What can other algorithmic stablecoins learn from Terra’s crash?

The spectacular implosion of the Terra ecosystem in mid-May marked the crypto industry. While there were some brave reviewers who realized how thin the razor’s edge was for TerraUSD (UST) – now TerraUSD Classic (USTC) – I think it’s safe to say most people don’t hadn’t expected Terra to fail so quickly, so dramatically, and so completely irrevocably.

I write this as the Terra community votes on a plan to reboot a kind of Terra 2.0 – a plan to save the Layer 1 ecosystem without the UST stablecoin. The old Terra, now known as Terra Classic, is completely dead. An ill-fated attempt to prop up UST holders printed billions of LUNA tokens, destroying their value and ultimately jeopardizing the security of the network itself.

The complete wipeout worth $50 billion seems to have made people decide once and for all that algorithmic stablecoins can’t work. But I think it’s important to have a more nuanced understanding of why the original LUNA failed and how others can learn from its lessons.

Related: Terra 2.0: A Crypto Project Built on the Ruins of $40 Billion in Investor Money

Stablecoins: New name for an age-old concept

The term stablecoin primarily refers to currencies pegged to the US dollar that aim to maintain a value of $1. But it’s important to remember that it’s mostly a matter of convenience. The same mechanics that underlie today’s USD stablecoins can be used to create coins pegged to Euro, Gold, even Bitcoin (BTC), Nasdaq futures, or some specific stocks, such as Tesla (TSLA).

It’s also worth noting that stablecoins aren’t really a new idea in crypto. Today’s stablecoin designs are closely tied to either how money works under a gold standard – for example, Maker’s Dai is a claim to material collateral just as the first banknotes were claims to a gold vault – or they are a reproduction of pegged currencies such as the Hong Kong Dollar.

The HKD is a very interesting example in all of this, as it is pretty much your ordinary “algorithmic stablecoin”. It is pegged to the US dollar, although not at a 1:1 ratio, and the HK central bank uses its vast reserves to keep the price of HKD within a well-defined ratio by trading it on the market . The latest audits put Hong Kong’s reserves at $463 billion, six times the HKD in immediate circulation and almost half of its M3, the broadest definition of ‘money’ which also includes assets that are not immediately liquid. (such as locked bank deposits).

Really, the only reason why HKD is technically not an algorithmic stablecoin is that there is a central bank that does market operations. In decentralized finance (DeFi), the central bank is replaced by an algorithm.

Related: Consequences of UST: Is there a future for algorithmic stablecoins?

Terra isn’t HKD, though

Confusing Terra with the algorithmic stablecoin space, in general, makes it unclear why Terra collapsed as badly as it did. It is important to realize how fragile the design of the Terra protocol was. In a nutshell, UST has been “collateralized” by LUNA, the gas token of the Terra blockchain. Since there was a fairly robust DeFi and non-fungible token ecosystem being developed on Terra, the LUNA token had inherent value that helped boost UST’s initial offering.

The operation of the mechanism was, in principle, similar to that of HKD. If UST traded above $1, users could acquire LUNA and burn it for its dollar value in UST. Basically, the system has assumed that the UST is worth $1, so the LUNA engraver can simply sell the UST on the market for, say, $1.01 and make a profit. They can then recycle the profits back into LUNA, burn them again, and continue the cycle. Eventually, the ankle would be restored.

If the UST was trading below $1, the reverse mechanism helped support it. Arbitrators would buy the cheap UST, exchange it for LUNA at a rate of 1 UST equaling $1, and sell those tokens on the market with a profit.

This system is excellent for supporting the ankle under normal circumstances. A problem with Dai, for example, is that it cannot be directly arbitrated for its underlying guarantee. Arbitrators have to “hope” the peg stabilizes to make a profit, which is the main reason why Dai is so dependent on USD Coin (USDC) now.

But we must also mention the extreme reflexivity of Terra’s design. The demand for UST that moves it above the peg causes a demand for LUNA, and therefore an increase in price. The keystone of this mechanism was Anchor, the lending protocol on Terra that guaranteed a 20% APY to UST stakers.

Where does the 20% APY come from? From extra UST minted to Terraform Labs LUNA Reserves. A higher LUNA price meant they could hit more USTs for the peg yield, thus increasing the demand for USTs and increasing the price of LUNA – so they were able to hit even more USTs…

UST and LUNA were in a reflexive demand cycle that, let’s face it, had all the elements of a Ponzi. The worst part was that there was no cap on the amount of UST that could be minted, for example as a percentage of LUNA’s market capitalization. This was purely driven by reflexivity, which meant that just before the crash, $30 billion of LUNA’s market capitalization supported $20 billion of UST’s market capitalization.

As Kevin Zhou, founder of Galois Capital and famous critic of LUNA and UST before its collapse, explained in an interview, every dollar invested in a volatile asset increases its market capitalization by eight times or more. In practice, this meant that the UST was grossly under-guaranteed.

Poke the bubble

It is difficult to determine the precise reason why the collapse started when it happened, as there were certainly multiple factors at play. On the one hand, Anchor’s reserves were visibly depleted, with only a few months of yield remaining, so there was talk of reducing yield. The market wasn’t doing too well either, as most of the big funds started to expect some sort of big crash and/or an extended bear market.

Some conspiracy theorists blame TradFi giants like Citadel, or even the US government, for “shorting” the UST with billions and starting the bank run. Either way, it’s crypto: if it’s not the US government, it’ll be a wealthy whale who wants to be known as the second coming of Soros (who notoriously shorted the pound then that it had a similar peg configuration, known as Black Wednesday. Although less spectacular than Terra, the pound lost 20% in just two months).

In other words, if your system can’t handle coordinated, well-funded attacks, it probably wasn’t a good system to begin with.

Terraform Labs sought to prepare for the inevitable, collecting a total of around 80,000 BTC which is believed to support the peg. He was worth around $2.4 billion at the time, which is nowhere near enough to redeem all of the UST holders who wanted out.

The first depegging event between May 9-10 took the UST to around $0.64 before recovering. It was bad, but not deadly yet.

There’s an underrated reason why UST never recovered. The LUNA redemption mechanism I explained earlier was capped at around $300 million per day, which was ironically done to prevent a UST run from destroying LUNA’s value. The problem was that LUNA crashed anyway, rapidly dropping from $64 to around $30, which has already lost $15 billion in market capitalization. The depeg event barely lost any supply of UST as more than 17 billion remained of the initial 18.5 billion.

With both Do Kwon and TFL silent for the next few hours, LUNA’s price continued its slump without any significant redemption activity, hitting lows in the single digits. It was only here that management decided to raise the repurchase cap to $1.2 billion when LUNA’s market capitalization had already fallen to $2 billion. The rest, as they say, is history. This hasty decision sealed the fate of the Terra ecosystem, leading to hyperinflation and subsequent shutdown of the Terra blockchain.

Related: Terra Collapse Highlights the Benefits of CEX Risk Management Systems

It’s all about the warranty

Successful examples of TradFi like HKD should be a clue to what happened here. Terra seemed to be oversized, but it really wasn’t. The actual pre-crash collateral was perhaps $3.6 billion (Bitcoin reserves plus Curve liquidity and a few days of LUNA redemptions).

But even 100% is not enough when your collateral is as volatile as a cryptocurrency. A good collateral ratio might be between 400% and 800%, which is sufficient to account for the compression in valuations mentioned by Zhou. And smart contracts should strictly enforce this, prohibiting the minting of new coins if the collateral isn’t ideal.

The reserve mechanism should also be maximally algorithmic. So in the case of Terra, Bitcoin should have been placed in an automatic stabilization module instead of opaque market makers (although here there was simply not enough time to build it).

With safe collateralization settings, a bit of diversification, and a real use case for the asset, algorithmic stablecoins can survive.

It’s time for a new design for algorithmic stablecoins. Much of what I have recommended here is contained in the Djed whitepaper that was published a year ago for an oversized algorithmic stablecoin. Nothing has really changed since then – Terra’s collapse was unfortunate but predictable, given how undersecured it was.

This article does not contain investment advice or recommendations. Every investment and trading move involves risk, and readers should conduct their own research when making a decision.

The views, thoughts and opinions expressed herein are those of the author alone and do not necessarily reflect or represent the views and opinions of Cointelegraph.

Shahaf Bar-Geffen has been CEO of Coti for over four years. He was also part of the founding team of Coti. He is known as the founder of WEB3, an online marketing group, as well as Positive Mobile, both of which were acquired. Shahaf studied Computer Science, Biotechnology and Economics at Tel Aviv University.

About William G.

Check Also

The Evolution of Cryptocurrencies and Digital Assets

Countries and governments are scrambling to accommodate these assets and institutions, while more and more …