As listeners know, this is a topic of great interested to me, and I was happy to have the opportunity to dig into it with Lex in a conversation that we’re releasing through both of our platforms.
In this conversation, Lex and I take stock of the BaaS and embedded finance space in a moment of market dislocation to see what it reveals about what works, what doesn’t and where the next opportunities are.
If you don’t already subscribe to the Fintech Blueprint, I’d highly encourage you to do so for the best analysis in the industry. For an example of their content, check out the post for this episode on rebank.cc.
Thank you very much for joining us today. Please welcome, Lex Sokolin.
Illustrated Transcript (courtesy of Fintech Blueprint)
Hi everybody. Really excited about today’s conversation. Will Beeson is back on the podcast and vice versa. Will, I am back on your podcast and I’m so glad to have a chance to talk with you.
Likewise, likewise. This is really good, really good subject matter, which is very relevant for both of us. I think we can have an interesting conversation.
Absolutely. So, the way I would set the scene is that it’s the end of 2022, and this has been the year of FinTech reckoning, the year of the destruction of dreams, the burning of pitch decks, the collapse of valuation multiples, the embarrassment of soft banks and sequoias on top of the usual conversation about fraud, bankruptcy, and theft. So, it is been quite a year and that’s painful, but it’s also a moment to see how things fall apart because when you see how things fall apart, you can actually see what the parts are that had made up the hole, and then you can start figuring out how the whole thing was actually put together and if it’s going to be put together again, are there mistakes that we can avoid ourselves from making?
Source: CB Insights
And so, I’m talking about this in a pretty abstract way, but the particular thing that we’ll focus on today is embedded finance and banking-as-a-service. And this idea that you can make contextual banks and they exist everywhere in the physical world, in the digital world, and this is the dream state. Well, it is still a desired state and likely where we will end up in the long-term, but we’re starting to see the foundations of some of that thesis start to crack from a regulatory perspective, from a business perspective, and you would expect that to happen in a market squeeze that we’re in today.
But maybe you can help us lay down the groundwork for what is it exactly that is starting to crack in the banking-as-a-service space? What are the symptoms of things going wrong or is there anything to worry about at all?
Yeah, yeah. Great. So great framing, and I totally agree that this is kind of the opportunity to take a small step back and think about the construction of the value chain, so to speak, the sponsor bank/ bank-as-a-service model, the neobank or consumer FinTech or business FinTech front end many neobank type providers are servicing the small and medium sized business space. And just reset in a sense, think about what we’ve learned, what’s working, what’s not working. I think today will be as much you and me unpacking this in real time as opposed to bringing any specific conclusions because I think there is a lot to unpack, there’s a lot to think through, and I think that’s what this exercise is today.
Source: Credit Suisse
I guess if I think about it, we’ve observed, those of us who’ve been in FinTech or following FinTech for the past, call it what, 10 plus years now, the very earliest simple era, right? I think you had Shamir on your podcast recently. I connected with him way back in… I don’t know, it was probably late 2020 or early 2021 at this point. What they were doing at Simple very early on, and I think they took months and months if not longer, Brett King, we talked too together about the early days of moving. Those were multi-year processes to find a bank that would work with them to basically provide their infrastructure, their backend so that they could deliver what was at the time the dream of universal consumer, neobanking.
Source: Brett King
So, we’ve certainly come a long way from there. I think the opportunity, and I won’t necessarily dive straight in, I think it’s so useful to frame this, but the opportunity is to think about the value chain, how it’s evolved over time, where we are now, and it potentially applies some of the lenses that are useful in other types of tech analysis. For instance, the infrastructure app cycle, thinking about attention accrual, value accrual development in each of those components of the stack and where we’ve gotten to and where we’re likely to go because I think we’re at an important inflection point. And I guess if I have to summarize it in just a couple sentences, I think the infrastructure was a competitive advantage early on.
Source: Chris McCann
I think the app side ended up proliferating so quickly into such an extent that it outpaced the sophistication and in some cases the scale and the stability of the infrastructure. I think there were ancillary service providers, so kind of infrastructure adjacent companies, Plaids and Alloys, and Socures and other types of RegTech or fraud or KYC, AML data aggregation type players that came about to patch some of the gaps in the fundamental infrastructure i.e., the banking infrastructure.
And that’s kind of taken us to a point where the value chain is brittle. I think the value chain is out of balance, the revenue’s available to the bank backend infrastructure providers versus these ancillary service providers versus the consumer and SME facing front ends. And I think what we’ll see coming out of this crunch and the further new building development we see going forward will be a realignment of that value chain and a reconfiguration, I think of some of those parts.
There’s obviously a lot going on in that. The parts that are sticking out to me, and I always happen to have a slightly more capital markets lens, but I think payments and deposits have similar analogies, especially with crypto assets. The symptoms that stick out to me is that pretty obviously, a whole bunch of subscale neobanks, digital investing companies, brokerages are blowing up, it’s hard to raise capital. The businesses don’t have the potential they used to. They are far less valuable because valuations are cheaper. And on the distribution side, which is I’m going to build an app and acquire customers, the playbook was a growth hacked blitz-scaling approach to grow into profitability.
And so naturally at the edges, companies are going to blow up. And even good entrepreneurs, experienced folks like Bill Harris who had founded Personal Capital, like his latest venture, he shut down because it couldn’t make the economics work. You look at Pipe, so Pipe is a revenue financing company. The entire founding team has just stepped down a few weeks ago after having raised like 350 million bucks. They stepped down saying, “We’re too inexperienced, we can’t run this company.” And so, all over the place you see companies at the edges blow up and then some companies you see getting hit with regulatory censure for blowing past regulation, sometimes through negligence and sometimes somewhat maliciously or on purpose.
So, this is an area where I think we’re both inspired by Jason McCullough’s work who focuses on all the neobanks that have failed. I feel like he’s got a tracker of them. And now the consequences of all of these clients, of the infrastructure providers are starting to put a drag on the infrastructure providers. And the ones that come to mind for me are looking at Silvergate, looking maybe at Evolve, Silicon Valley Bank is apparently under pressure. And so, this banking-as-a-service, selling the shovels scenario, which seemed very attractive when you have distribution that has been blitz scale funded by venture capital. You can’t sustain all that growth because things are so good, but actually, this company has locked out of its customers. What do you do? Or this company in the case of BlockFi, has literally taken all of your deposits and zeroed them out. Meanwhile, you have a credit card open with them and they expect you to pay back your credit card balance.
So, it’s creating these ridiculous, weird situations that emerge out of the complexity of an industry value chain that had atomized all of this infrastructure, created completely different technology layers so that if you wanted to plug into payments, you would go to Stripe, and if you wanted to plug into a banking charter, you would go to a different place. So that’s kind of what sticks out to me. Is that right?
I think it is. And a number of the points that you mentioned, for instance, the blitz-scaling on the app side, the competition, and as a result, now bankruptcy closure of a number of FinTech apps, I think are all symptoms or representations of the same underlying reality, which for me is the fact that innovation on the app side is much faster and demands from the app side are much quicker to evolve than the underlying infrastructure has been.
Source: Capital IQ, Pitchbook, F-Prime team analysis
So, in this case, the underlying infrastructure is banks’ ability and willingness to offer their balance sheet generally via API, to FinTechs that want to use it. So, companies have been doing this again for years. BBVA was quick to the open API based concept back years and years ago now, talking about being a software company and not a bank. Cross River, Green Dot, not to mention some of the sponsor banks that you mentioned evolved, certainly a number of others have been in this space for a while, but it’s tough to see rapid innovation, enhancement, improvement in the underlying banking-as-a-service infrastructure at the bank level that’s commensurate with the changes in evolution, changes and requirements on the consumer app side.
Source: 11:FS, adapted by Simon Torrance
And as a result, I think we’ve ended up in this kind of unbalanced situation. And as a result, you have apps that are unable to fully differentiate, unable to fully compete. You have ultimately people spending a lot of money, as you mentioned, VC backed to acquire customers, but ultimately not able to offer the quality of experience, quality of product that they’re looking for. And again, as I mentioned before, I think the economics of doing that, if you’re on the app side, are not what you would expect them to be.
So, I think actually, what this all adds up to is I think over the next few years we’re really going to see an interesting acceleration, an interesting opportunity for banks that are currently in the sponsor banking model or wish to be, regular regulatory headwinds and compliance concerns notwithstanding. I think there’s a tremendous opportunity for banks that want to be serious at the infrastructure level to really double down, take a more technology focused mindset, think more carefully about what types of tools, automated, oversight, compliance, data management, program management, they want to bring to bear on their offerings, which I think will help fundamentally drive the infrastructure forward.
Now, we’re not talking weeks or months, but I think fundamentally drive the infrastructure forward in a way which will ultimately support this embedded finance dream that a lot of people in the industry, myself included, view as being kind of a natural evolution.
I think there’s something to say about this dialectic between infrastructure and application. And I’ve written about this in more detail before, and I think the original sort of analysis of this actually, I believe it came from Andreessen, although I’m not a 100% confident. And the basic point, which is not that profound, but I think it’s helpful is there’s an ebb and a flow between infrastructure phases of innovation and application focused phases of innovation. And what does it mean to be in a phase? It means that there’s a lot of innovation and variety of startups, it’s easier to get capital for that type of business and there’s broader adoption and in the adoption, there are new things discovered that become the kind of established form factor for that particular outcome. So, discovering how to visually design a mobile app for financial services interactions. A couple of years of figuring that out and then it clicks and then everything looks that way.
Source: Mario Laul, Medium
And so that’s a phase where you would see a boom and then the opposite happens, which is you have an oversupply, let’s say, of applications and they get to a place where they are constrained in what they can do for a user because of the underlying infrastructure, and that infrastructure in the example doesn’t have to be financial. It could be, hey, real player in 2002 was trying to stream videos, but there wasn’t enough bandwidth to actually stream videos for another 10 years. And so real player maxed out its ability to perform on its thesis, to reach its thesis even though the thesis was correct and died as a company, and the experience that people had with it was very frustrating. But the flip towards infrastructure development, which was a stronger internet, the ability to stream and transfer more data, people getting much more comfortable with using the web as their kind of primary media consumption location, all those things made it possible for YouTube. Then the next phase in that category would be the development of the iPhone and then the emergence of TikTok.
It’s sort of the same thematic, but there’s a dialectic between this has gotten as big as it can and it can’t go further because the underlying infrastructure’s not there. And I think for banking-as-a-service or embedded finance, there are things that fundamentally look very similar to me as what’s happened with the sponsor banks and FinTechs, which is more of the original bank tech world. So, you’ve got your core systems, core banking systems with Fiserv and Jack Henry and FIS, and they’ve got thousands of widgets and connected integrations and there are 10,000 banks and every single bank uses one of those three platforms. It’s oriented towards the physical world. It’s generally speaking a pretty tough digital experience. It’s not real-time, it’s not API first, all these things.
But for the world of, “I’m just another local bank for my geographically distinct town,” they do a totally fine job. But when you get to the digital world where now you’ve got people who say things like, “I want to put money into the phone,” let’s say it’s 2008 or 2010, it’s very hard to try and extend FIS, I don’t mean to pick on these companies, but to extend any core banking platform into being API first. And then in addition, the infrastructure had to include not just… It had to include a lot of things that those platforms did not want to give up, which is number one, read data of all the customer accounts that had to be literally stolen through username password scraping.
It has to give up the ability to actually move clients’ money and have read right access to the underlying bank accounts, which absolutely no bank would say yes to unless completely demented in some sort of way. And it took a while for those people to show up and I mean that in a positive way. And APIs hadn’t been designed as a technology in a way where it was dead simple for a mobile developer to show up and integrate them into an application and integrate multiple sets of different API providers into an application and then tie up a bow with that experience.
And so, you had a shift into building out that infrastructure phase, both in terms of the banks that were available to provide their balance sheet as well as the big data aggregation companies, whether it’s I think Finicity or MoneyDesktop or then obviously Plaid as the latest iteration of it. And then once that was available, you started getting a lot of apps, like you say that end of the day started to feel very much the same because what they were doing is providing accounts, not omnibus accounts, but essentially omnibus accounts on the same system. You’re all connected into Cross River or Evolve or whatever.
And so the limitations of what the infrastructure can provide, even if it can give you a lot of the basics, I think have started to come through and developers for each one of these different apps that had to twist their feature sets and integrate other providers and create all sorts of weird combinations, you do end up in situations where you’re renting credit card capability from one provider, you’re renting deposit capability from another provider, your depositors blow up, but you’re still asking them to pay their credit card debt down because you don’t have a holistic system to actually tie this together and make sense of it. And so, I think that dialectic going back and forth is definitely shifting us towards out of the application phase and back into infrastructure.
I would add to that, almost picking up where you left off just there, like any business, as things evolve, you look for ways to optimize, you look for ways to build efficiency into the business to leverage newly available tools. And if you’re the infrastructure provider, i.e., the sponsor bank, you’re inundated with much more demand from FinTech programs who would like to work with you. And you are kind of recognizing that a lot of the requests from the most attractive FinTech partners stretch your capabilities. You have access to a whole range of FinTech picks and shovels of the types of companies we mentioned before that, Plaids, Socures, Onfidos and dozens and dozens of others.
Source: Proof of Capital
And so, I think it becomes natural to basically create a sort of operating model where there’s increased dependency on some of those ancillary service providers. So maybe it’s KYC, maybe it’s document verification, maybe it’s AML, maybe it’s fraud, maybe it’s a whole host of other things, but you’re in a sense relying on those third parties to feel important, to provide important services as part of this ultimate offering.
And because you’re looking to optimize and bring efficiency to the way that you manage FinTech programs, you’re probably putting good process and structure in place. But certainly, from the kind of technology startup perspective, the obvious solution would be like how can you automate this? How can you put tools in place which do compliance, which do types of governance, which are checking adherence to policies and procedures? And I’m not sure that the technical competency has traditionally been there with a lot of the sponsor banks nor the desire to invest in the technology to meaningfully automate that.
So, I think probably what has happened as a result is that there’s been kind of a decreased level of day-to-day interaction or oversight of increasingly complex FinTech businesses, especially if they start dealing in crypto or other financial assets which are outside of the traditional core competency of the sponsor bank, combined with increased reliance on third parties to provide some of the required services.
And I think as a result of that, you’ve kind of gotten to the situation we’re in today where there is increasing regulatory interest and scrutiny in many of these partnerships. OCC, I think is… I guess they stopped sort of taking action, but there was some sort of agreement between… It’s called Blue Ridge Bank and the OCC around BSA, Bank Secrecy Act requirements. I think specifically related to AML. There was Jason’s recent article as you mentioned about Evolve and potentially some of the challenges that are being reported there.
Source: Public Filings
So, I think it’s no surprise to anyone who follows this space that the aggregate has piqued the interest of regulators. And I think we would all expect increased rules and requirements to be imposed, which certainly for consumers, certainly for the industry at large will only be a positive thing. But again, I think it only further emphasizes the need and opportunity for investment and innovation, I think at the infrastructure level, the sponsor bank level.
The regulatory points are always so tough because I think that very real damage gets done to people’s finances by inexperienced companies and very rarely, but occasionally very large damage is done by people who have maleficent intent and that can be extremely damaging to industries. But part of me just has trouble with the idea that we need to regulate more how you can put a loan into a checkout widget or something.
Not that these types of failures are inevitable, and I don’t want to be glib about it, but it’s not clear to me what will be fixed through regulatory action, especially because these are often technology companies are really marketing companies that would be happy to comply. Generally speaking, and maybe I’m being optimistic or naive, but it doesn’t seem like compliance is the problem or a lack of a rule is the problem. So, I don’t know, how do you think about the question of whether more regulation will really help?
I think that anything through the eyes of a risk person, it’s more about the kind of tangible impact. So, the loan and the checkout flow, sure, consumer protection considerations, but that feels less fundamentally important to me than things like if I think about the goals of regulators, I think one is definitely consumer protection, potentially an even greater goal is protecting the structural integrity of the banking system. And I think that that is relevant in some of these conversations around quality of deposits, dynamics of deposits, sources of deposits. And ultimately, if you have too many BlockFis, there are implications around too many associated banks. What does that start to mean in terms of the broader trust in and credibility of banks more broadly?
Source: Federal Reserve of Chicago FMG
And then I think the other piece which is of fundamental in my view, moral importance, AML, right? The Bank Secrecy Act and KYC, AML requirements exist for a reason. I think it’s an important reason. There’s definitely a faction in the crypto community which would I think disagree with that point, but I think it’s morally quite relevant. And so, the importance of setting reasonable but also responsible AML requirements and then having those respected.
So, to the extent that any of the gaps in oversight and in outcome relate to money laundering, risk of deterioration of structural integrity of the banking system and as you mentioned, consumer harm, I do think regulation is important. Now, I think it’s most relevant around things like checking accounts and money movement and probably less relevant in other products in the financial services suite.
Source: Morning Consult
I wanted to ask you about a point you made earlier around the value chain. Another analogy that I always go to is manufacturing and distribution, and so infrastructure is kind of like the manufacturing of the financial product here and our applications are distributing it. And one of the value propositions of the distributors has been, “Hey, it’s really cheap. Are you tired of paying your bank a monthly fee? Well, get it here for free. Are you tired of paying too much to your payday lender? Get it here for a subscription fee instead. Are you tired of your bank charging you international money transfer fees inside of the FX? Do it here for almost nothing.” And that’s often cheaper, less features, but cheaper.
Source: 11:FS, adapted by Simon Torrance
And to do that, to be cheaper, your value chain has to be cheaper. And we thought that it would be cheaper because you’re not maintaining a bank branch footprint for distribution, you don’t have giant corporate overheads because in a lot of the time these aren’t big public companies, but you’re now kind of layering in so many different providers, Plaid for authentication and data aggregation, and then maybe Sardine for KYC and money transfer in, and then whoever it is for the cash account and then whoever it is for the credit account. And then maybe you’re going to add in some investing and crypto or whatever.
Direct deposit to be facilitated, easy transfer of payroll from your existing bank to your new bank. Or if you want faster ACH settlement times, the list goes on and on of amazing FinTechs providing really, really great services, but the costs add up.
Do you have a sense for what those costs look like? If we were to just create a caricature of it, what would the old costs look like and what would these aggregated new costs look like? What’s even the order of magnitude? How do we think about it?
The simplest answer is like your own personal budgeting, when you actually stop to take a look at all the subscriptions that you’re paying for you, your wife, your kids, quickly approaches your rent. I think it’s probably not dissimilar. I think there’s an irony, right? Because in whatever it was, 2008, 2010, 2012, there’s this idea of like, well, banks have a high-cost base and they don’t have significant competition that it feels like an oligopolistic kind of approach to pricing.
So, if we use technology, we don’t have branches, we don’t have people, we don’t have to charge $25 for an international wire, we can charge $5 and everybody wins. And I think in that Quaint and tech 1.0 world, that was true. I think as you start to have this proliferation of apps that are competing with each other as much as with traditional bank providers, and then you start to optimize for things like payment in and out, processing times and success rates on online transactions that require reapplying fraud analysis, things like data visualization across multiple accounts which have an underlying cost through a provider like Plaid or Yodlee. I think that’s when the cost structure starts to get away from you and it does ultimately mean that giving everything away for free becomes less and less tenable.
Source: Capgemini, Business Insider
If I were to guess, I would guess that you would end up with consolidation in the infrastructure space because what the infrastructure space needs is not just a wave of clients that try to win, pay them a bunch of money and then all die. They need a wave of applications that try them, use them, find product market fit, and most likely 10 of those applications crush everybody else, but the infrastructure is so locked in that it is industry level kind of monopoly or oligopoly as in the case of the core banking systems.
But right now, you’ve got a lot of the horizontal players saying, “Look, build on our APIs,” and a challenge on reaching the customer. So, one way it could go is the bundling of all of this stuff together again into a fat API first embedded finance like Super Bowl. But then is that the right outcome? Or maybe the right outcome is just, it was always hopeless to try and put scotch tape of APIs on traditional banking and it’s all computational blockchains and DeFi, that’s the only way to go. How do you think the industry dynamics will play out?
Yeah, so firstly, I think you nicely verbalized the way I think about this also, which is that there has to be a consolidation optimization on the efficiency side. I think one opportunity is for banks, sponsor banks that want to genuinely commit to this long-term. I think they can right now build a very strong competitive position by bundling, sounds dangerous, but finding ways to either independently develop or acquire the third or fourth leading provider of some of these services to wrap them into the sponsor bank or banking-as-a-service stack.
Source: Dealroom and ABN AMRO Ventures
I think the other opportunity that the ultimate winners in the sponsor bank space have is to rather than think of themselves, the way that many SaaS companies generally do, which is like, “Here’s our self-serve API, put in your credit card number, here’s the documentation, do whatever you want,” to think more strategically about how they can really deliver the success of the FinTech programs that they work with because I think unlike many SaaS offerings where it’s really about Stripe, about having every developer that’s doing anything related to payments, use the platform. And then maybe it ends up being like an 80/20, they’re like a small number of enterprise customers that are driving tons of revenue but that long tail of developers is really important.
Source: CB Insights
I think the sponsor bank model, you’re much more likely to have the handful of massively successful scaled programs generating nearly 100% of revenue and upside for the business. And I think therefore, the more you can do to engender, enable the success of your FinTech partner programs rather than signing up as many as you can handle and letting competitive dynamics naturally play out and hope that you end up with a couple great programs. I think that’s ultimately going to be the way that banks that are serious about this structure their businesses.
Source: Jeff Forkan, Medium
And what about on the application side? What’s a way to fail gracefully? End of the day, I guess this is venture capital’s fault that people are still building these mobile distribution pipes, but do you have any advice for people who are like, “Now is a great time to build a new neobank?”
I suppose my advice will sound relatively obvious and hopefully not overly conservative, but start with revenue generating wedge, which may or may not be a financial product. Think very hard about what parts of the embedded finance, the banking, the neobanking suite you expect to monetize versus which parts you expect to be loss leaders, and ensure that you comprehensively have a very strong revenue positioning.
Source: Saison Capital
But again, crucially, do not start with the goal of building a full service neobank without a very clear high margin, strong revenue generating wedge product that really provides the cash flow to support the time ultimately that it takes to expand your products, that to acquire customers, to build trust. I think there’s, at least based on what we’ve seen in the market to date, that’s the prudent strategy. If your goal is to build a company, raise money, deploy it responsibly, and have something to show for it after decade,
I would add to it that given how far we are into the FinTech journey and the web2.0 journey and potentially we’re at the end of the web2.0 journey and the beginning of the web3.0, one that freemium-as-a-distribution model is from my point of view, essentially dead. You can give people free stuff, but you can’t pay full acquisition costs for free customers. And so, you know, see that now pretty clearly with companies like TransferWise doing very well because they’ve got an engine that makes sense. And then lots of companies that sped thinking they were going to expand margins into their various hallucinations, really, really struggling with the public markets.
I think the other thing I want to add just in closing is that a lot of this discussion including some of our final points were around the neobanking model. And I think probably if you’re thinking about companies to start now, that’s not where you’re really looking, right? Embedded finance in many cases is about thinking about existing workflows, whether it’s business focused, whether it’s consumer focused, but that aren’t looking to replicate what banks do or traditional banking products or where people go for the traditional banking products. But to basically plug in a high value financial product into an existing consumer business workflow.
That to me, I think especially if you start to couple it with developments in the digital economy, a lot of the tools and dynamics that we’re starting to see come out of Web 3.0, that to me, I think starts to get really interesting. So, I think that’s much more my area of interest right now rather than the traditional neobanking model.
It’s been a pleasure to poke around this topic with you. I’ve definitely learned a ton. If our audience wants to learn more and follow along, where should they go?
Rebank.cc. They can find all of our content there. They can check me out on LinkedIn or Twitter, though it’s not my platform of choice.
We’re all on Twitter. We’re all in Elon’s world now. And then of course, check out The FinTech Blueprint @fintechblueprint.com and me at @Lex Sokolin on Twitter and various other social networks. Thank you so much for the conversation today.
Thank you, Lex.