1. Two economic engines
2. The difference is real
3. DeFi as a velocity machine
4. Which is better?
5. An open research question

---

When you are a bank, you can have a low reserve balance. You don't have to keep very much cash actually sitting in a vault somewhere. And so you can do a lot more productive work with your capital, with your deposits. This is part of that risk management thing that banks have gotten so much better at. And that's good. Ultimately, we kind of want society to operate at the limit of risk, but just below the limit of risk. That way, as many people as possible are doing big, adventurous things — creating stuff and buying houses and having happier families.

But stablecoins are 100% collateralized with the lowest risk assets. And so we just get less productive use out of the money.

This is the central tension of stablecoins versus banks. And I don't think it's been explored enough.

Two economic engines

There are fundamentally two ways to get more economic activity out of a dollar.

Capital efficiency (the bank model): Keep a fraction of deposits in reserve and lend out the rest. When you deposit a dollar at a bank, the bank keeps maybe 10 cents and lends the other 90 cents to someone else. That person deposits the 90 cents, and their bank lends out 81 cents. And so on. One dollar supports multiple dollars of economic activity simultaneously. Two people both pretend to have the same dollar at the same time. That's what a low reserve balance means.

Capital velocity (the stablecoin model): Keep 100% of deposits in reserve but move money faster. I create a stablecoin. I can put it into a Morpho pool, and then many people can come and go and use it. A single stablecoin can be sent, spent, lent, or borrowed many times a minute, controlled by humans or software, 24 hours a day, seven days a week. The dollar is never in two places at once, but it moves so fast that it might as well be.

The difference is real

This isn't just a semantic distinction. It maps onto fundamentally different financial architectures.

With a bank, I put a dollar in and the bank can actually give two people both a dollar simultaneously. That's improved reserve efficiency. The dollar is literally double-counted on the bank's balance sheet — once as my deposit, once as someone else's loan. The bank manages the timing risk: it bets that I won't withdraw my dollar at the same moment the borrower needs theirs.

With stablecoins, the dollar is never double-counted. It's 100% backed by Treasuries. But it moves so fast that in the same day, it could be used to fund a trade, collateralize a loan, pay an invoice, and settle a cross-border payment. That's improved velocity.

I probably need a third term because efficiency is really both velocity and reserve amount — how much actually needs to be backing the system. Efficiency might be best expressed as a combination of both.

DeFi as a velocity machine

On the stablecoin side, DeFi protocols create the velocity. Lending protocols like Morpho, Aave, and Compound allow stablecoins to be deposited, borrowed, and returned in rapid succession. This is equivalent to private credit — a financial instrument that's been around for a decade but has been supercharged by the composability and programmability of blockchains.

Where a bank loan might lock up capital for 15 to 30 years (a mortgage), a DeFi loan can lock up capital for minutes. The money comes back faster, gets redeployed faster, and supports more economic activity per unit of time. You can take out a loan that lasts minutes instead of years.

But this is serial reuse, not parallel reuse. At any given instant, there's only one claim on the dollar. The throughput comes from speed, not from leverage.

And speed introduces its own risks. As Marco Macchiavelli — a former Fed economist now at the Bank Policy Institute — pointed out, stablecoins create a "millisecond liability versus minutes of asset liquidity" mismatch. Someone can redeem a stablecoin instantly, but liquidating the underlying Treasury to fulfill that redemption takes minutes or hours. That's an eternity in a crisis. And unlike banks, stablecoin issuers don't have access to intraday credit from the Fed. So someone has to extend that credit — and the question of who, and at what cost, is unanswered.

Which is better?

The honest answer is: we don't know yet. And potentially, you can recreate some of the capabilities lost in capital efficiency with capital velocity. But I don't think this has been explored enough.

What's increasingly clear is that the two models are trending toward each other. Banks have a legal monopoly on improved monetization of collateral — fractional reserves, lending, the money multiplier. But stablecoins are getting faster, and regulators will eventually widen the aperture of what stablecoin collateral can include. The question becomes: how fast can banks improve their velocity (through better technology and faster settlement), and how quickly will stablecoins be allowed to improve their capital efficiency (through broader collateral and eventually fractional reserves)? The answer to both questions determines when — not if — the two systems converge.

Here's what we do know:

In favor of velocity: Stablecoins are faster, cheaper, global, and programmable. The speed of capital redeployment on a blockchain is orders of magnitude faster than in traditional banking. And velocity-based systems are arguably safer — there's no fractional reserve risk, no bank run in the traditional sense, because every dollar is fully backed.

In favor of efficiency: Fractional reserves have powered centuries of economic growth. The ability to multiply the money supply through lending is how communities fund infrastructure, housing, and businesses. A world where every dollar must be 100% backed by Treasuries is a world with structurally less credit available.

The synthesis: Ultimately, we probably want best of both worlds. Deposit-backed stablecoins could preserve the bank's lending function while offering blockchain-speed transfers. DeFi lending could add velocity on top of efficiency. But designing a system that captures the benefits of both without the risks of either is a hard, unsolved problem.

An open research question

This is the kind of thing that I think deserves more rigorous thinking. Under what conditions does velocity-based capital utilization match or exceed the economic output of fractional reserve banking? Is there a crossover point where speed compensates for the lack of leverage?

Consider a simple model: a bank with a 10% reserve requirement effectively turns $1 into $10 of economic activity through the money multiplier. A stablecoin would need to turn over 10 times in the same period to match that. Is that plausible? On a blockchain that settles in seconds, maybe. But not all of those transactions create the kind of productive credit that a mortgage or a business loan does.

The question isn't just whether stablecoins can move fast enough. It's whether velocity-driven economic activity is qualitatively the same as leverage-driven economic activity. A dollar that funds a 30-year mortgage creates a different kind of value than a dollar that cycles through a lending protocol ten times in a day.

This is genuinely one of the most important open questions in stablecoin economics. And I haven't seen anyone try to answer it rigorously.

---

Quotes

Reserve efficiency and why it's good

Source: Sam Broner / Common Prefix — Jan 20, 2026

When you are a bank, you can have a low reserve balance so you don't have to keep very much cash actually sitting in a vault somewhere. And so you can do a lot more productive work with your capital, with your deposits. This is part of that risk management thing that banks have gotten so much better at. And that's good. Ultimately, we kind of want society to operate at the limit of risk, but just below the limit of risk. That way, as many people as possible are doing big, adventurous things — creating stuff and buying houses and having happier families or whatever people do. But stablecoins are 100% collateralized with the lowest risk assets, and so we just get less productive use out of the money.

Two people pretend to have the same dollar

Source: Sam Broner / Common Prefix — Jan 20, 2026

With a bank, I put a dollar in and I can actually give two people both a dollar simultaneously. That's what a low reserve balance means.

Velocity as the stablecoin counterpart

Source: Sam Broner / Common Prefix — Jan 20, 2026

I would call that improved velocity as opposed to improved efficiency. I probably need a third term because efficiency is like velocity and reserve amount — how much actually needs to be backing it. And efficiency might be best expressed as a combination of both.

Morpho pools and private credit

Source: Sam Broner / Common Prefix — Jan 20, 2026

I create a stablecoin. I can put it into a Morpho pool, and then many people can come and go and use it. This would be equivalent to private credit, sort of a new financial instrument. Whereas with a bank, I put a dollar in and I can actually give two people both a dollar simultaneously. That's what a low reserve balance means. And so this is exactly the point — with stablecoins, you can have this very good velocity, but a bank can do something interesting where two people both pretend to have the same dollar at the same time. That's improved reserve efficiency. Ultimately, we probably want best of both worlds.

Speed compensating for reserve requirements

Source: Sam Broner / Common Prefix — Jan 20, 2026

You can send stablecoins faster. You can take out a loan that lasts minutes instead of years. And so, potentially, you can recreate some of the capabilities lost in capital efficiency with capital velocity. But I don't think this has been explored enough, actually — the kind of thing that I think you guys would be good at thinking through and trying to come up with a crisper presentation.

The money multiplier tension

Source: 30 min with Sam (Marco Macchiavelli) — Feb 5, 2026

Marco (former Fed economist, Bank Policy Institute): The problem is externalities. Because if I pay you what I don't have, either you block the entire payment system and you have daisy chains of failures, or you have to do what the Fed does — intraday credit, potentially overnight overdrafts. Now the problem with stablecoins is that at the moment you don't have it. Even with a skinny master account you don't have it.

Sam: I would say —

Marco: So then either stablecoins preposition and they cannot earn interest on their reserves, which means that you are back to 1990s where you have a reserve requirement and you don't earn anything on your reserve, so you're trying to keep your reserves as little as possible. Which means that if you have a big enough outflow, you're screwed.

Banks and stablecoins trending toward each other

Source: 30 min with Sam (Marco Macchiavelli) — Feb 5, 2026

The banks have a legal monopoly on [improved monetization of collateral]. And so the two are actually trending towards each other, where the question is — how fast can the banks improve their capital efficiency, and then how quickly will the stablecoins be allowed to improve their capital efficiency? And that might be how I would compare the two.