Algorithmic stablecoins are hot right now. But investors like Mark Cuban have lost money, and warn about their risks.

Never miss a story: Follow your favorite topics and authors to get a personalized email with the journalism that matters most to you.

Popular stablecoins like Tether and USD Coin are backed by reserves of U.S. dollars and other assets in an effort to keep these cryptocurrencies from fluctuating wildly in price like some other digital tokens.

Now, a newer kind of stablecoin—algorithmic stablecoin—is getting buzz. But instead of being backed by fiat currency or other crypto, it’s backed by computer code.

The price of these coins is based on algorithms that let traders create and destroy coins as needed to maintain their price. The result is supposed to be the same as their predecessor: a stable coin that is nearly always equal to the price of a fiat currency like the dollar.

But some have criticized this approach. Well-known investor Mark Cuban said he lost money when the algorithmic stablecoin IRON lost its dollar peg after a selloff by large investors. He has said stablecoins will be the first to be regulated.

“What is an algorithmic stablecoin? Is it stable? Do buyers understand what the risks are? It needs standards,” Cuban tweeted in September.

Others, like Ryan Clements, a professor of business law and regulation at the University of Calgary, say this version of stablecoin is inherently unstable, and could be susceptible to something like a “bank run” that could make them worthless.

What are stablecoins?

Before understanding how algorithmic stablecoins work, you first must understand regular stablecoins.

Some of the first stablecoins, including Tether, were created in 2014 as a response to major swings in crypto prices. Although tied to the price of fiat currencies like the U.S. dollar, they are easier to transfer between decentralized exchanges than dollars.

In case of a crypto downturn, investors can theoretically keep their assets safe by exchanging a cryptocurrency like Ether, that is prone to fluctuating prices, to a stablecoin like Tether, which is meant to maintain its value at $1.

The two most popular stablecoins are Tether and USD Coin (USDC), both of which claim to be backed by an equal dollar amount of assets. Tether claims its tokens are 100% backed by cash and cash equivalents, while USDC says it is backed with a combination of cash, Treasury bonds, commercial paper, corporate bonds, and certificates of deposit with foreign banks.

What are algorithmic stablecoins?

Algorithmic stablecoins, unlike regular stablecoins, are not backed by assets but rather an algorithm that puts in place incentives for traders to maintain the price.

The most popular algorithmic stablecoin is TerraUSD (UST), the third largest stablecoin by market capitalization behind USDC and USDT, according to CoinMarketCap. Other algorithmic stablecoins include Frax and FEI USD.

TerraUSD was created by South Korean crypto developer Do Kwon and his company Terraform Labs. It maintains its dollar value through a system that relies on traders burning or creating tokens for profit to maintain a stable price. This process works through TerraUSD’s mutually dependent pairing with another Terraform Labs cryptocurrency, Luna. Every time a TerraUSD token is minted, the equivalent of $1 in Luna is burned, and vice versa.

When the price of TerraUSD drops below $1, traders can burn TerraUSD, or remove it from circulation, in exchange for a dollar in Luna. This reduces the supply of TerraUSD tokens, and therefore raises the price of the token. If the price of TerraUSD exceeds a dollar, traders are incentivized to burn Luna in exchange for a dollar in TerraUSD, which increases its supply and drops the price.

This algorithmic balancing act is intended to maintain the price of TerraUSD at $1, and has been mostly successful so far. TerraUSD’s price has only dropped far below a dollar twice since it was created in 2020, according to CoinMarketCap In December 2020, it dropped to around 85 cents and to about 94 cents in May 2021.

Controversy around algorithmic stablecoins

Algorithmic stablecoins are gaining popularity.

In the last quarter of 2021, TerraUSD saw a 260% increase in its overall market value, according to CoinGecko. About 17.5 billion TerraUSD tokens were in circulation as of Monday.

Yet despite their increased adoption, some have criticized how TerraUSD and other algorithmic stablecoins operate. Clements, the University of Calgary professor, wrote in the Wake Forest University Law Review last October that algorithmic stablecoins only maintain their price because traders expect the coin to have value in the future.

“These uncollateralized digital assets, which attempt to peg the price of a reference asset using financial engineering, algorithms, and market incentives, are not stable at all but exist in a state of perpetual vulnerability,” he wrote.

The problem with algorithmic stablecoins is three-fold, according to Clements. First, the supposedly stable cryptocurrencies require a certain level of demand to operate as expected. If demand falls below a certain threshold, the system falls apart.

Second, because the stablecoins rely on independent actors with no legal obligations apart from their own self-interest, the system is fragile. Finally, he writes that in times of crisis this type of stablecoin could be prone to unclear information and uncertainty that can cause herd mentality that will hurt the system.

Ultimately, algorithmic stablecoins could be prone to “bank runs” if traders keeping the price afloat decide to take their money out of the system over a short period of time, Clements argues.

He writes in the report that algorithmic stablecoins should be regulated to guard against what he calls “a system that will be prone to runs, destabilization, and failure when reality deviates from the assumptions underlying the embedded incentive structure.”

Do Kwon makes a big bet on algorithmic stablecoins

Terraform Labs’ Do Kwon is undeterred by critics. Last month, Kwon bet a crypto investor who goes by the pseudonym Sensei Algod $1 million that the price of Luna, which depends on TerraUSD, would be higher in a year, according to crypto news outlet, The Block.

But TerraUSD may soon have more than an algorithm to back it.

Earlier this year, Kwon announced plans to obtain $10 billion in Bitcoin to hold as reserves to back TerraUSD. In early April, he told tech news site TechCrunch that Terra had purchased a total of $1.6 billion worth of Bitcoin so far, and was planning to purchase another $1.4 billion through the Luna Foundation Guard, a nonprofit organization which supports TerraUSD and Luna. The Terra Protocol, which underpins both TerraUSD and Luna, will obtain the next $7 billion in Bitcoin through users that want TerraUSD and are willing to exchange their Bitcoin for the stablecoin.