Fintech After The Crash
In February 2022, Dave, a US challenger bank with $380m of assets, was trading at $364/share, giving it a valuation of $5.2bn. This valuation exceeded Banco Sabadell, Spain’s fourth largest bank, which at the time had $287bn in assets, 12 million customers and a market cap of $4.6bn.
Today, Dave’s valuation is $95m.
It’s always easy in retrospect to diagnose the excesses of speculation. We laugh now about people selling million-dollar cartoon jpegs and about an $88-billion currency based on a joke about a dog. This was obviously a form of mass technological hysteria that was never going to end well. However, what most people don’t realise is that, while Crypto represented the Wild West periphery of financial services innovation, the core was just as badly hit. Dave is not an isolated case. As we show below, the fintech crash was worse than the dot-com bust.
Valuations of the technology sector as a whole have been negatively impacted by rising interest rates. As the cost of capital rises, the value of future cashflows are marked down. This disproportionately affects early-stage and not-yet-profitable companies since so much of their worth is reflected in their terminal value. But fintech startups were particularly badly hit as their products are about the movement of money itself, so changes to capital flows have an amplified effect which ripples through the sector.
In the fintech space, things are never going to return to where they were in 2020 and 2021. Easy money is no longer going to be thrown at non-viable companies and these companies will cease to exist. Many startups and VC’s who didn’t come from technical backgrounds mistook ‘tech’ investment for technology, expecting technology itself to disrupt the market. Instead, they missed what the internet really offers: which is distribution through completely new forms of sales and marketing, which is critical for an industry with very high customer acquisition costs (and still massive underprovision). Financial services are essential, but boring.
The good news, however, is that fintech is still a massively viable and flourishing sector for innovation: it’s the world’s most profitable vertical market with the most legacy infrastructure and digital services, still riddled with inefficiency and ripe for disruption.
Going forward, we see three areas where the rebound and differentiation will happen:
1. The role of embedded finance in leveraging the true innovation advantage that internet era technology offers: distribution.
2. The fundamental value of AI in a business whose core value add is centred on measuring risk.
3. The increased role of the smarter incumbents.
The incumbents were not disrupted and often have an unfair advantage, playing in a regulated sector which is not a free market. The future will undoubtedly consist of a hybrid of the best of the incumbents and the new entrants, and the incumbents’ role will also be important in shaping investment and M&A. Secondly, the evolution of embedded finance will continue to be the one that solves for the internet era distribution opportunity and lastly, the paradigm shift in technology (from deterministic to probabilistic software) that recent innovation in AI has created, opens up new avenues for real tech investment within the financial services sector.
We are starting to see the green shoots of recovery. Publicly traded fintech stocks are off their lows. And while private investment has yet to rebound, sentiment is palpably better. Nonetheless, we shouldn’t expect a repeat of the same animal spirits. The next phase of fintech will be different with more focus on infrastructure (and AI), solving bigger challenges, with better unit economics, a stronger focus on compliance, and with less hype.
The fintech prize is massive
When Jamie Dimon and others were proclaiming that Silicon Valley was about to eat banks for lunch, the sense of alarm was real. New business models based on internet technologies had fundamentally upended the media industry, among others, and VC money was pouring into fintech, attracted by the possibility of doing the same to finance.
What is more, the prize is so much bigger in fintech. Finance has the largest profit pool of any industry. Collectively, financial services players across insurance, banking and capital markets generated $6.5trillion in profits in 2021, over 30% more than the healthcare industry, and more than twice as much as in ecommerce.
That is a large prize to shoot for, with technology change providing the opening, and venture capital providing the fuel — and so new fintech company creation accelerated, along with valuations, until we got into bubble territory.
The bubble burst
The fintech downturn was more severe than many realise. Like the wider bull market, the correction in fintech was precipitated by rising interest rates. However, unlike the market in general, fintech didn’t correct; it crashed.
The decline in fintech stock values eclipsed the fall in the Nasdaq after the dot-com crash. It was only marginally less severe than the crypto crash of 2021/22. At the nadir, many stocks, such as Root, had negative enterprise values (i.e. their valuation was lower than the net cash they held on the balance sheet)!
The correction happened first in public markets…
…before moving to private markets
Disruption is hard
The fintech correction was brutal because disruption is hard. That is to say, VC investment did not necessarily overshoot per se, but rather there was a realisation that a lot of money was flowing into business models that wouldn’t deliver. When interest rates started to rise, the flaws in these businesses become more apparent (and in the case of VC funding, interest rates made these business models more costly to prop up) — and the bubble burst.
Market commentators have turned widely on challenger banks, but the same flaw in challenger banks is evident in many B2C models. It is expensive to produce financial services products, especially those that need to conform to regulatory compliance requirements. And it is expensive to acquire time-poor and attention-poor consumers, especially when your product lacks virality and network effects. The mistake that most B2C companies made was to try to do both, to manufacture and distribute their own services, thinking that offering a simpler and less expensive alternative to existing financial products would be enough to gain share. However, the reality is that it is very difficult to achieve the same scale advantages in manufacturing that large incumbents enjoy. And it is difficult to obtain the distribution advantages of either large financial firms (who benefit from trust, as the SVB-induced flight to safety emphasised) or from non-financial firms, such as WeChat, whose network effects and product virality enable them to acquire customer cheaply and then embed financial services.
There are some global consumer winners
There have been consumer fintech firms that have broken out, who have achieved the kind of “escape velocity” level of scale where they are now profitable, or have a path to sustained and material profitability.
These were companies that typically had a “killer” first product, with sufficient virality to enable them to acquire a critical mass of customers. Examples include:
- PayPal, which at the start of the e-commerce boom allowed online payments between individuals and companies
- Ant Financial, which launched the first mobile payment app in China
- Revolut, whose FX functionality had even my 75 year-old Dad demoing the app to his friends
- Nubank, which understood that Brazilians were being ripped off by their credit card providers
- (Transfer)Wise, which also saw the gap to offer better rates on foreign exchange transfers.
All of these “winners” were then smart enough to quickly bundle together other products to grow customer lifetime value or to offer their killer product through third party distribution channels to keep CAC low and drive down average costs — or did both, as Wise has done.
But we predict more B2B winners
In general the returns are less asymmetric with B2B fintech. That is to say that the chances of success are higher when operating within the paradigm of existing businesses, helping to make them more efficient or linking them together. But fewer of these players will achieve the same break-out success and valuations of companies like Ant Financial or PayPal.
Within the B2B space, we see three main pockets of opportunity, serving the main actors of an increasingly networked value chain.
Manufacturing and distribution of financial services will continue to separate
When digitized, financial services no longer need to be distributed through physical distribution outlets — or indeed by financial services companies at all. And, in fact, many non-financial companies are in a position to do a better job. They either have access to more data on their customers to suggest the right services at the right time (which, in turn, will improve financial inclusion) or they have much higher engagement, giving them more opportunity in terms of time and customer attention through which to cross-sell other services (including financial services). Or, in many cases they have both: better data and higher engagement. So, we expect financial services to be increasingly distributed through new, non-financial channels over time.
Manufacturing of financial services, on the other hand, has different properties. Where the power law in distribution is governed mostly by convenience, virality and network effects, manufacturing remains mostly about economies of scale. More specifically, the scale economies that come from spreading the high, but largely fixed costs of compliance, IT and operations over high volumes of transactions and large asset values. Therefore, the production of financial services is likely to stay the realm of regulated financial institutions who are likely to get bigger over time given the importance of scale (but within limits imposed by geography, regulators and diseconomies of scale).
So the opportunity in fintech lies within this new industry landscape, namely in making incumbents more efficient, in creating new routes to (SME) customers, and in bridging between different ecosystem players.
Making financial services manufacturing work better
Opportunity continues to abound for fintechs working to make smart incumbent financial services companies work better. By that, we mean they are not looking to challenge the status quo by displacing the existing players, but by upgrading the technology infrastructure to introduce the promise of “fintech” — lower costs, greater convenience and better user experience — within the established order.
Cloud-based systems of record are a case in point. By improving the costs and scalability of performing commodity tasks, they allow incumbents to meet the demands of 24/7 digital finance while freeing up the budget and giving the operational flexibility for financial institutions to adapt their business models, such as becoming banking-as-a-service providers. These modern systems of record also provide the consolidated data infrastructure to massively improve efficiency, but also to allow financial firms to exploit artificial intelligence (see later).
A well-known example would be Mambu for core banking, which has gained significant market share over recent years, but there are plenty of lesser-known and more exciting newer entrants like Tuum (in banking), fundcraft (in fund administration), ALT21 (for hedging) and Metaco (for digital assets).
There is a big role for vertical SaaS in distributing financial services to SMEs
We predict that financial services distribution to consumers will move increasingly to (large) consumer platforms.
Our rationale is two-fold. Firstly, that consumers’ needs are reasonably uniform and can mostly be met with simple products. Second, that the biggest opportunity to increase financial services take-up is offering the right product at the right time.
As a result, products will be increasingly embedded at the point of need. Product insurance will be sold when and through the same channel as the product is purchased, product financing will be sold when and through the same channel the product is purchased, payment will be made when and through the same channel as the product is purchased, etc. Even when financial product purchasing is less moment-in-time, there is still the opportunity for new distribution channels, such as a retailer using loyalty points or cashback to incentivise a consumer to put money into a saving or investment account or a life insurer steering a consumer into a balanced investment portfolio.
Since the consumer needs are likely to surface on the platforms where they spend most of their digital lives, the opportunity is likely to accrue mostly to the largest platforms.
However, in contrast, we think the SME space is likely to be much more fragmented.
The rationale here is that businesses needs are more complex, but also more heterogeneous. What a hairdressing salon needs in terms of, say, insurance or working capital is likely to be very different to a doctor or a lawyer or microbrewer. This lends itself to a more fragmented application landscape, which has in turn given rise to vertical SaaS platforms that provide unified end-to-end applications tailored to the requirements of the segment they serve (like Toast for restaurants or OpenSolar for solar contractors) as opposed to horizontal applications like SAP for ERP, which need integration with other applications and more customisation.
These vertical SaaS platforms become the ideal medium for distributing financial services to SMEs. This is because they have both high engagement (a restaurateur uses Toast every day), and the data to be able to offer and price financial services correctly. In fact, as a16z says, fintech scales vertical SaaS or, put another way, fintech and vertical SaaS converge. What is more, the embedded opportunity for vertical SaaS is still largely untapped. Public Vertical SaaS companies generate more than 60% of their revenues from embedded finance (as opposed to subscription revenues), yet we estimate that more than fewer than 30% of vertical SaaS platforms globally offer any embedded finance services.
And there is a need to bridge between the actors in a more networked ecosystem
As discussed above, we believe that manufacturing will continue to split from distribution of financial services. Distribution will increasingly be through non-financial channels, which enjoy a mixture of brand strength, engagement, and utility. Manufacturing will be dominated by large, scaled incumbents, but outside of regulated domains, there is likely to be greater fragmentation, especially as BaaS providers like LHV UK lower the cost of manufacturing for fintech entrants. In addition, there will continue to be a plethora of companies, in areas such as regtech, market data and security, providing key ancillary services to this ecosystem. In short, therefore, financial services will become more networked, and less vertically integrated, and there is an opportunity in bridging between these different actors.
We see the opportunity space as orchestration. This orchestration could come in the form of establishing connectivity and data standards, as players like Plaid and Tarabut Gateway do for Open Banking or that Calastone does in the fund administration space. But we see an equally big opportunity in helping to bring together disparate financial services for distribution through single distribution channels.
In our consulting work, we have seen how embedded finance projects can get de-railed. The promise is easily understood, but the path to execution is not. There is often a chasm between financial services providers and brands, both experts in their domain, but unable easily to cross the aisle. Furthermore, when a brand wants to offer multiple services, and do so in an intelligent and proactive way, then it needs software logic that sits above the underlying group of suppliers.
And so we see a clear gap for platforms that can do this: mediate between supply and demand, abstracting away complexity, and delivering a seamless and intuitive customer experience, even when multiple financial products are embedded. Theoretically, incumbent financial institutions could perform this role, but to the extent they are looking at embedded finance, they are mostly focused on BaaS only, not orchestration (with Standard Chartered being a notable exception). Instead, we expect orchestration to be provided by specialist platforms, which are already starting to emerge, such as Kayna, Alicia, additiv and Toqio.
Like in any industry, AI is critical, but not in the way many people think
With the advent of ChatGPT, the power of AI and its potential near term disruptive impact became suddenly apparent, an iPhone moment for AI. Since its launch, many people, including us, have been trying to figure out how best to exploit this step change in computing capabilities. Use cases abound, and the foundation models will even write the code for you.
Opportunities lie at every level: in the infrastructure players providing the raw compute (the processors, the cloud service platforms and the models); the providers that enable existing companies to exploit this step change in AI capabilities ( players like Jaid or RegGenome); and, the companies, not AI startups per se, that use AI as previous generations of companies used, say, cloud computing to find fundamentally better ways of delivering services.
Although there will be a Pre-Cambrian level explosion of new startups in sectors outside of financial services where fundamentally new use cases are yet to be discovered, within the finance sector, the focus may be on those that exploit financial services’ core purpose — understanding and managing risk.
Moreover, while large language models accentuate the advantage of certain large companies with large proprietary datasets, they also highlight the challenges that many of these same companies face with utilizing their data assets. Too many large financial institutions have data locked up in on-premise, siloed databases, and are unable to adequately expose this data through APIs, or enrich it with the right metadata. As a result, many of the financial companies should look first to changing the core systems of record before embarking on ambitious AI projects. And in the same vein, there is an opportunity for many of these systems of record to step up to fill this gap: by becoming systems of prediction not just ledgers, like a core banking system with predictive risk management.
The future is exciting, but probably a bit less dramatic
There are many reasons to continue to be bullish about fintech. Valuations have reset. The hot money has gone. Many of the bad actors have been flushed out. Business models have been tested through a downcycle. A goodwill approach to regulatory compliance has given way to a much more robust regime.
This is now a time to build and scale the infrastructure for digital banking. It will be transformational, but different and more measured. And recognising the extent of the crash and the business models that didn’t work is an important first step because the fintech landscape has changed irreversibly.
This blog was based on a talk I gave at an Aperture Investor Network meet-up in Geneva, Switzerland. At Aperture, we invest in companies in which we are operationally active, based on a privileged, first-hand validation of team and business model, companies such as Metaco (which exited to Ripple earlier this year). And we also syndicate these investment opportunities to our network of co-investors. If you would like to join our investor network and attend future events, you can register here.