A Data-Driven Exploration of UST’s Collapse

Riyad Carey
Kaiko
Published in
8 min readMay 11, 2022

--

Subscribe to Kaiko’s weekly research newsletter here.

Introduction

TerraUSD’s spectacular rise ended in chaos and confusion this week following a dramatic depegging. The formerly third largest stablecoin by market cap dipped to as low as $0.3 USD, with virtually no reserves or liquidity remaining. While more details will be uncovered in the days and weeks to come, the data tells an interesting story.

Background

TerraUSD was launched in 2020 as one of the first decentralized stablecoins relying on a mint and burn mechanism with LUNA to maintain its peg; both tokens are native to the Terra blockchain. UST’s mechanism has allowed it to grow extremely quickly because it does not need to be overcollateralized like Maker’s DAI stablecoin. Instead, Luna is burned to create UST, which decreases Luna’s supply and has contributed to Luna’s large upwards price movement, the same works in reverse: Luna is minted when UST drops below its peg.

This mechanism can create rapid downwards price movements in the event of a depegging or a rapid reduction in UST supply. This past week proved the risks of such a mechanism: large UST sells and depegging crashed Luna’s price, and UST has thus far been unable to regain its peg.

Luna and UST Price

UST was announced and released in conjunction with Anchor, a lending protocol native to the Terra blockchain where users could earn 20% interest on UST deposits. In the early months, under 20% of circulating UST were deposited in the protocol. More recently this figure has held above 50%, making it the dominant use case for the stablecoin and driving UST’s growth from $180mn in January 2021 to $10bn in January 2022.

This allowed anyone to earn easy money with Anchor, triggering massive demand for a stablecoin with limited use beyond Terra, where it is used on a variety of DeFi applications. With UST’s massive surge in market cap, recent efforts had been made to expand its use on other blockchains, with Terra announcing the creation of the Curve 4pool (composed of UST, FRAX, USDT, and USDC) in an attempt to topple the dominant 3pool (DAI, USDT, USDC).

As charted below (from an old Kaiko weekly newsletter), UST never generated significant volumes on centralized exchanges, which pale in comparison to Binance USD (BUSD), to say nothing of USDC and USDT.

UST and BUSD trade volume on centralized exchanges

Depegging and Crash

The rapid growth of UST’s market cap driven by Anchor deposits led the Luna Foundation Guard (LFG) to begin buying Bitcoin in February to create a “forex reserve”. This was intended to create a “further layer of support for the UST peg using assets that are considered less correlated to the Terra ecosystem.” However, this mechanism faced two problems in the current downturn: (1) Bitcoin has gone down significantly, likely below LFG’s average buy price; and (2) according to Terra founder Do Kwon, the forex reserve mechanism had not been launched.

To add to the unfortunate timing, the Curve 4pool was not fully operational either, and Do Kwon stated that Terraform Labs had removed $150mn from the UST+3pool (the top UST pool on Curve) to deploy into the 4pool next week. This timing has led some to speculate that it was a coordinated effort to depeg UST, although nothing has been proved.

While the Curve UST+3pool was in a vulnerable state with low liquidity, large UST sell orders began rolling in on Binance, including a $10mn sell order. As the sell off started, the Curve UST+3pool quickly became unbalanced, and currently stands at nearly 90% UST and 10% DAI, USDC, and USDT, indicating that users were swapping out of UST for stablecoins that are seen as more reliable.

Liquidity Crisis on Centralized Exchanges

While most UST usage and activity happens on decentralized protocols, centralized exchanges are still extremely important in the price discovery process. Liquidity is extremely important to maintain the price of a stablecoin, which is why market makers play such an important role.

Stablecoin order books all have a similar shape, typically in the form of two giant buy and sell walls. For example,the USDT/DAI order book on Binance shows that liquidity is tightly maintained immediately around the mid price, with virtually no liquidity at any other price levels.

Whereas for a non-stablecoin trading pair such as BTC-USDT, liquidity is maintained in a V-shape, with constant fluctuations as Bitcoin’s price moves up or down.

Today, the UST-USDT order book resembles BTC-USDT’s, even though it is a stablecoin pair. That’s because UST is in pure price discovery mode.

Following the large sell orders and price dip, Binance’s UST-USDT liquidity immediately around the mid-price virtually disappeared. What is interesting to note is that in the run-up to the collapse, liquidity on the ask side of the order book increased considerably. Upon depegging, bid depth collapsed while ask depth fluctuated widely, ranging from 2 million to 20 million, likely contributing to the difficulties that UST had in re-pegging.

Far from being just Binance, UST liquidity evaporated on all major pairs on centralized exchanges.

Yet what is interesting to note is that when we look at 10% order book depth (a wider range from the mid price than 2%), we can see that after the initial depegging, ask depth soared to nearly 300 million. Could this massive sell wall have prevented UST from re-pegging? We don’t observe a similar trend on the bid side of the order book.

We noticed a similar phenomenon on Coinbase, where ask depth surged after UST depegged.

Additionally, funding rates and volumes suggest that Luna was and continues to be aggressively shorted.

Ultimately, the poor liquidity on centralized exchanges likely played a huge role in UST’s depegging, along with a series of events exacerbated by a rush of UST out of Anchor as users became concerned about the dreaded UST/Luna “bank run”. UST deposited in Anchor fell from over $14bn to about $6bn at the time of this writing.

Source: Anchor Protocol dashboard

This rush out of Anchor and Terra was further evident in volume on Wormhole, a bridge allowing users to move tokens from one blockchain to another; it processed nearly $2bn in outbound volume from Terra on May 8 and 9 alone.

Aftermath

This saga is far from over, and UST currently rests at around $0.45 at press. Luna has plummeted from highs of over $110 to under $1. Much of the crypto community has reveled in perceived comeuppance for Do Kwon, who had tweeted things like “By my hand $DAI will die.”

The Luna Foundation Guard has loaned $750mn worth of Bitcoin to OTC trading firms to protect the UST peg and has loaned $750mn of UST to accumulate more Bitcoin. On May 10, Do tweeted that he was close to announcing a “recovery plan” for UST, which is rumored to be a $1bn raise from Jump Crypto, Celsius, and others. In another tweet thread, he announced that UST will become collateralized.

Will UST return to its peg and, if so, can it regain trust? Only time will tell.

My Take

Having been involved in the Terra ecosystem since mid-2020, I have seen it grow from a small blockchain focused on mobile payments in South Korea to a DeFi powerhouse. In my opinion, and with the benefit of hindsight, there were a few major miscalculations as Luna and UST experienced massive growth:

  1. Anchor: In the early days of Anchor, the yield reserve was relatively stable because of the demand for borrows. However, as yields around the DeFi space dried up post-DeFi summer, Anchor’s 20% began to look increasingly appealing. It became clear, especially this year, that Anchor was running too hot and bringing un-sticky capital into the Terra ecosystem. I had long viewed Anchor’s yield reserve as Terra’s marketing budget, but the advertisement was simply too good and Anchor thus became UST’s primary use case (and a systemic risk) rather than allowing UST to grow organically from its other uses.
  2. Bitcoin buys: The thesis for adding Bitcoin as a forex reserve was flawed from the beginning: crypto (including BTC) is too volatile and correlated to provide any meaningful protection for the UST peg. Again, it was good marketing, as Bitcoin maxis and the media were more interested in Terra, but it did not add meaningful protection for the peg.
  3. Low liquidity on centralized exchanges: As evinced by its BTC loans to trading firms to protect the peg, Terra has relationships with market makers who are experienced in providing liquidity (and profiting from it). DeFi protocols have long incentivized users who provide liquidity on DEXs, and Terra should have considered methods to deepen liquidity on centralized exchanges. Peg protection protocols on Terra like White Whale and Kujira are of little use if most of the selling activity is happening on illiquid CEX pairs as shown above. While providing liquidity isn’t as splashy as huge Bitcoin buys, it is clear that the former would have been more helpful than the latter in this instance.

--

--