Pod Insights a16z Podcast The Open Source Movement in Fintech About this episode Imagine that financial services could be built like Lego blocks — that developers could easily combine different pieces to discover hundreds of new use cases. That's the transformative potential of the open source movement on banking, according to a16z fintech general partner Angela Strange. In this video and accompanying op-ed (https://future.a16z.com/open-source-financial-services/), she unpacks how open source has the potential to modernize payment infrastructure, expand connectivity beyond traditional banks, combat financial crime, and increase accessibility for three billion underbanked people worldwide.
Transcript 0:15 Speaker 0 I'm excited to be here today at the first move FinTech developer conference, thanks to Wade, Bob, and the Move team for inviting me to speak. And it's fitting that it's a developer conference, because I believe that developers are gonna lead what is the biggest transformation in financial services that we've seen to date. Now, I used to make this joke to room fulls of bankers maybe five years ago and say, hey, how many of you work at a bank? Predictably, people would put up their hands. And I'd say, well, if you did the same job but you worked in a loft, or maybe you wore a hoodie, then you'd be a fintech company. And then generally people would chuckle and they think about these small fintech companies off to the side.
1:02 Speaker 0 But now, in a 25 trillion dollar industry that has long been dominated by suits and lawyers, the power is moving to developers. Why? Part of the reason is that open source is finally coming to financial services in a big way. And I'm gonna talk today about the power of open source, specifically in financial services, the companies and the projects that are leading the way, what that means for us and the broader industry. But first to understand what's going on, it's worth a quick walk through history, focusing specifically on how banks are built. The answer, hundreds of years ago, is well, marble. Hundreds of thousand ton slabs of it. In fact, even the ledgers in the really early days were marble.
1:48 Speaker 0 The ledgers moved to paper, the marble buildings moved to brick and other materials, but banks for a very long time were concerned with keeping physical money safe. And it actually wasn't until the 1960s with the advent of the mainframe, COBOL, that software came in. Eventually banks could interconnect, they could do interbank transfers, we could wire money across the country. And now banks for the first time had to be concerned with keeping virtual money safe. Then came the desktop computer, we didn't even have to go to the branch, we could just access our banking account from our laptops in the comfort of our home. And eventually we could do a lot of our banking via the mobile phone.
2:33 Speaker 0 So this brings us to where we are today, call it the third era of banking, where the modern cloud has allowed many of these core systems to move up into the cloud. And so now, as we'll talk about a little bit later, you can get many of those layers as a service, very similar to what Amazon did for compute and storage. Deposit accounts as a service, credit card issuing, some compliance. And really what this did is it took banking away from something that only a bank could do to something that any company can do. Lyft can add bank accounts for its drivers, Procore can do lending for its construction workers. And so now money has gone from physical at a bank, to virtual at a bank, to virtual to pretty much any company.
3:20 Speaker 0 So that's a lot of progress. And many would say that we're pretty much done for the near future. And this is where I would disagree in what we're gonna talk about today. There's demand from users and there's demand from developers to go even further. So what might that look like? What if we exploded those layers even further into the most basic primitives? So what is a primitive, not derived from anything else? So in financial services, that could look like a ledger, or it might look like a payment library that's built right down to the network layer. And then what if those primitives were open source? Here's what we'd get. Essentially, Lego, but for banking. They'd be composable, you could combine any combination of any piece that you wanted to into any use case imaginable.
4:11 Speaker 0 And then they'd be open source, which means that not only Lego the company could build the building blocks, but the smartest people anywhere in the world could contribute. So before we dive into what this looks like for financial services, let's just take a use case from everyday life that can help illustrate the power. Imagine 100 years ago you're building a home. You don't really have that much choice. Would you like the two-bedroom home or the three-bedroom home, and what color would you like your roof? Over time, you could customize more and more. How many bathrooms do you want? What's the number of floors? So eventually you can now source your windows from Italy, you can get your toilets from Japan.
4:57 Speaker 0 And the three major impacts are roughly analogous to some of the power of open source. One, you can leverage contributions from the smartest people in the world. Two, which results in best in class parts and pieces, and then ultimately, way more flexibility and choice, so you can put together crazy combinations that you wouldn't have been able to even think of before. Now, before jumping into financial services, let's take an example from the broader enterprise software landscape. Git, popularized by GitHub. Now, version control existed prior to Git, but as many of you know, it was fairly clunky software, or it could even look like, hey, are you working on this piece of software? Calling down the hall, just to make sure that you don't override someone.
5:47 Speaker 0 Along came Git, and here's what happened. It changed entirely how software teams worked. There's now global collaboration. You don't have to be in an office, you can be anywhere in the world. What's results in best in class libraries? And then most importantly, you're not rebuilding stuff that people have built all over the place. You're able to grab the status quo and then focus on building the new.
7:06 Speaker 0 And to be clear, there is a long runway ahead of companies that still need to add existing financial services, and I'm very excited about this phase and plan to continue to invest here. But let's start with taking a quick look at what this phase has enabled and some of the drivers that are gonna push this further.
7:31 Speaker 0 So let's say, for example, I wanted to build a neo bank for Denver mountain bikers. Here's what just a few short years ago, this process might have looked like. This is a crude banking stack. I need to get a license, like sure I could apply for one of those, but that would take a very long time, as anyone who has gone to this process would attest to. So I'm probably gonna partner with a sponsor bank. Then I need a core system, a ledger to keep track of my money. I need to get money in and out. I might want to issue a mountain biking card. Maybe I want to issue loans, I need connections in with the credit bureaus. There was a lot of regulatory frameworks around this because we're dealing with money, interactions with various different agencies.
8:20 Speaker 0 And then of course, anything to do with money has lots of fraud. So now I have combined all of this stack, mostly with legacy players. It's probably taken me a year and a half to two years, and I still haven't built the new creative thing that doesn't exist in the market that I set out to do. But that's changed, thanks to many of you in this room. I can now get all of this as a service.
8:52 Speaker 0 Lending as a service, there's also great infrastructure players for that. But now what I'd say is while we've made a ton of progress, and we've gotten better and better at launching existing financial services. Users and developers are demanding more forward progress. So let's take a look first at the user side of the equation. We talked about how banking has moved from something that banks do to something that any company can do. Consumer companies, enterprise companies, so it stands to reason that because those companies have gotten much more creative in terms of the use cases that they're trying to solve, the financial services that they're trying to add needs more customizability.
9:40 Speaker 0 Geography.
9:53 Speaker 0 But I would invest in a bank for each of those reasons, regions, and it's because it's got different regulatory structure, different Infrastructure, different consumer needs. But now here's the problem. Enterprise companies, take Shopify, which are global, are trying to add financial services. For example, merchant accounts. It's still very difficult to add global financial infrastructure. And then finally, we have billions of people that are left out of the financial system entirely. And there's a growing set of highly talented entrepreneurs, often living in these communities that know the products that need to be built. What if they could just grab open source libraries and get started?
10:40 Speaker 0 I think that could dramatically speed progress. And so consumers are demanding evolution. Let's look at developers. 40% of bank's code is in COBOL. It's taught at no top schools, there's a shortage of COBOL developers, any COBOL developers here, you have great job security. But the thousands of fintechs built are obviously built on more modern infrastructure. But it's nearly impossible not to interact with legacy tech. So let's say you're building this once and it's very tedious and frustrating, fine. But what's happening now is there's more new companies, there's more developers changing jobs. And so, as my partner David Yulovich likes to say, any time developers find themselves building something over and over again, they're very smart.
11:31 Speaker 0 They'll find a way to automate it. Better yet, they'll find a way to open source it. So not only will they not have to build it again, the community can help make it better. So now there's thousands of developers looking for off-the-shelf solutions to repeatedly frustrating problems, and potentially also thousands that are willing to start the next open source projects. So I'd say that while this third as-a-service era will continue to solve many problems. There's both demand and supply to continue to evolve further. And that's why I believe that we're on the cusp of this fourth era, which is going to be driven in large part by open source primitives. So let's dive into what some of those benefits could be.
12:19 Speaker 0 Standards and reliability. There's lots of standards and payments. Many of them are pretty tedious to build to. Great example is the 497-page PDF nature spec for ACH. Or sometimes they're applied slightly differently, as anyone who's written to the ISO 8583 protocol would tell, the messaging protocol for interchange. Reliability. Payments has thousands of edge cases. Very difficult for small teams of even the best developers to fix. Open connectivity. We talk a lot about open banking. But this is now thousands of banks around the world who are building fairly similar things, potentially a use case for open source. Tapping global networks. There's many problems that have seemingly been intractable.
13:07 Speaker 0 Any money laundering for one. What if we could pull from the best minds all around the world to collectively help solve this problem? And finally, access. Now, open source is not about free software. But by the industry collaborating together to build infrastructure that makes more sense for the industry to own, who might stand to benefit the most, it's the people that are building for and the people in the bottom third of the socioeconomic pyramid. So the bottom line is open source, when applied to fintech could have tremendous benefit. So now let's look at some examples. Let's start with move. We're gonna bring back our banking stack. Now, as we discussed, we can get each layer as a service.
13:54 Speaker 0 And so move does two primary things. First, it explodes some of those layers into their most basic primitives and right down to the network layer. So let's take payments, libraries for ACH, for RTP, for wires, and then two, it open sources them. And so what does that do? Let's take standards. Standards exist, but they're old and they're tedious, as we discussed. ACH was created in 1970, Metro 2 defined to the credit bureaus in 1997. If you want to create receipts for 1099, that was in 1910, it was created in 1918. So with move, not only do you not have to rebuild the standard, you now have a modern reference point. Next, reliability. ACH, very widely used, very prone to failure.
14:41 Speaker 0 So who would be better to fix all of these edge cases? A small team or a large community. And that's with Moves Done with many of you in this room, including payments engineers at some of the largest banks that have run trillions of dollars through the move libraries. And these are mortgage payments, auto loans, all sorts of other payment types that were previously only accessible, COBOL on mainframes, that helped discover 1600 edge cases that the community was able to write test cases against and fix and harden and benefit the library for all. Next, let's look at connectivity. Often, the first step to building a banking app is to help your consumers get data out of their existing financial services institution.
15:30 Speaker 0 So if you're building a budgeting app, you'd want to get all of your transactions, or even if you're a mortgage originator, you might want to let people get the last two months of their income statements. And this has led to the rise of aggregator companies. Plaid in the US, Belvo and Latam, TrueLair in Europe, and many others. But unlike when Plaid started in 2012, in a very closed US ecosystem, there's now open banking regulation that is starting all around the world. So you have to ask: does it make sense to have more proprietary companies all around the world, or is this perhaps an application for open source? So let's look at Europe with PST2. Banks are required to provide access to their transaction data.
16:18 Speaker 0 So now you have banks all across the continent basically building the same thing. Enter to SOBI, which helps run the open banking project. And it does three things. One, it provides libraries to help with integrations into legacy systems. It provides a starting point for the APIs that everybody's gonna have to use. And then perhaps most usefully, it keeps up with the ever-evolving regulation. So those libraries are always up to date, and that's one fewer things that the banks need to do. And so imagine how much more open and connected our ecosystem could be if banks were all using the same open source framework. And it was similarly, there's open source projects that help maintain those connections.
17:06 Speaker 0 Now, I'm glossing over a very complex set of regulatory frameworks that make this difficult, but we could take an inspiration from crypto. There's 60 plus crypto exchanges. And so is there a proprietary company that's sitting on top of all of these? Of course, not crypto by nature, more default open source. There's library assesses exchange, which are a consistent interface for providing trading and accessing market data. Default, open source. So now let's look at the power of tapping global networks. Any money laundering, term that was coined in the 1920s and 30s because mobsters used to mask their money through laundromats. Interestingly, it actually wasn't until 1986 that the first law was passed making this illegal.
17:56 Speaker 0 Since then, there's roughly two trillion dollars a year in money that's laundered. 30 billion dollars is spent by banks and financial institutions in North America to try and prevent this crime. These software companies result in hundreds of thousands of alerts, 95% are false positives. Just imagine the compliance people that are required to process these. And then even then, you probably feel like you read about a different fine every other week, which is true. In just one year in 2019, there were 20, or there were 10 billion dollars in fines. And so after all of this, only 3% of the two trillion is actually caught. Because here's what's going on. Each financial institution individually is monitoring hundreds of thousands of watch lists of suspicious people, politically exposed people.
18:50 Speaker 0 And then they're trying to match entities. Think how difficult it is to do John Smith, the money launderer versus John Smith, the average citizen. You're continuously evolving different rule sets to be able to cap suspicious transaction, and then you're keeping up with ever evolving regulation. And then what makes this problem worse is that if one financial institution gets much better, it doesn't actually solve the problem. It just transfers the problem. If I'm a money launderer, why am I gonna go to the guy that's best in class? I'm gonna go over to the weakest link. And so here's where open source could come into play, and it's already started. Watchmen and open sanctions help provide search across many of these different sanctions lists and how to ingest data in a consistent way.
19:43 Speaker 0 And then there's an early project at Red Hat, which is helping share entity matching rule sets, how to deal with large volumes of data. So you could imagine a world whereas everyone was able to use. The same consistent open source software with the best in class to date, and then if somebody figured out how to make it better, they could also contribute back. And we might finally be able to make a dent in what is very much felt like an intractable problem.
20:21 Speaker 0 Lastly, let's talk about inclusion. There's roughly three billion people worldwide that are left out of the financial system. And there's a few reasons for that. One of them is that many banks are actually pretty happy from a profit perspective. If you look at banks in Latin America, they actually have the highest return on equities in the entire world. But there's also a software infrastructure reason. If it costs me the bank 10, 20, 30 dollars in software cost just to have an account open, it just financially doesn't make sense unless I'm gonna keep a lot of money in that account, I'm gonna take out a mortgage, all sorts of different loans. And so this is where the power of the open source community contributing to a shared resource, in this case a shared core, could really stand to benefit.
21:12 Speaker 0 Mifos is an example here, which is now under the Apache license. And it's enabled its partners, which are micro lending institutions, to spin up peer-to-peer lending. Or for instance, enabling tiny micro merchants to accept QR code payments. And many different applications that are built for the bottom end of the socioeconomic period pyramid, which has previously been left out. And this brings us nicely summarized to this quote I love by Paul Ramsey, who's an open source advocate, which is you get what you pay for, everyone gets what you pay for, and you get what everyone pays for.
21:54 Speaker 0 And so now we've seen the power of open source to drive standards and increase reliability, like we've seen with Move, to help with connectivity, like we've seen with the open banking project, to tap global networks, like the Watchmen and Open Sanctions list, and importantly to increase access, like we've seen with Mifos. And having seen the accelerating pace of innovation over the last several years, from the Marble Bank to mainframe to as a service to now open source primitives, it's impossible not to believe that this is gonna lead to the next era of billion dollar companies. And so what does this mean for all of us in this room?
22:40 Speaker 0 I think it changes how we think about and how we build infrastructure. To start, you could start to look for an open source library. Even just a few years ago, that would have been not even worth doing. Second, if you're using one, consider contributing back to the holistic good of the industry. But more likely, since we're in the earliest of innings, there isn't a solution to the problem that you're trying to solve. And so if you find yourself building a tool or a boring piece of infrastructure, think about how many other people might be doing the same thing or who might want to use the same thing. And maybe your boring tool is the next infrastructure company.
23:27 Speaker 0 And now I called this talk from marble to code. But I could have called it that every financial services company will spin out a FinTech company.
23:42 Speaker 0 And the implications here are actually even broader. And so developers are obviously the creators and the sellers of these primitives. But perhaps an even more transformative shift is that developers are now the buyers. The buying power in many cases has moved from the C Suite who needed to approve 100,000 budgets, to PM and developer teams who can get started for free, or extremely low cost. And so now we have developers selling to developers. Which brings us back full circle to where we started, where I believe that we're on the cusp of a massive change in financial services, which is gonna be led by some of the best developers in the world.
24:31 Speaker 0 It's 2021, and we are at the first FinTech developer focused conference, to which I say, it's about time. Thank you all for having me, and I look forward to meeting you over the next few days.
Previous Cloud Wars and Company Wars: Play Nice But Win Next Crypto Security and the New Web3 Mindsets for Users Page updated August 25, 2026 · transcript and summary are machine-generated
