Over the course of a few days last week, we saw what was perhaps the most dramatic flame-out in the history of cryptocurrencies.
Just a few weeks earlier, Terraform Labs the creators of the TerraUSD and Luna cryptocurrencies, was riding high.
Luna was valued at US$86 per token on 5 May – down from the US$120 peak, but most tokens were down at the time – and it was ranked as high as number 6 in terms of market capitalisation among cryptocurrencies, with a total market capitalisation of US$28 billion.
On 13 May, the blockchain had even been shut down by the
validators – those responsible for verifying the transactions on the blockchain – though it would later resume block production with some functionality disabled.
A VC darling, backed by the likes of Pantera Capital and Kinetic Capital, Terraform Labs saw its first initial coin offering in 2019 but started to really take shape in January 2021 after a round of venture capital funding.
The founder of TerraForm Labs, Kwan Do-hyung (better known as Do Kwan) claimed he had created something unique, capable of weathering both bear and bull markets.
Investors – retail and institutional – agreed. Luna job function email database exploded from a market cap of US$2.5 billion in June 2021 to a peak of over US$40 billion in April this year.
An algorithmic stablecoin
To understand how this happened, you need to have a bit of an understanding of how stablecoins work, and how Terra/Luna worked in particular.
A stablecoin is a currency that is pegged to the value of an asset – most commonly the US dollar, but it could also be euros or gold.
Even though it’s traded on an exchange, a stablecoin should always be valued 1:1 with the underlying peg.
There’s a relatively complex taxonomy of types of stablecoins, but most broadly they can be broken down into two types: backed and algorithmic.
The world’s three most popular stablecoins – Tether (USDT), USD Coin (USDC) and Binance USD (BUSD) – are all backed stablecoins.
That means there is a central company that holds a reserve equal to the number of tokens in circulation. In theory, Tether holds $1 of real-world assets for every USDT.
The companies then guarantee that they will trade, 1:1 for the asset, so you can trade 1 UDST for 1 US dollar at any time.
Natural arbitrage takes care of the rest.
So, let’s say that USDT slipped to US95c. An arbitrageur (or rather an automated bot owned by one) could and would then buy as many as they can, since they can immediately trade it in with Tether for $1. An instant, no-risk profit for the arbitrageur, and an increase in demand that will push the price back up to the peg.
Likewise, if the value of USDT went above the peg, say US$1.05, an arbitrageur bot could go to Tether and say, “here’s some amount of US dollars; please mint me an equal amount of new USDT.”

They would immediately sell this new USDT at a profit: that is, buy at $1 and sell at $1.05. The increase in supply would naturally drive the price back down.
Algorithmic stablecoins (sometimes called algos, not to be confused with the cryptocurrency Algorand), however, work a little differently.
They maintain a peg by using supply and demand manipulation managed by (as the name suggests) an algorithm.
If the value of an algo stablecoin drops below the peg, then they reduce supply of the stablecoin to drive up the price. If it goes above the peg, they automatically mint new stablecoin tokens to increase supply and drive the price down.
Enter TerraUSD (UST), an algorithmic stablecoin with a unique mechanism for managing the price.
While TerraUSD was a stablecoin, it was inextricably linked with Luna (LUNA), a free-floating cryptocurrency, and the two worked in tandem to maintain the peg.
Luna, being a free-floating smart contract platform, is priced at whatever the market decides (US$120 at its peak; now US$0.0022), just like most cryptocurrencies.
The key to the Terra/Luna system was that UST was ‘backed’, in a sense, by LUNA. LUNA serves as a kind of counterweight to keep UST at its peg.
The way it worked was that if UST slipped below the peg, then the blockchain algorithm would mint new LUNA tokens, and use that LUNA to buy UST at its current exchange rate (sellers would be incentivised by arbitrage, much as in a backed coin, to make the trade).
The purchased UST would then be ‘burnt’ (ie. destroyed). That would drive up demand and reduce supply of UST to return the price to its peg.
Conversely, if UST went above the peg, it would mint new UST and use it to buy LUNA tokens. That LUNA would then be burnt. The increase in UST supply would drive the price down.
So, you have a ying and yang system: LUNA is burnt to mint UST when necessary to maintain the peg, and UST is burnt to mint LUNA when necessary.
Terra/Luna had numerous critics, who noted that this entire system was built on sand, and the weird ouroboros of Terra and Luna could easily come crashing down.
But for a time the system worked, until suddenly it didn’t.