In the wake of the unexpected collapse of the Terra stablecoins Luna and UST, the financial world has learned a harsh lesson. People need reliable financial services that aren’t going to wreak havoc on people’s finances in a matter of days.

Advocates of TerraUSD (UST), a dollar-pegged stablecoin, anticipated it would upend existing payment systems throughout the world. However, it was quickly depleted as investors panicked and rushed to withdraw their cash, resulting in a violent, self-enforcing bank run. The crypto market capitalization of almost $400 billion was wiped out as a result of the crisis, which bankrupted many investors.

Terra’s demise might have short-term and long-term consequences for the crypto industry and beyond, especially when lawmakers and regulators who are wary of the technology take a closer look at the damage.

Let’s begin by looking at the entire chain of events that unfolded.

Dissecting the Crash

Without prior knowledge of the blockchain, Terra’s rapid rise and downfall can be difficult to explain effectively. Indeed, many of its supporters used obfuscation and rhetoric to mask some of its glaring shortcomings.

Terra has its own blockchain, much like Bitcoin or Ethereum. Its flagship product is the UST stablecoin, which is linked to the US dollar. Stablecoins are utilized as safe havens by crypto traders when markets in DeFi (decentralized finance) get choppy: rather than changing their more volatile assets into hard cash, which may be costly and trigger tax implications, traders simply exchange them for stablecoins.

Some stablecoins get their value from being fully backed by reserves: if investors choose to withdraw their funds, the stablecoin’s foundation should theoretically have enough cash on hand to refund them all at once. On the other hand, UST is an algorithmic stablecoin that depends on code, continual market activity, and sheer trust to maintain its peg to the US dollar. The theoretical support for UST’s peg was also provided by its computational connection to Terra’s basic currency, Luna.

Investment in UST surged in the past months as a result of a platform called Anchor, which provided a 20% return to those who purchased the token and committed it to the protocol. When this opportunity was announced, many detractors quickly compared it to a Ponzi scam, claiming that Terra could not possibly give such a large return to all of its investors. Terra team members recognized this but compared the pricing to a marketing expense to create awareness, similar to how Uber and Lyft gave heavily reduced fares at the start of their operations.

What started at 20% and was billed as “stable” gradually began to fall following the passing of Proposal 20 in March. This proposal indicated that if Anchor’s reserves rose by 5%, the interest rate would rise. If these reserves fell by 5%, the interest rate would fall as well.

Furthermore, if there were more lenders than borrowers on the platform, this rate was predicted to fall by 1.5 percentage points per month. With interest rates predicted to decline, the primary use case for UST began to wane. On April 23, for example, Anchor held more than 72 percent of the UST in circulation. The entire point of this stablecoin was for it to be invested in Anchor.

When it became evident that the 20% interest rate was not going to persist, UST holders began to flee.

On Friday, May 6, there were around 14 billion UST in Anchor. By Sunday, the tally had dropped to 11.7 billion. Note, UST was still pretty much tied to the dollar at the time; thus, around $2.3 billion in capital was gone over the weekend.

And, given that UST holders were only engaged in the Terra network because of Anchor, their departures meant they had no purpose for that UST. As a result, we experienced a massive outflow.

You have two alternatives for exiting UST. The first option is Terra’s widely popular burn-and-mint system.

Holders can use this technique to exchange one UST for one LUNA, destroying the UST in the process. Whenever 1 UST goes below $1, this presents an arbitrage opportunity, as speculators may buy the discounted UST and swap it in for $1 in LUNA, generating a little profit. The inverse is also true: if UST trades for more than $1, you can swap (and burn) $1 of LUNA for that UST.

The second option is to use the stablecoin exchange Curve Finance.

When a stablecoin has a minor price fluctuation, savvy arbitrageurs will often rush over to DeFi’s biggest liquidity pools on Curve and trade the discounted stablecoin to whatever alternative has maintained its peg.

For instance, if DAI is trading at $0.99, investors will purchase the discounted DAI and sell it for USDC (in this case, $1) to profit. Normally, this buying pressure drives the price of DAI back up to $1. This is also true for UST.

Now, let’s take a look at why things went so wrong for UST.

First, there’s the burn-and-mint exit. The UST supply cratered as it burnt, whereas the LUNA supply mooned. This was also not a painless exit, as consumers encountered a number of technological challenges.

Remember that Terra is a blockchain network with gas fees for operations. Along with rising gas prices, the network is also constrained in terms of how much UST or LUNA can be burnt or minted at once. Because things were slow and clogged, exchanges began to pause withdrawals.

This burn-and-mint method could potentially have an effect on the LUNA pricing.

Swapping and burning UST for LUNA generates more LUNA, decreasing the supply and reducing the price of this token. Furthermore, as the price of LUNA lowers, exchanging 1 UST for $1 worth of LUNA will require more and more LUNA to reach the $1 threshold.

At some point, LUNA’s price might drop so low that there isn’t enough liquidity to absorb all the UST flowing in.

As for Curve Finance, here’s how it looked.

Over the weekend, UST fell by less than $0.02 as Anchor users jumped in and began trading UST for any other stablecoin, such as Tether’s USDT or Circle’s USDC.

Eventually, the pool that permitted these transactions (dubbed the “UST + 3Crv” pool, which also included all of the major stablecoins) became imbalanced, meaning there was considerably more UST in the pool than the other stablecoins.

What happens when you sell UST for USDC on Curve?

If you sell UST on Curve for USDC, you will add additional UST to this pool while removing USDC. The pool will eventually include more UST than USDC. To correct course, the pool begins to offer that UST at a discount with the aim of convincing arbitrageurs to execute the opposite transaction (and rebalance the pool).

This is why we started seeing a tiny de-pegging over the weekend—Curve was doing what it has been doing since its inception.

The issue in this scenario was that the opposing trade, which would rebalance the pool, was not taking place. Despite the reasonably profitable arbitrage trade, it appeared that no one wanted to be in control of UST. As for why, keep in mind that Terra’s flagship app, Anchor, has begun to falter.

Soon one investor sold almost 85 million UST tokens in return for 84.5 million USDC tokens in this pool. This, of course, increased pressure on UST’s dollar peg as Curve continued to generate the discount in the hopes of incentivizing arbitrage traders to rebalance the pool.

A $0.02 devaluation on Sunday became a staggering $0.32 by Tuesday. Simultaneously, the $64 LUNA token plummeted below $30.

Around this time, UST’s market value began to overtake LUNA’s, implying that the latter would no longer be able to soak up the former, resulting in a state of collapse. This is when the Luna Foundation Guard (LFG) naturally intervened.

It put a ton of non-UST stablecoins into the Curve pool to help the stablecoin establish its peg. It then apparently began deploying its Bitcoin holdings to a “professional market maker,” who was effectively directed to spend BTC when UST fell below the peg. As such, UST rose from $0.64 to $0.93, but momentarily.

Constant exits via Curve depleted the bailout liquidity. It’s also unknown whether the BTC was ever utilized to defend the peg.

Eventually, when people abandoned UST and subsequently sold their LUNA, the price of LUNA fell so low there wasn’t enough headroom for UST, resulting in a massive amount of bad debt.

Do Kwon and the Terra community doubled down and expanded the amount of LUNA that may be coined at one time. But all it did was speed up the spiral. On May 8, LUNA had a circulating supply of 343 million. By May 12, the total had risen to 32.3 billion.

Despite the negative feedback loop, the Terra community offered three more emergency actions, all of which came down to just setting fire to as much UST as possible (without having to mint LUNA on the other end).

What are the lessons to be taken away?

The most apparent is that cryptocurrencies are risky, but you presumably already knew that.

Let’s look at the shortcomings in the design. In theory, when you peg one asset with another asset as collateral, there is always the possibility of under collateralization or depegging. Even if it is ten times collateralized, the collateral asset might fall more than ten times. Nothing is 100% stable.

The greatest irrational design fault, according to Binance CEO Changpeng Zhao, is believing that minting more of an asset would raise its overall worth (market cap). Printing money creates no value; it only dilutes current holders. Exponentially minting LUNA exacerbated the situation.

The overly aggressive incentives were the second basic issue. Anchor’s 20% fixed APY is specifically designed to encourage (in-organic) development. Incentives may be used to entice users to join your ecosystem. However, in order to sustain it, you must earn “income,” that is, more revenue than expenses. Otherwise, you will run out of funds and have to declare bankruptcy.

However, in this scenario, the definition of “income” is muddled because the project team most likely considered their own token sales or price appreciation of LUNA as “income.” This strategy is flawed. Yes, people come in because of the incentives, and the value of LUNA rises. And more people are receiving incentives. But no value is generated.

We must always remember that high APYs don’t always indicate healthy projects.

It’s now clear that the entire system was founded on a self-perpetuating, superficial premise. While Terra did have an ecosystem with certain use cases, the ecosystem’s expansion did not keep pace with the incentives employed to attract new users. The development was “hollow.” This resulted in its complete collapse.

The key takeaway here is to avoid chasing high APY.

Closing note

The crypto industry is still in its infancy, and decentralized software platforms like Ethereum provide virtually limitless opportunities for developing alternative financial products. These services have not been tried and tested in the same way that traditional finance has, and things may go wrong quickly and severely.

More specifically, while not intrinsically bad, algorithmic stablecoins are especially risky. A stablecoin that is dollar for dollar-backed by actual cash in a bank somewhere is significantly less likely to lose its peg to the dollar. An algorithmic stablecoin, although always partially backed by something valuable someplace, is complicated. And while everything may appear to be great 99 percent of the time, when the markets begin to crumble, the mechanism preserving its value may fail. This is why looking at the fundamentals over the incentives is important.