Published in


Collateral Damage

In the aftermath of the Terra collapse, those of us interested in stablecoins should start to take stock of what was learned. One of the lessons you will hear is that a stablecoin should be backed by external collateral.

Before we assess this advice, let’s start with some preliminary definitions:

Internal collateral: a collateral asset whose value depends on the stability of the coin.

External collateral: a collateral asset whose value is independent of the stablecoin.


So what does this have to do with the recent collapse of Terra? Terra was an algorithmic stablecoin protocol based on the Seigniorage Shares model. That is, the stablecoin UST was backed by internal collateral — the equity-like token LUNA. When the price of UST fell below the dollar, the protocol would start minting LUNA. The LUNA would be exchanged for UST, which could be burned to reduce supply and reestablish the peg.

The value of LUNA was based on the future growth of Terra. If UST couldn’t maintain stability, it wouldn’t grow, so the value of LUNA would fall. Minting more LUNA while the value was falling would put even more downward pressure on the price, causing the value of LUNA to collapse. And so it happened.

Blurring the Line

Based on the Terra example alone, it appears that relying on internal collateral is untenable. But before we settle on that conclusion I want to run a little thought experiment. Suppose a hypothetical protocol called stableETH uses ETH as collateral. Before long, stableETH becomes so popular that a large portion of available ETH gets used to mint stableETH. Then, imagine that a panic arises and people want to exit the stableETH protocol.

What happens to the price of ETH in this example? Well, when everyone exits stableETH it will cause a surge in the available supply of ETH. This will cause a fall in the value of ETH, which will likely cause a rush of people trying to get rid of their ETH before the price falls even further . So, what started as external collateral (ETH) started to look more like internal collateral when stableETH became a significant part of ETH.

Global Currency

If a currency gets big enough, virtually everything becomes internal collateral, in the sense that virtually all assets are correlated. Take a look at what happened when the US dollar started to lose it’s peg recently (i.e., inflation above target) and people started to expect rate hikes (i.e., that the Fed will start to sell assets to reduce inflation).

The dollar is so big that it’s correlated to just about everything. So when the Fed tries to stabilize the dollar, every asset class suffers, even BTC!

Of course, there are a lot of other things going on in the economy right now, but the point is that the difference between internal backing and external collateral isn’t always so clear.

The implication is that relying on external collateral only works when the stablecoin protocol is small relative to the collateral. In the case of a global currency like the USD, there may not be any such collateral available. So, in the short term is is probably wise for crypto stablecoins to use established collateral. But if we expect any cryptocurrency to become large, we may have to look at other ways to ensure stability.

Debt and Equity

In addition to internal collateral and external collateral, we should also make a distinction between debt collateral and equity collateral. The difference between debt and equity has to do with the balance of risk and reward. Debt gets paid first (at a set rate of return) and equity gets paid last (with a variable return).

Not only was LUNA internal collateral, it was internal equity-like collateral. I say equity-like instead of simply equity due to the unique properties of the seigniorage shares model, but the difference is not crucial here. The bottom line is that equity gets paid after debt, so it will lose value sooner than debt.

We should expect that backing a stablecoin with debt should be somewhat more reliable than backing a stablecoin with equity. Stablecoins are a form of liability, but they are not the same as debt. Stablecoins are subject to runs because people can “redeem” them any time, whereas debt is a promise to pay sometime in the future.

Thus, you can (theoretically) avoid a run by exchanging an immediate liability for a future promise to pay. Using debt to stabilize a token may sound familiar. The Basis protocol (Terra founder Do Kwon’s previous project) used a bond-like token called Basis Bonds to achieve stability. Basis crashed as well, so clearly using debt instead of equity isn’t a cure-all.

Debt-to-Revenue Ratio

The big question when issuing debt (or equity) to stabilize a token is whether the market thinks your promise to pay (or your future revenue) has any value. If people lose faith in the protocol entirely, neither one will work. But this “faith” isn’t necessarily based on nothing. One way to give people confidence in the ability of your protocol to pay your debts is to make sure your debt payments are less than your revenue.

Consider a simple example where a stablecoins protocol has a business model of collecting fees for certain types of transactions. Then imagine that the protocol can also issue bond tokens that pay out $1/yr perpetually (where $ is used to signify the native token), and these are valued at $20 (i.e., based on a 5% discount rate).

Let’s say there are 100K active users of the protocol that own an average of $1000 in stablecoins, and pay an average of $100/yr in fees. These active users hold a total of 100M stablecoins, and generate a total of $10M/yr in fees. The 100M stablecoins could (theoretically) be converted to 5M bond tokens, which would represent $5M/yr in payments.

However, now imagine that there are also some speculators that also hold 100M stablecoins, but they don’t really use the coins as much, generating only a total of $1M/yr in fees. These coins could also be converted into debt representing $5M/yr in payments.

The active users generate an income stream sufficient to justify issuing them stablecoins because the rate or revenue per coin is 10%, which is more than the 5% rate on the debt. But the speculative users generate revenue at a rate of only 1%, which is less than the rate on the debt. So issuing to active users results in a financially stable coin, whereas issuing to speculators does not.

Now, you don’t really know who will be an active user and who will be a speculator, but you can stop issuing coins once your fee ratio dips below the rate on your debt (or some more conservative threshold). Of course, many protocols (including Terra) have been so eager to get more adoption that they pay users to speculate on their coins rather than collecting fees from them. This, of course, is not sustainable at all.


So here are a few take-home lessons:

  1. External collateral is a simple way to make sure a stablecoin is sustainable, but it isn’t the only way.
  2. If a stablecoin becomes big enough, its collateral could go from being independent of the protocol (i.e., external collateral) to something more like internal collateral.
  3. Stabilizing a coin with debt is less likely to cause a death spiral than stabilizing with equity
  4. The number of stablecoins backed by internally issued debt should depend on the fees collected from the issued coins.



Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store


Patent Attorney, Crypto Enthusiast, Father of two daughters