Charles Birnbaum

The through-line

Birnbaum approaches fintech investing as a student of history rather than a forecaster. He is explicit that prediction is not the point: "I actually think a lot of what we do in venture is very similar to the study of history. It's like you look at the past to try and understand the present, predict the future. And I don't think predicting the future is actually possible. So I just think understanding what has happened before is the best method" [1]. That instinct produced his framing of fintech as a sequence of waves, usually set off by regulatory change or interest rate shocks, and his argument that you cannot disrupt the industry without understanding its history [1][4]. The same historical sensibility makes him unusually modest about domain mastery. After more than a decade covering financial services he says flatly, "I feel like I'm an expert in none of them," because fixed income asset management, long short equity, mortgage, commercial insurance and life insurance are each deep enough to absorb a career [1].

The second constant is a preference for zooming in over zooming out. He describes the waves piece as an attempt to do something different, then adds that he "more enjoy[s] kind of double clicking and zooming in because I think that's really where more of the interesting nuance happens" [1]. Applied to the present moment, that same discipline makes him cautious about the AI cycle: he expects a major shakeout among current AI companies, with the enduring ones still to be formed [1].

On the five waves of fintech

The first wave was business model innovation, concentrated roughly between 2008 and 2013, when "the incumbents were kind of shattered by the lack of confidence that the public had after the financial crisis" and new brands could be built off a single sharp point solution: SoFi with student lending arbitrage, Robin Hood with free trading, Betterment with democratized access to diversified equity portfolios [1]. Those companies also drove the unbundling of financial services, which in turn created the conditions for rebundling through software and APIs [1]. A third, less glamorous wave is industry-specific software finally bringing whole sectors into the cloud. He is blunt that property and casualty insurance and mortgage "are still not really that far along in their digital transformation," and that while the conversation gravitates to crypto, AI and decentralization, "a lot of these companies are still using manual workflows and pen and paper" [1]. The pressure applied by the first wave is what forced incumbents to move at all, and he stresses that "these are not things that happen fast. I mean, financial services and healthcare are both super slow industries to move to digital world" [1]. The most recent wave is what he calls orchestration layers, and looking forward he points to the intersection of healthcare and fintech, particularly products aimed at the problems of aging populations in developed economies [1]. He has also examined why the embedded fintech wave has progressed more slowly than expected, and what might catalyse the next round of innovation [4].

He holds the taxonomy loosely. The buckets are "probably not that accurate because nothing's that simple or neat," and the first wave, business model innovation, "is still happening" rather than being closed off [1].

On infrastructure, APIs and the pendulum between best-of-breed and platform

The rebundling wave produced companies selling critical infrastructure as APIs, Stripe and Plaid outside the portfolio, Alloy, Lithic and earlier stage companies such as Basis Theory inside it [1]. The structural consequence matters more to him than the individual names: founders can now build on this infrastructure "and not need as much capital as SoFi and Robin Hood and Betterment needed to get going" [1]. Asked where startups can still win as the large API companies expand, he treats it as a recurring dynamic rather than a fintech-specific question: "in any software market, this is kind of that constant pendulum that swings back and forth" [1]. A company spikes on one piece of the puzzle, becomes best of breed, and then either stops there and gets acquired or becomes a platform itself. That is the path of Twilio and Salesforce, companies that started doing one thing and now do a lot more, and he sees several of his own companies on that multi-product journey [1]. Becoming the platform others must work with requires either genuine product development and listening to customers, "not just growing for the sake of growth," or well-chosen acquisitions of products and teams that fit the ecosystem, as with Twilio's purchase of SendGrid [1].

On orchestration layers, and why you have to earn them

The orchestration layer is his term, self-consciously nerdy, for a position he thinks is genuinely valuable: sitting between a customer and the sprawl of data providers and point solutions, delivering access to all of it "in a manageable way that's very specific to your use case," whether that use case is fraud prevention, B2B payments, or the front-end experience of approving a loan [1]. Both startups and incumbents need that layer. But he is emphatic that it is not a starting position. At the peak of market frothiness "we saw a lot of startups that were kind of at the beginning trying to be those orchestration layers and it's really hard. It's really hard to start there" [1]. The right to orchestrate is earned over time, which is why the established API platforms have a better shot at it than a company that begins with the ambition [1].

On distribution as the permanent constraint

Asked for the biggest challenge facing fintech over the next five to ten years, he does not hesitate: "It's always been the same and it's distribution" [1]. The zero interest rate period masked the problem, because abundant venture capital let companies acquire customers aggressively despite long payback periods. SoFi is his example of that strategy working, going from student loan refinance to a full-fledged bank with its name on a football stadium, but he notes it was only possible because they raised a great deal of capital to go direct through digital advertising, direct mail and traditional media [1]. That route is now closed for a period, because raising capital is much harder, and "finding durable distribution channels takes time. It doesn't happen overnight and that's really where the value gets built" [1]. Product advantage alone will not substitute. Occasionally a founder builds something 10x or 100x better, but in financial services the window is short "until the incumbents just have to react," which is why Schwab and Fidelity now offer free trading too. Robin Hood used that hook long enough to build a large, loyal base, and now has to offer much more [1].

On founder market fit and the "why now" question

Because no investor can be an expert in every submarket, the underwriting has to rest elsewhere. For early stage fintech, he says, "it's all about the founder market fit and respecting the why now question like why does this not exist?" [1]. His view is that when a company does not exist in financial services, the reasons are typically big, messy and structural, and an investor has to get comfortable with those before taking a swing [1]. Time in the industry mainly buys respect for the difficulty: "The history gives me a respect for how hard it is, but it doesn't necessarily make me an expert in any of them" [1]. His own focus on financial services came through mentorship under Rob Stavis, who invested in disruptive financial services models before the category was trendy, having run a fixed income arbitrage desk and used technology and software to arbitrage fixed income securities in the 1990s and early 2000s, backing expert networks, the first peer-to-peer lender, and Betterment at the Series A ahead of other firms [1]. That grounding in wealth management is a thread he has continued to trace, discussing the evolution of modern wealth management alongside another practitioner in the field [3].

On what an investor can and cannot do for a founder

His portfolio spans companies of a few people still hunting for product market fit and companies of a few thousand navigating a higher interest rate environment, and he calibrates involvement accordingly. At the earliest stage the honest answer is restraint: "you cannot help them find product market fit," so "you just have to let them cook and listen," weighing in only when a specific question lets you spare a founder something you have watched someone else try, "but you don't tell them what to do" [1]. Once traction arrives, the work shifts to team and structure, which he describes as the hard part, since the people who got the company to the first phase do not all grow into the next one. He interacts with as much of the team as he can so that his feedback on which role and which set of experiences would unlock the next level is grounded in observation rather than theory [1]. He also sees part of his job as translating a moving target founders cannot see: what future investors will need in order to get excited [1].

At the later stage the skill is reading the room on a larger board, being supportive when others are being harsh on the team, and asking the hard questions when nobody else will [1]. Underneath all of it is a fixed conviction about the limits of the role: "we don't run these companies. I mean, I think as a minority investor, you have to have the mentality that you backed a set of entrepreneurs to do it and you have to coach them and support them as best you can, but you can't run the company" [1][2]. He suspects never having founded a company makes that easier for him than for the former CEOs turned investors he has watched struggle to let go [1].

On experience, and what tenure does not confer

He is consistent in refusing to treat seniority as an edge on market conditions. He tells the junior people who join the firm, some straight out of college and some as associates with no investing experience, that "they don't necessarily have a much weaker perspective on the market conditions than I do cuz it's always changing" [1]. The relevant data arrives collectively, every Monday, in discussions of live deals and what the team is seeing in the market, and that picture is simply different from the one that existed when he joined [1]. He is equally candid that the shape of his own path was unplanned: he knew twenty years ago he wanted to invest but had no route to it, never recruited at venture firms in business school, and describes the eventual move as "pure luck in the end" [1]. Even his portfolio management approach carries the same disclaimer: "I'm still figuring that out" [1].

Takeaways

• Fintech moves in waves triggered by regulatory change or interest rate shocks, and the honest use of that history is to understand the present, not to predict the future [1][4].
• Distribution has always been the binding constraint in fintech, and the venture-funded direct acquisition playbook that built SoFi is off the table while capital is expensive [1].
• Product advantage in financial services is temporary; incumbents like Schwab and Fidelity eventually copy the hook, so the base you build during the window is the real asset [1].
• Orchestration layers are a valuable position but a terrible starting point, and companies have to earn their way there over time [1].
• API infrastructure from companies such as Stripe, Plaid, Alloy, Lithic and Basis Theory means new fintechs need far less capital to launch than the first-wave neobanks did [1].
• Property and casualty insurance and mortgage remain largely pre-cloud, running manual workflows and paper, which is a slow but real opportunity [1].
• At the seed stage an investor cannot manufacture product market fit; the job is to listen, occasionally save a founder wasted time, and never dictate [1].
• Expect a significant shakeout among current AI companies, with the enduring winners still unformed [1].