Happy Wednesday, Fintech Takers!

I spent a large chunk of time yesterday watching the publicity strategies of Robinhood and the Federal Housing Finance Agency play themselves out. 

It was quite amusing.

I doubt he cares, but someone should tell Vlad Tenev that his answers to questions about the tokenization of private company stocks aren’t … great. He defended the concept on CNBC by saying, “In and of itself, I don’t think it’s entirely relevant that it’s not technically an equity instrument.”

Not comforting!

And the FHFA? All I can say is that if you aren’t following Bill Pulte on Twitter, you are missing out on … something. The Director of the FHFA spent some time yesterday tweeting updates on initiatives that were already underway to increase access to homeownership (incorporating rent data, adding the VantageScore as an option, etc.) and claiming credit for them on behalf of himself and President Trump. He also encouraged lenders to pull both Vantage and FICO for every application (why?!?) and tweeting my new all-time favorite quote (about credit scores): “It's just math. Predictive math.”

Yes! I love it so much. If you would like a Fintech Takes “It’s just math. Predictive math.” t-shirt, reply to this email and let me know.

Now, let’s dive into the latest edition of the Fintech Takes podcast! 

— Alex 

---

3 BIG IDEAS FROM THE PODCAST

In typical Not Fintech Investment Advice fashion, Simon Taylor and I started with a few interesting companies (that we’re definitely not giving investment advice on) … and let the conversation take us wherever it wanted to go.

In this case, the conversation wanted to go to real talk about billing for the LLM era, mortgages as an employee benefit, and the regulatory endgame for stablecoins that offer yield (even if they’re not supposed to).

Tune in for the full conversation here

And read below for my three big ideas... 

#1: LLMs Are Breaking the Billing ModelCopy anchor linkCopied

The way we bill for software hasn’t kept up with how software is being used.

Modern billing engines were built for SaaS and e-commerce. 

But the economics of LLMs and AI agents look nothing like those models. They’re metered, variable, and highly context-dependent (the billing challenge here starts with how to even measure what you’d be billed for in the first place).

AI-native companies need usage-based billing, token tracking, GitHub access provisioning — and they need it now.

Stripe dominates when it comes to billing infrastructure for SaaS and e-commerce, but even Stripe is starting to get outpaced for certain edge cases (every bleeding-edge infrastructure company eventually falls back a bit as they move up market to enterprise customers).

If AI-native businesses are spending millions on usage before their first ARR dollar hits, they’re going to need new tools. Not just “payments for AI” but actual infrastructure that understands how computing is used, metered, and priced from the ground up.

What’s powerful here is timing. Cloud billing controls and DevOps tools showed up too late. They were a correction to a problem (excessive cloud computing bills). LLM billing startups are arriving during the Cambrian explosion, so their infrastructure might actually shape the market instead of chasing it.

🎬 DIRECTOR'S COMMENTARY

That was the subtext of our discussion on Polar — an “LLM billing engine” that lets AI-native apps meter usage, manage entitlements, and spin up checkout in seconds.

Polar is solving for billing first (as opposed to payments), which is how they’re unlocking growth. It’s the DevOps lesson all over again: companies don’t realize how much they’re overspending (or how poorly they’re tracking usage) until it’s too late. Polar helps LLM startups skip that learning curve.

#2: Mortgages as a Perk (and the Return of Company Towns)?Copy anchor linkCopied

We talk a lot about embedded finance, but what happens when mortgages get embedded?

It’s a throwback with a twist. Company-owned housing isn’t new (think Dearborn or Hershey), but it feels new when it’s delivered through HR software, not factory bosses. And it’s not about tech — it’s about talent retention in an era when housing affordability is the issue.

Multiply Mortgage pitches mortgage-as-a-benefit, turning a broken housing market into a hiring and retention strategy. Companies can’t always outpace rent with raises, but they can route benefits through financial partnerships that remove friction from the housing ladder.

If you need employees in a specific location, a discounted mortgage starts to look like table stakes (especially if that location has a high cost-of-living).

Mortgages are one of the most complex consumer products imaginable (if not the most complex). But they’re also most people’s biggest monthly expense. Multiply’s bet is that streamlining the process and passing savings through employers gives them the edge.

But that edge dulls if interest rates fall (will they ever at this point?). Add to that the bigger macro wildcards — if unemployment rises and competition for talent cools, employers have less reason to offer fringe benefits like this at all.

It’s a smart model, but it’s tightly coupled to an economic moment that could shift at any moment.

#3: You Can’t Stop Stablecoin Yield (But You Can Reroute It)Copy anchor linkCopied

Stablecoin issuers have a clear directive under the Genius Act: no interest, no yield, no APY. The goal is to keep stablecoins from acting too much like banks.

But here’s the problem: users still want yield. That demand doesn’t vanish just because it’s banned at the source — and other companies (that don’t issue the tokens) are figuring out how to deliver it anyway.

What’s emerging is a clever workaround:

Instead of depositing stablecoins and earning interest (which would break the rule), you swap them for a token that represents something like a money market fund — similar to a savings account yielding ~4–5%.

The fund is held by a traditional custodian (like Vanguard or JPMorgan), and the swap happens instantly on-chain. From the user’s perspective, it feels simple: “I moved my dollars and now I earn 4%.”

No laws were broken; it was just a swap. And now your new token represents the same dollars… just bearing yield.

One company playing this game particularly well is OpenTrade which offers “yield-as-a-service” by handling the regulatory gymnastics behind the scenes.

So why does this matter?

Because it shows how crypto and fintech are figuring out how to route around regulations that don’t evolve fast enough. 

And they’re doing it by building new “middle layers” that connect users, apps, and assets in ways regulators didn’t anticipate (which is what happens when we try to regulate this stuff based on an imperfect understanding of how the technology works).

---

WHAT I'M LISTENING TO

#1:  How the Attention Economy Is Devouring Gen Z — and the Rest of Us (The Ezra Klein Show)  🎧 Copy anchor linkCopied

Ezra Klein and Kyla Scanlon. Enough said.

#2:  Trent Sorbe, Chief Payments Officer at First International Bank & Trust (Fintech Business Weekly) 🎧Copy anchor linkCopied

This came out a while ago, but I finally caught up on some podcasts and this one stood out. Two folks who really know what they’re talking about when it comes to BaaS.

---

WHERE I'LL BE

💻 Fintech Open House — July Edition | July 16 | ZoomCopy anchor linkCopied

Dive into fintech’s frontlines at our Fintech Takes Open House (truly open to all). We’ll tackle whatever’s breaking in the moment (the BaaS reboot, digital wallet turf wars, embedded creep, and the ever evolving regulatory maze — or not). Wherever the conversation goes, we’re going with it. Sign up here to join.

---

Thanks for the read! Let me know what you thought by replying back to this email. 

— Alex