The Stablecoin Trilemma

Basis, a cryptocurrency project trying to build a token with a stable exchange rate against the US dollar, is shutting down. Basis was one of the smarter projects attempting to build a post-dollar currency exchangeable 1:1 for dollars, so if they couldn’t pull it off, it must be

Basis, a cryptocurrency project trying to build a token with a stable exchange rate against the US dollar, is shutting down. Basis was one of the smarter projects attempting to build a post-dollar currency exchangeable 1:1 for dollars, so if they couldn’t pull it off, it must be a hard problem.

There’s a running joke that cryptocurrencies are just relearning the last 100 years of financial history (“so that’s why we limit margin loans. So that’swhy banks get audited.”), but this joke is widely seen as a joke made at the expense of Bitcoin. I don’t really see it that way: if a group of people rediscovered the last hundred years of physics in ten years, we’d expect them to quickly pioneer newer and better methods; to make fun of the crypto community is to assume that existing financial institutions have it all figured out.

And if you think that’s true, you need a stock chart that goes back longer than ten years.

So, Basis learned that pegging a small currency to a big one is hard. In macroeconomics, people sometimes talk about the “Impossible Trinity,” the fact that, mathematically, you can have any two of: free capital flow, fixed exchange rates, and sovereign monetary policy.

There’s a similar trilemma for Stablecoins:

Fully Backed

The easiest way to issue $100 million worth of crypto currency that’s stably pegged 1:1 with the US dollar is to start with $100m in a bank somewhere. Most of us don’t have that kind of money handy, and most of the people with that kind of money aren’t excited about converting it into an asset that doesn’t bear interest.

So the first thing many Stablecoins do is try to bootstrap from a small fiat currency base to a large monetary base.

We know this is a hard problem, because nation-states try to do this, and fail, even when their central banks are indirectly collateralized by the net present value of the country’s future tax revenues. The 1997 Asian financial crisis, for example, was precipitated by Thailand breaking their currency peg with the dollar. As long as they maintained the peg, their export surplus got recycled into domestic investment, which eventually led to a real estate boom that led to higher leverage and, eventually, a painful unwinding. Maintaining a peg is difficult even with a growing economy to back it.

There are countries that can maintain a dollar peg, and there are two categories of these:

  1. Countries that would benefit more from somebody else’s central banking infrastructure than from trying to manage their own (these countries are, generally, very poor), or
  2. Countries whose peg is massively over-collateralized by their sovereign wealth funds. (These countries are very rich — not just rich enough to credibly promise to exchange dollars for Riyals or Dinars in any amount, but rich enough that if US monetary policy hurts their economy, they can afford offsetting fiscal policies).

A peg that’s not over-collateralized is a martingale bet. It will hold for a while, and then it will fail.

Basis’ protocol was a variant of this: the basic idea, as I understand it, was that they controlled the supply of their currency by either selling more (if it rose above the peg) or by issuing Basis-denominated bonds and using them to buy back the currency. But note that these bonds only have value if the currency rallies, or at least depreciates unexpectedly slowly in the future. This means that selling Basis short is a good trade, as long as you can borrow enough and have more capital than they do: eventually, it will reach a death spiral where the market interest rate for the bonds is so high that currency holders expect extreme dilution, at which point they sell, at which point the market value for the bonds starts to fall, too. Eventually, the natural prize reaches zero.

This is preventable on the margin if your “central bank” intervenes with dollars, but it’s still a Martingale bet until the central bank can fully buy back every token at the stated value. (Until then, a sufficiently brave enough short seller will eventually be willing to borrow and sell the tokens until the central banker runs out of dollars to buy them.)

Not Money Laundering

Good news! There’s a way to fully collateralize your token with US dollars. There are investors sitting on lots of cash who are just itching to exchange it for something instantly transferable around the world.

These investors are money launderers, though, and there’s a term for someone who accepts cash from a money launderer. It’s “Money launderer.”

The rules manage to be both nuanced and surprisingly broad, but the short version is that if you’re in a business where you allow customers to deposit and transfer large sums of cash, you should either a) know who those people are and how they came to possess so much cash, or b) expect to get in legal trouble.

The civil libertarian in me says that this is completely unfair. How can we presume guilt, just because someone has more money than the average person will earn in a lifetime in the form of hundred-dollar bills in a duffel bag? Surely, there’s a reasonable explanation! And surely, that person shouldn’t have to provide it!

But scratch a libertarian and you’ll find an economist, and here’s what an economist would say: the business of providing units of account, medium of exchange, and store of value is a business with powerful network effects — perhaps the strongest network effects in the world, and, if not, second only to language. Because of this, it’s prone to monopolization. And we all know that monopolists, especially government-owned monopolists, don’t have a strong incentive to provide superior customer service.

So the current legal standards re money laundering might be unfair, but they aren’t unexpected. If you’re a business owner, you need to deal with them.[1]


You can construct a toy stablecoin that doesn’t suffer from these two issues. Suppose I put a dollar in a safe deposit box, and collateralize one ByrneCoin with that dollar. I have a stablecoin, and as long as my bank isn’t destroyed it will be collateralized. But to be useful, a stablecoin needs to have enough liquidity to service the economy it’s attached to, and in crypto the size of the economy fluctuates a lot. You basically find stablecoins short on hard currency whenever times are good. (Interestingly, this same dynamic has occured in various frontier markets in the real world. The American Colonies, for example, didn’t have enough pounds, so they printed some fakes.)

To allow inflows is to raise your risk of violating anti-money laundering statues — and the faster you grow, the more you suffer from a Market for Lemons problem: if your stablecoin might be money laundering, people who want to treat it as equivalent to legal cash will stay away; their legal cash can’t get confiscated by the Feds if things go wrong. Whereas to people with dirty money, your coin might still have better odds. Maybe they expect half of their cash to be stolen, confiscated, or eaten by mice, but they think there’s only a 20% chance the coin gets shut down. They’ll switch from cash to coin. This means that as soon as your coin has a remote chance of money laundering difficulties, dirty money flows in, clean money flows out, and the odds of difficulties rise to 100%.

A stablecoin can shrink its circulation in response to lower demand for currency, which is a neat trick most fiat currencies can’t manage. However, you have to be pretty pessimistic to think this is the main feature you need.

Resolving the Trilemma

The trilemma isn’t quite a real trilemma, since all the problems are somewhat solvable. For example, if you start out with an enormous amount of capital, and attract capital from well-vetted institutional investors, you can achieve all three criteria. The questions are: who could do this, and how?

The who is: Someone who wants the advantages of both fiat currency (stable value) and cryptocurrency (ease of transfer, smart contracts). And that someone will probably be a player in the crypto space. If cryptocurrency prices never recover, there won’t be that much interest in Stablecoins, but if they do recover, the surviving crypto firms will have that much more collateral.

But, as I never tire of pointing out: cryptocurrency volatility is a function of its low valuation. Savings are just speculation slowed down so much it achieves a phase change and starts to behave differently. But really, they’re the same act. Someone buying a share of stock, or a token, or a dollar, is making the same bet: in the future, this will be exchangeable for goods and services on favorable terms. With the dollar, you get no price appreciation but no volatility, either; as assets get riskier, the prospect of price appreciation increases.

Most assets get less volatile as they go up in value, and currency-like assets should be an extreme case of this: at a low valuation, a crypto-currency is a bet that there’s a small chance it will be a vehicle for saving money in the future, and of course small chances fluctuate more. If it rises, the implicit odds go up, so the expected volatility goes down.[2]

Extrapolate that far enough in the future, and you can imagine a world where some crypto-assets are a default part of investors’ and central banks’ reserves, at which point their volatility would be similar to other reserve-like assets.

And if that happens, whichever coin wins is itself as close to a stablecoin as you’re likely to need.