The $100T API opportunity

I keep coming back to a moment I had with a payments-engineer friend in 2014. He showed me a Stripe integration he had just shipped: seven lines of Python, a test card, a working charge. The thing that struck me wasn't the elegance. It was the realization that he'd never had to learn what Visa actually was. The whole card-network stack had been hidden behind an API that just worked. Twelve years later I look at my own corner of the world and the contrast is almost embarrassing. The people building modern fintechs are still being asked to learn what FIX is.

Stripe didn't invent payments. They turned a thing that already existed into something a developer could call from a notebook. The ten years since have proven that turning legacy infrastructure into clean APIs is the largest software opportunity of the decade. Capital markets is next, and I think it's bigger.

The numbers

Global equities trade roughly $115T notional per year. Fixed income, another $25T. FX, $7.5T daily. Commodities, derivatives, repo, securities lending. Together, the global capital stack clears a multiple of that.

What fraction of that touches a developer-friendly API. Less than two percent, by my math. The rest moves through Bloomberg terminals, prime broker portals, FIX gateways with documentation written in 1992, and PDF-based onboarding flows that take six months.

What's still on PDFs

• Onboarding to a prime broker. 90 to 180 days.
• Adding a market. A sales call, a custom contract, often a wet signature.
• Tax documents. Emailed, paper-mailed, sometimes faxed.
• Corporate actions notifications. Per-broker, schema-incompatible, easy to miss.
• Settlement instructions. Typed by humans into proprietary terminals.

Each of these is a developer experience that a fintech founder in 2026 wouldn't accept from any other infrastructure category. They accept it because they have no other choice. That's the gap.

What the next five years look like

The on-chain capital markets layer has been building piece by piece. Stables in 2014. ETFs in 2022. T-bills in 2023. Equities in 2024. Private credit in 2025. Eventually I think ten percent of global flow runs on this layer, not under two.

What changes when 10% of $100T runs through APIs. Three things.

• Speed of fintech iteration. Today building a neobroker takes a year of broker integration. Tomorrow it's a weekend.
• Capital efficiency. Idle margin doubled across asset classes is the unlock. Collateral that today only works in one asset class becomes globally portable.
• Composability. The same kind of programmability that DeFi unlocked for stables comes to the rest of capital markets. Money markets and perps on tokenized equities, ETFs, bonds, credit.

Capital markets is what payments was in 2010. Mostly working, mostly invisible, and mostly inaccessible to the developer who wants to build on top.

Flo is the SDK layer. We don't replace the broker-dealer; we abstract over them. We don't replace the exchange; we make programmatic access to it look like a single SDK call. We don't replace the BVI issuer / Cayman SPC legal structure; we wrap it in a contract partners don't have to think about.

The bet is simple. The next decade of fintech gets built by developers who never wanted to learn what FIX was, on top of an infrastructure layer that makes capital markets feel like Stripe felt in 2014.