
So since you have so much experience with credit cards across years, vintages, geographies, customer types, I'm super curious, what are the best and the worst types of credit card customers? And maybe we all know the answer to this, but I'm just curious if there's anything that might surprise people on what are the best customers for a credit card company.
Frank Rotman:
It's not a surprise. I mean, first of all, it's different by segment. So if you think about a new to credit or a subprime customer, they're worried more about getting access to credit than they are about the cost of the credit. So very different dimensions that you're competing on when you're dealing with not having a lot of information about customers or trying to figure out if customers are rehabilitating themselves after experiencing some credit issues.
For those customers, identifying the low risk customers within a high risk segment is really the key to winning. Figuring out who you can give credit to and have them use it responsibly is the key to success because it's not a demand issue.
Turner Novak:
Yeah. Who doesn't want money?
Frank Rotman:
Right. So it's about making sure you can get it in the right hands.
And in credit-served segments, the prime and super prime credit segments, you're competing on very different dimensions. It's not about access anymore. A lot of times it is around the cost of the credit. And a lot of times it's around the rewards that you end up getting or the cache that you end up getting for having that cart.
So you're competing on status. You're competing on rewards programs. You want to make sure that you have enough transaction volume running through the system that you can make money at the margin. You're hoping that the customers may be occasionally revolved, but by and large, they tend to be just transactors that are using it as a convenience vehicle. So that's a very different segment.
And then on the business side, you can imagine there are businesses that run a lot of money through cards, and the key is finding the responsible businesses that are going to use the card in a way that makes the economics work. But every segment is different. So a credit card isn't one thing, it's a bunch of different things.
Turner Novak:
And then you mentioned a little bit some of the common pitfalls in fintech, and just generally having lessons learned to share with founders.
What are some pretty common trip ups that you see founders run into? Common mistakes, easily avoided. Hopefully some people listening can be like, "Oh, that's smart. I won't do that."
Frank Rotman:
Well, I mean, the interesting thing is, if you want to simplify the world into a framework, there are businesses that an MBA, talented MBA, as an outsider could look at, analyze, tear apart, build back up, and they would actually understand it pretty well. And then there are industries where if they did that, they would get it precisely wrong.
When you're dealing with highly regulated industries like most financial service products, an MBA would get it wrong. An outsider looking at how you would build a credit business, they would tear it apart and they would put it back together, and they would be so wrong in how you would go about launching a credit business or where the dimensions are, where the things are that you should care about, how you would build credit policy.
And you see this all the time. You get an ex-Googler who just thinks about data as data, and they're like, "Wow, these credit card companies and these lenders, they're not using all the data that's out there. We're just going to be better at using data than everyone else."
And you look at them and say, "Have you ever heard of Reg B?" Right? Which is fair lending and what data you're allowed to use in models and how you can use it and when you can get in trouble with the regulators. They don't understand what it takes to actually not just take data and build a predictive model, but actually do it in a compliant way in an industry that is actually watching these things very closely.
So operating in a regulated industry means that you have to understand how it actually works, not the way you want it to work. You can't brute force your way in and say, "I don't like the way this industry works. I'm going to change it." Right? Regulated industries are not a space for entrepreneurs to say, "I'm going to impose my own will on the space." There are people who have done that over time, but you better be prepared for giant fights.
Uber got into a giant fight knowing that there was some gray space about employment law and everything that they were doing. And they had to fight those battles in order to win.
But in financial services, are you going to fight the battle to get the laws changed? Because in the meantime, by the way, you have to be compliant. There's not a choice to say, "Let's not be compliant and then fight a battle in the court of law." No, there are rules, there are regs, there are associations, there are regulators that hold you accountable. And they can actually shut you down. And it happens quite a bit.
So for a lot of startup founders, what they want to do is they want to get their MVP into the market as quickly as they can. And in the FinTech space, a lot of times you're assembling vendors, you're assembling third parties in order to be able to manufacture the product that you're selling into the market.
You might need licenses, you might need charters, you might need to borrow them, work with a third-party who has those licenses and those charters in order to get into market. You have systems of record. You have processes that you have to march customers through when you're onboarding them, whether it's KYC or AML-type procedures for every time that you move money. You have to follow all of these regs.
A mistake that a lot of startups make, they look for the third-party providers that are the easiest to work with and might actually be cutting corners, which is why they're easy to work with. So in some ways, being too loose is a feature for getting launched, and it's a flaw for the stability of whatever you've learned and the ability to remain an entity that's not in trouble. Because if your third-party banking vendors end up getting in trouble, you're going to be affected.
So you've seen what's happening with the big Synapse and Evolve fiasco that's out there right now. It is a big mess. And that's what happens if you have third-parties that you host yourselves on or you rely on that might not have everything buttoned up.
So we spend a lot of time with our FinTechs as they're launching, making sure that they have the right assemblage of Lego blocks. We don't like cutting corners. In fact, we insist on doing things the right way. And from a governance standpoint, when we're on the board, we make sure that everything is being looked at through the right lens and a very conservative lens.
So if you come to QED, don't expect us to say, "I love the fact that you're about to get in a fight with the regulators. Let's go do it." And I know of some firms who they are like that. I'm not going to name them. There are some firms that they love the idea of getting in a fight because they feel like that's what's going to differentiate their companies because if they win the fight, they're the first ones to innovate in the space.
Turner Novak:
I mean, I could see that. That's true. If you win the fight, no one else fought, it's differentiated, I guess. And if you staff up on the team to win the fights at the firm level - it's not your strategy obviously - I guess I could see it.
Frank Rotman:
Yeah. And if you win the fight, by the way, then everyone else has won the fight and you fought the battle. So it's not like you get unique positioning if you win the fight. Like, new regs get written, and if the new regs get written, everyone has to adhere to those new regs.
We are friendly with the regulators. We are friendly with the banking ecosystem. We are not anti-bank. We think that there are a lot of very good banking partners out there. A lot of the FinTechs actually are building to support the banks, not to compete with them.
But there's a lot of room and space for great companies to be built. There's no reason to be antagonistic to different parts of the ecosystem.
Turner Novak:
Switching gears a little bit more towards the VC asset classes as a whole. You've written quite a bit about it. I'll let you talk through your thesis on how it's all changing, but I think it'd be really interesting just to go a little bit deeper on what do you think is going to happen over the next couple of years.
Frank Rotman:
So I wrote a piece a few years back called The Three Body Problem, and it really defines kind of a framework for thinking about the VC ecosystem as an asset class. So I encourage anybody who wants to go into depth, feel free to find that piece and let me know what you think. I mean, my DMs are open, I'm happy to talk.
Turner Novak:
I'll throw a link in the show notes. If you're listening, you want to read it, we'll throw a link so you can read it easily.
Frank Rotman:
Yeah. The general thesis is that you have three major constituencies in the ecosystem. You have the founders that are building great companies.
You have the LPs that are basically funding the VCs, basically giving them money to do their jobs. And then you have the VCs that are basically trying to sift through all of the founders to find the ideas that are great ideas with the founders that have the capabilities of building those great ideas out. So you have three major constituencies.
There's been a Cambrian explosion of startups. So you've got this gigantic base of the pyramid with all of these companies that are being formed.
But at the same time, you have a Cambrian explosion of VCs out there that are saying, "Look, we believe that we have insight into this sifting exercise. We believe we have insight that can differentiate the good from the bad and figure out which ones have potential and which ones don't, or which founders have potential and which ones don't."
And you have a Cambrian explosion in the LP ecosystem where whole new parts of the money machine are now looking to get exposure to VC as an asset class. VC as an asset class was very tiny if you go back a couple decades, and now it's actually a much larger ecosystem because the LPs are deploying more money.
So this is like a giant matching problem. The founders are looking for VCs. The VCs are looking for the LPs. The LPs are looking for VCs. The VCs are looking at companies. It's just, like I said, a big matching problem. The most stable VCs are the ones that actually have a product that can be consumed by the LPs and can be consumed by the founders, and it just fits both of them incredibly well.
So if you think about what the VCs are as a product, it's about access. It's about the ability to analyze. It's the ability to win deals. It's the ability to support the companies on the other side. So it's kind of a see, analyze, win, support, and then ultimately exit industry.
There's a product that's wrapped around that, and a VC defines that product to the LPs and says, "Look, if you deploy capital and give the capital to us, this is our see, analyze, win, support, and exit methodology. This is why we are going to be able to produce outsized returns," and you'll be able to describe that product to them.
At the same time, once you have the capital, you have a product that you have to be able to describe to founders, right? Like, "If you take our money, here is our product. Here is the advice that we will end up giving you. Here are the resources we will end up giving you." Ultimately it's a combination of money and advice and access.
Turner Novak:
Is it usually the same pitch to both? Am I saying the same thing to a founder that I'm also saying to an LP? Or do you think they kind of speak different languages?
Frank Rotman:
They are similar, but not the same, right? A founder doesn't care if you have access to other founders, right? They don't care about that.
Turner Novak:
They probably also don't want you saying, "We get the lowest valuation and we buy the highest percent of your company also."
Frank Rotman:
That's right. If part of your value proposition is, "We're Sequoia and we get a Sequoia discount because everybody knows that we can help move the curve on the outcomes," that's a piece of the product for the LPs, right? Not a piece of the product for the founders that are basically paying that premium in order to have Sequoia. But it's the same thing.
It's that Sequoia has the ability to move the curve on outcomes because of the ecosystem they've built, the access they have to companies, the advice they give, the ability to deploy capital when other people couldn't. They're pieces of their business that actually make sense to founders and make sense to the LPs.
So for a VC to really have a strong product, you have to be able to ask and answer the question why. An LP that has perfect information on all of the VC opportunities, why would they deploy money within your firm? Why would they give it to you? Where does it fit in their portfolio? And why is it a smart decision for them to deploy money with you?
And at the same time, you have to have a product that for a right-hand tail founder with an amazing idea, why would they take your money? What about that value proposition would make them pick you first? Not just pick you if there's a lineup and they've got 10 choices to assemble like a team, but why would they pick you first?
Because by the way, in the VC ecosystem, if you are not getting positive selection and being picked first, you're getting adverse selection, and you're in the exhaust of someone who is being picked first. In an ecosystem where the power law defines everything, you don't want to be in the exhaust of other firms, right?
So imagine if you're a second tier VC and you don't get to see the companies first, you only get to see the companies that the tier one firms already said no to. It doesn't mean that you can't succeed. Tier one firms are not always right. They turn away businesses for all sorts of reasons, but you're looking to find the cream of their exhaust. That's really hard to do. Again, it doesn't mean you won't succeed, but it's really hard to do.
Turner Novak:
So what do you think people get wrong then in terms of some of that positioning? Or maybe what do people do right? What's a winning strategy? What's a losing strategy?
Frank Rotman:
I think that a losing strategy is when a VC is trying to compete in a bucket that they don't belong.
Let's, again, simplify the world into the two types of businesses. There are consensus businesses and non-consensus businesses, right? So consensus businesses are ones where pretty much everyone understands the trend, understands why the business is being built, understands the economics of the business and what it will take for the thing to succeed.
There's still a bunch of operational risk associated with a lot of the businesses, but consensus-type opportunities, most people would say, "Yeah, there's an opportunity here and this probably is going to work, but they've got to work through a lot of stuff in order to build the business."
But for consensus-oriented deals, if everybody sees it in a similar way, then the player that's going to win is the one that's going to be most helpful to the business or has the lowest cost of capital to the business, offers the best terms to the business. There's a way of winning a consensus-oriented deal that a lot of the big generalist firms that have deep pockets have earned the right to invest in a lot of the consensus deals, right?
Then there arenon-consensus deals, where maybe everyone would call you crazy because 10 things have to go right, and if only nine of them go right, you still fail.
Turner Novak:
That's so hard.
Frank Rotman:
It's hard. And you look at it and most people, most investors will look at it and say, "I don't want to take that risk. I don't want 10 coin flips in order to invest."
Turner Novak:
And a lot of people just say, "We'll just pay up in the next round once you de-risk that. Once you figure that out, that channel, new product, hire the CTO," whatever the thing is.
Frank Rotman:
That's right, collapse it from 10 coin flips having to work to five. And then maybe we'll talk to you again and we'll fund the business. So the problem is when you have someone who has not earned the right in a category to try to compete.
So I can't tell you how many Fund 1 GPs are out there with a strategy that says they really want consensus deals. And you're like, "Wait a minute, if this is your Fund 1, how are you going to compete with Andreessen? How are you going to compete with Lightspeed? How are you going to compete with Sequoia? Why you? Why would they take your money?"
And the answer is they probably shouldn't. And if the answer is they shouldn't, then if you end up winning a deal, that actually does say something, that probably is adverse selection at work. It doesn't mean that you can't be right every now and again, but building a whole portfolio about trying to win deals out of someone else's exhaust is not a winning strategy.
So for a lot of Fund 1 managers, the answer is you have to build a track record by taking non-consensus risk. You have to be the one to say yes when everyone else is saying no to a very risky idea, and then you have to be right.
So if you're a new Fund 1 manager, you have to have a differentiated strategy in the market to basically build that track record. And by the way, there are Fund 1 managers that can compete for consensus deals because they bring something very specific to the table for a specific type of deal.
Maybe they were an early employee at Uber, and they've got this giant network that they built and they have connections within an industry, and then someone wants to fund a business that competes in the space. They might be the perfect investor that could help open doors for them.
We come from the banking ecosystem, and we have a Rolodex that's very different than a lot of other people's Rolodex. So we've earned the right to actually compete for consensus deals in financial services because we can open lots of doors, but we also have earned the right to compete for non-consensus deals because we understand the machinations of how you assemble these products. We probably have a better idea, due to our intuition and experience, about the odds of certain things working.
We have the ability to put a learning agenda together in a very organized fashion. I always like to learn as much as you can for the least amount of money and the least amount of time. And we can help assemble that learning agenda because we understand financial services and what it takes to get things done.
But again, a big mistake is trying to compete for a type of deal that you have no business actually winning.
Turner Novak:
How does that most commonly play out? Would it be you and me are competing on a deal, and it's a perfect QED, it's down the fairway, you guys are probably going to win it. Do I just come out and say I'll double the valuation? Is that what most commonly happens?
Frank Rotman:
Yeah. So for second-tier firms, and I always hate that term because it's creating-
Turner Novak:
Yeah, it's so derogatory.
Frank Rotman:
It is derogatory. And I don't mean it that way, but I think people understand a less well-known firm, a less well-seasoned firm, however you want to categorize it.
A lot of times, they have to win on price. And for companies that are at their series B or series C where they feel like they have a good board and they have good advisors, sometimes price matters. And that might be the differentiated strategy where they pay up to win until they build a track record of success, at which point they might not have to pay that premium because now they're attracting the right deals.
Turner Novak:
And that hurts the LPs too because the returns will be lower than if you just had a lower price.
Frank Rotman:
Yeah. You're in a pickers game at that point where it still works for LPs if you're picking the right companies. But on average, if you're paying more than everyone else, it means your returns should be lower than everyone else's unless you're a better picker. And that's a hard thing to do over time.
Turner Novak:
It's a hard thing to prove too. It's pretty hard to just say, "We're going to pay a higher price and we're never going to be wrong." That's an okay strategy if you've done it over a long period of time and you’re going to continue to do it. But I don't think you can go out with a Fund 1 to an LP and just say that's what's going to happen. You have to prove it out.
Frank Rotman:
And it does happen a lot where we'll be competing for a deal and there are two or three other firms that are in there, and then all of a sudden, it turns into a bit of a price war because someone else is like, "Oh, my God, we're going to lose this to QED. Look at what QED has to offer. The best thing we can offer is price." So we'll give them more money, quantum of money, or we'll give them a higher price. Or we're seeing founder bribes come into the market when-
Turner Novak:
Founder bribes?
Frank Rotman:
Yeah. So a founder bribe would be something like, "This is the value of the company, but we'll set aside some extra stock and reload you so you don't take the dilution from the round." So the founder bribe is basically saying, "Let your other investors take extra dilution, but we're going to keep you whole."
Turner Novak:
Like we're going to do a stock refresh, we're going to grant you some shares in a couple months or...
Frank Rotman:
Which effectively gives a higher price to the founders than to the investors. So there are interesting ways that other firms can compete, but ultimately, the advice most seasoned investors will give to most founders is that you should not let price determine who you pick. The players that are around the table, as you're building the business, it's a seven to 10-year relationship that you're building with them, maybe longer.
And these are people that you're going to be relying on for advice and connections, and maybe even additional capital checks downstream. So choosing the right partner when you're playing a multi-turn game is really important. It's not a transaction. It's not like I won and you lost, like a transaction. This is a relationship that has many turns left to go before you get to the finish line.
So picking the right partner is more important than picking the best price.
Turner Novak:
Yeah. So on price, do you feel like 2021 valuations are coming back? When I say that, what does that invoke in you?
Frank Rotman:
Yeah. So I was a pretty vocal proponent of just seeing the ecosystem for what it was during peak madness and saying this is a bit crazy.
Turner Novak:
And why is it bad, by the way? If I'm a founder and think my company should be valued more, can you give us a quick why is it bad for a valuation to be too high?
Frank Rotman:
Building a business using venture capital is about de-risking your business in stages.
So every founder, let's go all the way back to pre-seed or seed when it's just an idea. You start with an idea. You start with what I call a problem statement. And that problem statement is basically saying, "Let me tell you about something out in the world that hasn't been solved and is causing profound pain to a large number of people, or businesses, or whoever your target audience is."
And you try to convince the venture ecosystem that this problem is profound and that ultimately, there's going to be economics in it if you can solve this problem. People are willing to pay you to solve this problem. So that's the first statement every founder makes, a problem statement.
The second statement they make is a solution statement. And the solution statement is saying, "Remember that big problem that we just talked about? This is our idea about how we can solve that problem. We think it can be built. We think we can build it with a limited amount of money. We think that we're going to be able to get paid for this solution because it's making a profound problem go away. And we think that ultimately, you can build a business by pairing this problem and solution statement together."
The third statement they're making is a go-to-market motion statement, which is saying, "Guess what? The biggest tax in most businesses isn't finding customers and getting paid for the good work that you're doing solving a problem. So here's our go-to-market motion and here's how much it's going to cost us to basically scale the franchise and go from subscale burning money to scale and making money." The go-to-market motion helps define that. So that's the third statement.
The fourth statement is a founder-market fit statement. "We believe these problems are tough, but we believe we are the right team to do it and here's why."
But then there's the fifth statement, and the fifth statement is a financial statement. And the financial statement is basically a magical spreadsheet that describes an amazing business with a bunch of assumptions in it, but it plays out how the founder sees the business being built over the next three years, five years, seven years, but just expressed in numbers.
And what happens is if the assumptions are correct in that spreadsheet, you build an amazing cash machine. There's no reason to start a business if you can't make up a spreadsheet with assumptions in it that defines a very, very good business.
So the point of venture capital is to take these five statements, break it down into pieces, take that magical spreadsheet at the end and say, "Wow, I have all these assumptions. How do I narrow the range on that assumption? How do I learn about whether I'm right or not right? How do I get proof or anti-proof?"
And a lot of it is putting money into the market as an asset allocator, right? So you as a CEO, you as a founder are an asset allocator and you're putting money into the market to do one of three things.
You're either building product. You're testing product. Or you're putting money into scaling your customer base and capturing economic rent. The first two things burn money. The third thing is the only thing that creates enterprise value.
So the real key to venture capital is being very, very good at putting money into those first two buckets to learn as much as you can, as quickly as you can for as little money as you can so that you can put money to work in the third bucket, scaling your customer base and capturing economic rent so that eventually, you become a real business. That's what venture capital is about.
Turner Novak:
And then the real business is you take the spreadsheet, the amount of cash that's generated, collected by all the economic rent, and the company is basically valued on what kind of cash you can produce at the end of the day.
Frank Rotman:
Yeah. At the end of the day, the value of a company is the intrinsic value of the cash machine you've built, plus the option value of what you could build in the future based on the surface area you're covering and the quality of the team.
So the intrinsic value of the machine is saying, "How much money can I put into the machine and how much money will come out of that machine?" If I have a machine that I can put a dollar into and it produces $10, that's a good machine, but it's only interesting if I can put more than a dollar into it.
If I can only put a dollar into it and it produces $10, I don't really care that much. It's irrelevant. I can't build a big business. The unit economics look fantastic, but how do I scale that thing into something interesting?
But if I take that same machine and I say every month, I can put $10 million into it, and within three years, I get $100 million of economic rent back, that's a machine that's incredibly valuable. And the goal is putting money in motion at scale, figuring out how to put money into a machine at scale that produces a lot more money out of the machine at scale.
So when you talk about valuing companies, the first thing you have to do is say, "How much money is it going to take me to de-risk this thing and build that machine?" Because the more money it's going to take me to figure out whether I'm right or wrong, the bigger the outcome needs to be in order for me to justify putting that money in.
So when you start funding companies with too much money too soon, and they get sloppy in the deployment of that money, because now they have money and it was cheap money, cheap money gets spent pretty easily.
When it's expensive money and you're not sure there's more of it, you become much more disciplined. And if you think about what a venture outcome looks like, a typical, very good venture outcome, not a marginal one, but the type of venture outcomes you're looking for, the ultimate enterprise value upon exit, relative to paid-in capital, is usually 10x or greater. So if I build a billion-dollar business, a public print for a billion dollars, if I put $100 million into the business, the venture capitalists are going to do just fine. That is a venture capital style of investment that can work.
And if you're building a $10 billion company, you could put a billion dollars or maybe two billion dollars to work because there's enough money that's coming out of the machine on the other side that it justifies putting that amount of money in.
Because the 10x enterprise value to paid-in capital probably means that the VC's made 7x, right? Because let's say the founders, upon exit, and the employees own 30% of the company on exit. So 10x enterprise value to paid-in capital gets 7x. That probably means the last check written into the company gets a 2.5 or 3x.
And the check before that probably gets a 3x to 5x, and the check before that probably gets a 5x to 7x, and the check before that gets a 7x to 10x. So the way venture capital works, that last check getting a 2.5x, and the check before that getting twice those type of returns, it means that you need to be efficient in your deployment of capital.
So if a company is funding itself at too high evaluation too soon and raising too much capital too soon, it breaks this equation, it makes it very difficult for future capital to come in. It makes it very difficult for venture capitalists to earn their return.
And people might say, "Oh, why do you care if venture capitalists make their return?" Well, if they stop making their return, there isn't another dollar that's going to go into the ecosystem. It doesn't work unless you're underwriting to these type of outcomes.
Turner Novak:
So an example of that playing out might be a company is raising at a $1 billion dollar valuation and it's just not justified. It should be $100 million or something.
And I might be saying the ultimate outcome here might actually only be $2 billion. So if you want me to invest at $1 billion instead of getting my 10 to 7x that I might really need to be underwriting to make all this whole economic ecosystem really work, I'm really only underwriting a 2x.
And we also might be pulling the valuation forward as well to the point of the machine might not even be working yet, so I don't even know if I'm going to get my 2x either.
Frank Rotman:
Yeah, a 2x could be fine if there's a lot of certainty surrounding the deployment of money. And private equity, for instance, can make a 2x work because let's say that it's a three-year investment until exit. A 2x is a 25%+ IRR, which is not bad.
Turner Novak:
That's great. Yeah. Everyone shoots for that.
Frank Rotman:
But you can't do that if there's a lot of risk associated, if there's left hand tail risk associated, if you don't know where that money is going. So if you're deploying into a 2x return in this world of either deploying money into building something, testing something, or scaling something to generate enterprise value, that might be comfortable if 90% of the money being raised is going into sales and marketing where you have a machine and you know what the outcome is going to be for the money that you put in.
But if you're basically raising capital to build product and test product and refine your assumptions in your magical spreadsheet, you can't underwrite to a 2x return and make that math work.
So when you fund a business at too high a price with too much capital, it makes the next round very difficult. Founders don't have an appreciation for this, but in the VC ecosystem, the easiest deal to get across the finish line within a partnership is a company that's a rocket ship. And a rocket ship is defined in so many different ways, but it's not just customer growth, it's accretion of enterprise value.
It's about actually looking at the economics of the business and say, "Have they materially improved from the last time they funded to this time? Has the intrinsic value of the business gone up? Has the option value of the business gone up or gone down?"
The option value can go down in a business if you find that it's harder to build than you thought. The option value could go up if a team ends up deploying the capital incredibly well, and now they have more customers, and those customers love them so much that now you have more optionality than you did before.
But the value of a company is a combination of the intrinsic value of what's been built i.e. the machine, plus the option value of the surface area that you've covered with the quality of the team. So again, if you fund it wrong, you just break the ability to raise capital in the future and you actually put your business at risk.
Turner Novak:
Yeah, it's not intuitive if you don't really understand what's going on underneath. You just think, "Higher valuation, that's great. Who doesn't want a higher valuation?"
One thing I really wanted to ask you about was crypto. You did not really do much with it for a while. You started to dip your toes in, you had an NFT profile picture at one point. You started to get really curious, tried to understand it, and it seems like you're back out again.
I'm interested in just understanding the evolution of dipping in and then pulling back out of crypto.
Frank Rotman:
QED's brand, if you want to boil the ocean, is about being the best advisors to our companies about anything FinTech, bar none. We are the best advisors that you can have. And there's this thing called crypto that had evolved over a number of years from this niche little industry into something that would be in the headlines all the time and be explosive. And businesses were being built and funded and grew into gigantic companies almost overnight.
And our companies were asking us about it. And they were saying, "What do you think of crypto? Should we do an ICO? Should we be crypto? Should we launch a token? What do we do?" And I found that we did not have the ability to be the best advisors to our companies because it was something that we didn't know very well. And there's a long origin story behind why I looked at it and then turned away from crypto early on, but I just determined it wasn't okay anymore for us not to know how this crypto thing really operated.
Turner Novak:
Yeah. Because it was $3 trillion in market cap total, I think, at one point. It was massive.
Frank Rotman:
Well, it's just not having perspective on a new form or fashion of digital money felt like a big gap in our ability to give advice to our companies for something that either was going to be important and they needed to know about it, or wasn't going to be important and they needed to know that as well. But we weren't prepared to actually live up to our brand about being the best advisors to our companies.
So I took it as a personal mission to learn about crypto, and I was challenged by actually one of our CEOs who basically said, "If you're going to do this, you should learn in public because there are so many other people who want to go on this journey, but they're scared."
And they understand me and my persona is about being a first principles' thinker, break everything down into its component pieces, understand how they work, and then rebuild everything through an agnostic lens to understand it for what it is.
So a bunch of people were actually excited for me to go on the crypto journey, because they knew that I wasn't going to just become a believer just because, or raise a bunch of capital to deploy, and then you had to be a believer because you're putting money into the ecosystem.
Turner Novak:
Yeah, I remember you were very pragmatic just about, "Hey, this is what I'm seeing. This is how it works and relates to traditional financial services, and how it intersects. And here’s what I think crypto still needs to figure out."
I remember that framing, I always really liked as someone who's... I feel like I'm probably, overall, in a similar position as you on a lot of this stuff. So I really liked your framing.
Frank Rotman:
It was very interesting because the only way I know how to learn is to become a practitioner.
So I loaded up a bunch of wallets with different types of crypto, and then went on a learning journey to say, "What are the things that I need to experience in order to understand how this works? Who are the founders I need to talk to? Who are the people who are well-known in the ecosystem that I should connect with to really learn?"
And every week, I was writing about my experiences. And in fact, my most viral tweets were ones in the early days of me being on this crypto journey.
Turner Novak:
Yeah, that's how crypto gets you. The viral content. Everybody follows it.
Frank Rotman:
It's interesting because I wrote a piece called The House Money Effect, which was my way of framing an observation about how people behave in the crypto ecosystem. It had three million impressions. It just went viral overnight, and a lot of people started reading about it and reading some of my other things, which was really an agnostic view of different components of what was happening in crypto.
What was interesting is I got introduced to NFTs and NFT projects. And people forget, NFT is a technology first, and then there are these projects that are associated that use NFTs that tend to be more focused on community and the artwork and the images. And usually there's a project behind the scenes. I joined a couple of prominent NFT projects in order to understand what that part of the ecosystem looked like.
And what was really interesting is the day that I started to get ingrained in the NFT ecosystem, was the day that I polarized my audience, which I don't write for an audience, I don't care about the audience per se.
I write because it's a way of expressing positions, and it's a way of founders getting to know me before they get to know me, the way I think and the type of advice that I give. I think it's very helpful in the ecosystem. But I'm not writing to basically build a follower base or do anything with it. It's just a way of getting interesting content, hopefully, out into the world. But it polarized my audience.
And the minute that I started writing about NFTs, a bunch of the traditional FinTech people stopped reading the things that I was writing, stopped paying attention to the traditional things that I was trying to look at.
I got really interesting DMs from people who were like, "Are you crazy? Are you no longer interested in FinTech? Have you gone to the dark side? Why are you over there? Why are you doing that?" It was interesting.
And I'm like, "Well, can't someone learn? Can't someone throw themselves into something and say, 'Look, I'm exploring. I'm trying to get to understand this ecosystem, and then writing about my experiences. Why, all of a sudden, do you think that I'm a less credible person to learn from just because I changed my profile picture to that of an NFT project, or I changed my address to a .eth address instead of the normal address?'"
A lot of this was just part of my journey of trying to experience everything that I could experience in crypto. And again, I have a lot of learnings. I've now spent a couple of years actually interacting with different protocols, with different teams, with different technologies. And there are pieces of the crypto ecosystem that I'm super bullish about, and there are parts of the crypto ecosystem that I'm super bearish about.
But now I feel prepared to actually have a perspective on things and be able to give advice to founders. And we can now talk to teams that are now deploying crypto technology as part of their stack, and give the proper advice to them about the right way and the wrong way to do things.
By no means are we crypto experts like some of the crypto-only or crypto-native funds are. But we actually understand the financial service ecosystem. We understand how to build financial service products. So we actually understand building a business using crypto in a very different way than a lot of the other crypto-native firms.
And we're investors in companies like Bitso, which is like a version of Coinbase in Latin America. That's the type of business that we like and we understand fairly well. We're investors in a company called Shakepay up in Canada, which is a combination of a neobank with some crypto functionality.
So we think about it as infrastructure, and the infrastructure now just has very different building blocks than the infrastructure of traditional finance. But now we understand it enough that we can give advice.
And by the way, I know that you are asking why I had a profile picture, and I took the profile picture down. I eventually just went back to my original profile picture of me, because it seemed to play well both in the crypto world and in the traditional finance world.
And it was a good balance of just saying, "Look, this is who I am. I'm a real person." And it seemed to be a good balance, but it was amazing how how many people almost took offense that I was leaning in and trying to learn about crypto.
Turner Novak:
Yeah. It is very polarizing to your point of you're in, you're out. Whatever you pick, people have an opinion on it.
And I'm curious then, just in terms of you said there's some things that you did invest in, some things that you have not invested in, what are you bearish on in crypto? What would you maybe advise founders don't spend too much time or tilt too far towards this direction?
Frank Rotman:
It's easier to talk about the things that I find interesting.
Turner Novak:
Maybe that's a better way to take it then.
Frank Rotman:
I find stablecoins incredibly interesting.
I mean, if you think about the ledger problem in banking, you have 10,000 depository institutions in the United States alone. If you consider community banks, credit unions, and traditional banks, you have 10,000 different ledgers being run.
Turner Novak:
And what does a stablecoin do then? How does it interact with the ledger?
Frank Rotman:
Well, the concept is instead of having 10,000 ledgers, that when you want to move things from one ledger to another, you have an interbank settlement process that you've got to move things from one to another. And then internationally you also have tens of thousands of ledgers as well. And the interbank settlement process there is even more complicated because you're settling currency as well as you're settling the actual ledger issues.
The concept of a stablecoin is having a single ledger that has good money on it at all times. And that you only have to worry about the on-ramp and the off-ramp of getting that good money onto the ledger at all times. Then if you literally settle everything intra-ledger instead of inter-ledger, it reduces friction massively.
There's so much money that's moved internationally between ledgers as well as domestically from bank to bank that the concept of a stablecoin with good money on it that operates 24/7 is a magical thing actually reducing friction of all of these ledgers.
So I think stablecoins are real. I think a lot of trade that's being done internationally is going to be settled using stablecoins rather than settled using the traditional interbank settlement process.
And I think that that bodes well actually for the US dollar, because a lot of these stablecoins are going to be US dollar denominated, they're going to be pegged to the US dollar. So I find that a very interesting development. Any company that's dealing with money movement should really pay attention to what's happening in stablecoins because I actually think it's real.
Turner Novak:
So thinking about stablecoins as a ledger, how is it different from just say the Fed or the Bank of England because those... Are those a national ledger or is that a different thing?
Frank Rotman:
Well, every bank is holding money within its own ledger. So it's not like we have a federal government that knows how much money I have sitting in a bank account.
I have my money. Some of it's sitting at Wells Fargo and some sitting at JPMorgan, and my money is sitting in different ledgers. And if I want to move that, I have to move it through a set of rails. So I can move it through ACH, or I can move it through Visa MasterCard rails, or I can move it through the SWIFT network.
But each of these have settlement processes associated with them. And the concept that you could have a single ledger that everyone agrees that has good money on it, backed by a particular currency that everyone understands and you could affect the settlement process directly on that ledger, it means that it can operate 24/7 and be settled almost instantaneously.
Turner Novak:
So right now there's the delay. It's like three-day, something like that for an ACH, and it's because-
Frank Rotman:
There's potential for fraud every time that you're moving between ledgers. There's a lot of issues with it that are solved through stablecoins. And then the only friction in the process is the on-ramp and the off-ramping of money into the stablecoin itself.
But with the stablecoins, if they're built the right way, these ledgers can operate 24/7. So I find that real.
I find NFT technology very interesting. So the projects, people can draw their own judgment on whether NFT projects make sense. But really all an NFT is it's a container for metadata, and it's a container for metadata where you understand the source of the data being published, and then you have controls about who actually owns and can share that data. But it's a container for metadata.
So if you think about a container for metadata where you understand the provenance of where it came from, you understand the oracle of how the data ends up getting into that container. There's a lot of really interesting things you can do.
So think about all of these databases that sit behind walls. So think about, for instance, if you wanted to prove to someone that you graduated from a particular college taking a particular classwork and got particular grades, the only way that you can prove that is by actually giving permission to someone to go to the National Student Clearinghouse, pay them $20 or $50, and basically pull down that data about you. You don't have the ability to prove to anyone. If you actually walked into an employer with a physical degree, they would laugh at you like, that's not proof. That could be fake. The only actual proof is something that you do not have access to.
But imagine with an NFT if when you graduated, they dropped this NFT into a sovereign wallet that you control, and you can actually prove that the oracle that's publishing data into that container for metadata was the university.
And the university basically publishes data on when you went to school, all the classes you took, all the grades that you ended up getting. You now have a container that you have provenance and it's provable, it's controlled. You would have the ability to basically say, "Look, I have evidence on the blockchain through this container for metadata that I went to school. I have proof."
Imagine doing that with credit bureau data that's changing every month. Why can't you have a container for metadata with all of your credit bureau information sitting in it so that you could prove to someone about the liabilities that you have at banks, about your ability to pay bills, about your credit score? Why can't there be a container for metadata that is managed out of a self-sovereign wallet that you control?
Turner Novak:
So this would be less... It's less like the profile picture, the JPEG that you're making money from a price going up basically, and more so-
Frank Rotman:
NFT is a technology. And it's a technology that uses the blockchain in a very particular way to basically prove that certain data exists and has been published by certain oracles, and that is controlled in a very distinct way.
So you can think about all of these proof points that you or someone else should be able to put out into the world. If you buy a watch, what about the provenance of that watch? Where it came from? Who owned it?
There can be a container for metadata with an oracle that actually publishes data to it that's provable so that you know that it's actually gone through certain hands and it's been owned by certain people. So the provability I think is something that's really, really valuable if it's applied through NFT technology.
So I'm looking forward to the next five plus years. I think you're going to start to see NFT technology take root in some forms and fashions. People might not think of it as NFTs. It might end up being embedded into wallets like Apple Wallet is just native to your phone. But the technology behind the scenes I find very interesting.
Turner Novak:
Yeah, I've been kind of trying to think of practical application. It sounds like the US Government or each government entity might have a layer where it's your ID, they prove that you are Frank. It's not a fake ID.
Maybe the college university, the credentials, the licenses that you have that might be at the government level. So instead of having to interface with the State of New York, you just see what's public. This is a more easily accessible database, essentially.
Frank Rotman:
Your employer could publish your income to an NFT that's in a self-sovereign wallet. And when you go to apply for a mortgage or to go to a landlord to rent an apartment, you should be able to show with proof that you're employed by this employer, and this is the amount of money that you end up making.
Instead, they have to go through a company called... It's a product called The Work Number, which is owned by Equifax now, but you have to go through a third party in order to be able to prove things about yourself that should be provable by information that you own and control.
Turner Novak:
So does Equifax launch that product or is there a startup opportunity there, or how do you kind of suss that out? What'll happen?
Frank Rotman:
Yeah, I mean, that's part of our job is to figure out whether the incumbents are going to end up deploying technology on a new S-curve, or if startups are going to do it. And then they become targets for the big incumbents to buy, or they can compete with the incumbents.
But The Work Number, as an example, is a product that was a sub product within a company called Talx that was bought by Equifax for $1.2 billion. So this is not a small little thing, this concept of being able to prove that you're employed and the ability to prove how much money you make, to verify for different reasons. That's a very, very big business, and I think that there could be an S-curve disruption coming if done right.
Turner Novak:
Another interesting topic I've heard you hit on is the lending supply chain, how it all works. I'm super curious in understanding a little bit more because I truly don't know.
I'm sure you know 100x more than me about this, but if someone were to ask you for the Lending Supply Chain 101, can you talk through it?
Frank Rotman:
Lending is a very simple thing in practice. I should say, very simple thing in theory and very difficult thing in practice.
So what you're doing is that you are giving money to someone for a particular use in proceeds. And for that money, they agree to a schedule of payments, and that schedule of payments unfortunately has some volatility and variability associated with it. So you're trading cash today for an annuity of payments, a stream of cash flows in the future associated with those payments.
Now, in order to make that trade, trading money today for a stream of payments in the future, there's a whole bunch of things that you have to do to put that product into the world.
You have to actually round up money that you're going to be selling to someone else, and that money has to come from somewhere. And a lot of times it comes from insurance companies or it comes from banks basically raising a bunch of deposits and then deploying consumer deposits, transforming them through maturity transformation into loans. But you have to find a source of capital.
Then you have to find a source of customers. Then you have to actually underwrite those customers and you need risk models associated with them. Then you need servicing associated with those customers because you're collecting those payments. And if payments aren't coming in, then you have to figure out how to collect against them or recover against them if they haven't paid you in a long time.
So there's this supply chain that involves everything from licenses that you need in order to operate to sources of capital because you don't own all of this money that you're actually lending out to someone else. You have to take it from someone and give it to someone, and then create these balance sheets and payment flows. You need to round up the data on the customers and the history of their ability to repay and build ultimately risk models associated with the volatility of the payments that are going to be coming in the future.
And there are bunches of operational components that are everything from underwriting the customer to trying to suss out fraud, to figuring out if a business is real or whoever it is that you're lending to.
So the supply chain consists of lots of different things that you have to assemble. Some of them you can manufacture yourself, some of them you can actually assemble from third parties, but you need one of everything in order to make a loan out in the real world.
And there's a lot of innovation taking place in lending because the ways that people are assembling those building blocks are different based on what their business model is. So you can manufacture to sell, you can manufacture to hold, you can run a balance sheet, you can borrow money from players like insurance companies, or you could buy a bank charter and raise brokered deposits. There's so many ways of assembling the building blocks that each model is different in its own way.
Turner Novak:
What do you usually find if you're building a lending company, are there certain good areas to compete on? Where does the hole in the market usually end up being, or is it always different?
Frank Rotman:
It's different depending on the category. If you're building a mortgage business with 30 year money, that's very, very different than a two-week payday loan. And the ecosystem of who ends up buying those assets is very different. Those are not anywhere close to the same business. They both are called lending businesses, but they don't even look alike. They don't look anything alike.
Turner Novak:
Yeah. You might get paid back the entire payday loan before you get your first mortgage payment.
Frank Rotman:
Correct. Right. So how you assemble the capital, how you do your risk underwriting, what your operational infrastructure needs to look like, how you originate customers, it's all different if it's a short duration loan than long-term. If you're direct to consumer, it's different than if you're embedded lending.
I mean, think about buy now, pay later type loans where you're embedded in the checkout flow. That is a very different type of business. And in fact, who's paying you to originate that loan? It could be very different than if you're going direct to consumer. So understanding the nuances of the business that you're building and understanding the cash flows and how you manufacture the product, it's very specific to the type of product you're building.
Turner Novak:
So I had one more question. This was from... I don't know if you know Chris Hladczuk? He was a Head of Sales or Chief Revenue Officer at Meow.
Frank Rotman:
Oh, yeah. I know him quite well.
Turner Novak:
Oh, nice. He said you called the top of the market in 2021 in a board meeting. I don't know if he was there or if just Brandon and Meow told him, is this true?
Frank Rotman:
There are one or two pieces that I wrote at peak madness that basically declared that this doesn't work. The way the ecosystem is working is kind of broken. And I was spending time with companies basically saying, "This isn't going to last forever. We might be at the peak."
Turner Novak:
So do you raise money? Do you try to sell? What do you kind of talk people through at that point?
Frank Rotman:
This is probably a harder topic to deal with in a short answer, but it was a little bit about declaring things were crazy and then not knowing when it was going to end, and then eventually realizing you have to play the game that's on the field. And then playing the game that's on the field right before it ends.
So we have a little bit of stuff in our portfolio that is like that as well, but we were late to the game of capitulating. So a lot of funds have major problems where a lot of what they ended up deploying into was companies at peak valuations. You can look at Coatue’s and the Tiger’s, they deployed massive amounts of money at peak valuations. We still have our problem children that we're still working through, but we kind of saw the game that was being played and tried to advise our companies the best we could.
Turner Novak:
It was definitely a hard thing. I think the hardest part is being someone who started their firm in January of 2021, the game on the field was the worst possible game you could play. That was tough for me personally.
Frank Rotman:
I wrote a piece, it was on fundamentalist versus revolutionaries, and that was literally at peak madness.
Turner Novak:
I'm trying to remember if I read that one. We'll throw it in the show notes.
Frank Rotman:
Yeah, let's see if you can find it. Because a few people have now pointed out that it was literally at the peak when I published that piece, and it lays out the two models and about how they're very, very different.
And I was uncomfortable because I'm more of a Charlie Munger type investor than a Marc Andreessen. I believe a lot more in the intrinsic value of the businesses than in just overpaying for the option value of what a great founder can do. I find that irrational.
Turner Novak:
And that's the piece, is basically there's those two types?
Frank Rotman:
There are these two types of investors, Traditionalist versus Revolutionaries. And they think about businesses in very different ways.
Turner Novak:
Yeah, I could definitely see that. Is there a way to combine both?
Frank Rotman:
That's what we tried to be, where you realize that you're not only... You don't get the option value for free. This isn't private equity where you underwrite the business as it looks with the cash flows you know it can produce, and that's all you pay for on a rational multiple.
Turner Novak:
Yeah, it's a venture capital.
Frank Rotman:
Right. It's venture capital. So you have to pay for the option value of what can be built, but you have to understand what that option value is. And the problem is so much money was deployed over the past handful of years where people knew nothing but ZIRP that everybody thought the option value of a company was a decacorn.
Turner Novak:
Yeah, I think I tell when I'm talking to my LPs, I'm like, the biggest mistake I made was instead of thinking the default outcome was $1 billion, it was $10 billion. That was just straight up a mistake.
Frank Rotman:
Yeah. And for us, when we got into this business, we thought that the default outcome was to snap a product out from a bank, build it better, and then when you get to scale, sell it back to a bank for $200 to $400 million because that's what we had seen in banking. That's what we thought.
So in order to make that work, you had to build a business with under $20 million of capital deployed to get all the way to the finish line, and you are funding companies at sub $10 million valuations. And the first checks that were being written were like $2 million into companies. So it was a very different game that was being played back then,. And then ambition got bigger.
Turner Novak:
Which is always fun.
Frank Rotman:
Yeah. The art of the possible was discovered. When we sold Braintree to PayPal for $800 million in an all cash transaction, it was one of the largest all cash transactions ever in financial services.
Turner Novak:
Yeah, I remember you told me that.
Frank Rotman:
I mean, you now look at Stripe, which is now a hundred times the size of what we ended up selling the company for. No one would've anticipated that was the type of outcome. When I underwrote Credit Karma, I did not underwrite it to a $7 billion outcome. But our first check-in was at sub-$10 million.
Turner Novak:
So I mean, that's the beauty of VC. You underwrite some kind of intrinsic value and there's some option value, and hopefully the option value is a lot higher. I mean, it's impossible to really know. Maybe it is possible. I don't know. Maybe that's what makes the best VCs the best.
Frank Rotman:
Well, this is the thing I talked to Will Quist about this all the time. The zip code of the outcome is generally knowable. And this is where people get it wrong. They believe that any venture outcome with a talented founder is going to be a public print. And there are companies that are never going to be public prints.
Turner Novak:
Just because of the inherent business model necessary to make that work?
Frank Rotman:
It could be the business model. It could be the sector they're in. It could be the space they're in. It could be the product that they're launching.
Even talented founders, if you start in the wrong space or the wrong product, you don't have enough surface area to cover to expand into something bigger. You don't have brand permission to expand into something bigger.
So I could build a niche specialty lending company for mineral rights, and we've got one on our desk. It's about mineral rights.
Turner Novak:
Really? Okay.
Frank Rotman:
I can see how you build a good cash flow machine doing this. So think about mineral rights the same way that you would think about a HELOC. Instead of selling them, why don't we give you a HELOC that can tap into the underlying cash flows that you have from your mineral rights? You can build a business, but how big of a business can you build?
You can size that market, you can understand it, and you can say, "Look, you might think you can build a $10 billion business here. But I think you can only build a $500 million business here."
And you have to fund it appropriately. You have to scale it appropriately. I'm the cheap bastard around the table with some of these specialty lending companies that say, "You have to make money at low levels of scale," full stop. You have to assemble yourself so that you're making money at very little asset production because we don't know how big this thing can be.
And we ended up selling some of our lending businesses to banks and we did fine. But we did fine on a $300 million transaction because that's where probably it was going to cap out and we knew that.
Turner Novak:
So it's really knowing what the exit is going to look like when you invest. I mean, you need to have some idea of is this going to be worth $500 million or $50 billion? You need to have some kind of understanding.
Frank Rotman:
If it ends up being an order of magnitude bigger than you thought, everyone should be happy about that and no one should have paid for it.
Turner Novak:
Yeah, that's fair. And once everyone's paying for it, that's when you get in trouble.
Frank Rotman:
Well, no, when one party is paying for it, that's when you get in trouble. So if the venture capitalist pays for it by basically funding a business at a price that doesn't make sense, that's a problem.
But the point is, if it's bigger than everyone thought, then everyone gets a bigger outcome than everyone thought. So who cares if you funded it at... There's a sharing of economics and the founder will say, "I gave too much to the venture capitalist." And the reality is nobody thought it was going to be a order magnitude bigger than it was. Everyone should benefit from that.
Turner Novak:
Well, Frank, this is an awesome conversation. This was a lot of fun. Thank you so much for coming on.
Frank Rotman:
Yeah, hopefully you get a good edit out of it and we'll see it in the real world.
Stream the full episode on Apple, Spotify, or YouTube.
Find transcripts of all other episodes here.
Share this post