Bezzle, Reflexivity and the upcoming collapse of LUNA and UST
TL;DR investors should sell all their LUNA and UST to avoid losing all their capital.
[This article has changed and will probably change in the future. I might change the wording of some phrases, add new data or correct mistakes. If you think there is a mistake, feel free to contact me and send me arguments and/or data that proof I’m mistaken, if that’s the case, I will happily correct it.]
A few weeks ago I came across the Terra protocol and its native tokens LUNA and TerraUS (UST). They struck my attention immediately. UST is a stablecoin that intends to represent $1USD, currently there are around 1̶1̶,̶2̶B̶ ̶U̶S̶T 16,3B UST in circulation and is 1̶0̶0̶% 90% unbacked. 1̶0̶0̶% 90%! The peg of UST is supposed to be kept through an arbitrage mechanism using LUNA. All LUNA tokens are currently valued at around $̶2̶3̶B $38B. LUNA value comes from the fees paid for transacting within the protocol. This design immediately raised a lot of red flags so I decided to dig deeper into the protocol. What I found would be hilarious if a lot of retail investors were not about to lose all their money. UST is an obfuscated ponzi scheme. Real demand for UST is practically nonexistent. Terraform labs (TFL) have been able to keep the appearance of real demand for UST by creating a decentralized lending platform called Anchor that subsidizes, in a completely unsustainable way, borrowers and lenders to borrow and lend using UST. The crypto bull market has allowed an influx of retail money into the ecosystem that has helped to push prices up and keep the illusion alive. This is however not sustainable, exit liquidity in the ecosystem is limited. As soon as selling pressure starts to accumulate, everything will collapse. Whales hold significant amounts of both UST and LUNA and just the exit of a few of them can cause the collapse. That is precisely what happened to TITAN, another stablecoin protocol, that went to 0 in June 2021. Titan was 75% backed by USDC.
This article is organised as follows. First I briefly explain two economic dynamics that have led to a mess of this magnitude, creating all the fictitious value in LUNA and UST. While these dynamics help explain what is happening, the ultimate people to blame are founders, venture capitalists and whales that have relentlessly shill and prop the demand of UST and LUNA in a completely deceptive way. Ultimately, these same dynamics will destroy all the value of LUNA and UST. Finally, I proceed to outline the workings of the Terra ecosystem, how the TerraUS-LUNA arbitrage mechanism works, how Anchor works and other red flags.
Bezzle
Bezzle is a simple yet very important concept to understand the dynamics of asset bubbles. The Bezzle is the temporary gap between the perceived value of a portfolio of assets and its long term economic value. For simplicity, let’s assume that there is just one asset. The price of the asset is determined by the intersection of supply and demand at a given point in time. Owners of the assets value their portfolio using that price. There is nothing wrong with valuing the asset using the latest market price, as long as its price is equal or less than its long term economic value (usually, the discounted cash flow you can obtain from the asset). If the price is below its long term economic value, an investor will come and buy that asset pushing its price up. The danger comes when the asset trades above that value. By valuing their asset at the latest market price, the supply and demand in a single point in time affect the perception of wealth of all the owners of the asset. This illusion of wealth can lead to an exuberant enthusiasm which in turn can feed a reflexive process (more about reflexivity below) that ends up when enthusiasts about the asset run out of money and there is no one else to bid the asset at those prices. The price of the asset will collapse and all those paper gains will disappear. When the asset in question is highly speculative, as it is the case with cryptocurrencies, the amount of Bezzle that can be created is enormous.
If you are interested about the concept of Bezzle you can read more about it here.
Reflexivity
Now let’s turn our attention to reflexivity, a concept first applied to finance by George Soros. Reflexivity refers to circular relationships between cause and effect embedded in human belief structures. Basically, it refers to a feedback loop. Reflexivity is mainly driven by psichological biases and leverage and is, alongside with Bezzle, the main driver of asset bubbles.
Let’s illustrate how a reflexive process might work. Early buyers of an asset start to see how the value of their portfolios slowly go up. They might buy more, or tell about their gains to other people ,some of which will buy more, which will in turn increase the price of the asset. Note that the gains that drive this reflexive process are for the most part Bezzle (if all holders of the assets were to try to convert it to real money its price would collapse). If there is the option of using leverage (there is in the case of crypto), it will boost this trend even more by adding more money to the demand of the asset. FOMO will kick in, further accelerating the upward trend. At some point however, the trend will peak, the number of potential buyers decreases (most people who wanted to buy already have bought) or/and the number of sellers increases (people wanting to cash out). Now the trend starts reverting fueled by the destruction of Bezzle, fear and lack of demand. Wealth was an illusion.
Terra, LUNA and UST
Terra is a decentralized blockchain designed to enable algorithmic stablecoins. Its main stablecoin is TerraUS (UST), which intends to represent 1USD. UST is 1̶0̶0̶% 90% uncollateralized. LUNA gives you a claim to transaction fees (if you stake it) and is used to support the peg of UST through arbitrage using the following mechanism:
If UST > $1. You can mint 1 UST by depositing $1 in LUNA. By doing that, you increase the supply of UST and drive the price down towards its peg. All the deposited LUNA is burned (reducing supply).
If UST < $1. You can deposit 1 UST and receive $1 in newly minted LUNA. By doing that, you are decreasing the supply of UST, driving up the price of UST towards its peg.
Note that there is no need for UST to be higher than 1$ to be able to swap 1$ worth of LUNA for it. You can do that at any time. This is a very important point to keep in mind and I talk more about it in the bitcoin section of this article and in my follow up article.
A more valuable LUNA makes the system more robust and a less valuable LUNA makes the system more vulnerable. This mechanism is highly reflexive. When there is “high demand” (keep reading to see the reason for the quote marks), that is, UST > 1$, the system takes LUNA out of circulation, making its price more prone to go up. On the other hand, when there is less demand for UST, the system has to print more LUNA to keep the peg, diluting its value precisely at the time when it is more needed. The relation between the marketcap of LUNA and the marketcap of UST is a very important datapoint to watch. The chances of a death spiral increase if LUNA marketcap drops below UST marketcap.
Demand dynamics of UST
Right now there are around 1̶1̶,̶2̶B 16,3B UST in circulation. In order for these tokens to be printed there must have been some previous demand. From where did that demand come? Mainly from Anchor. Around 75% of all UST sit on Anchor right now.
Anchor is a decentralized lending platform similar to Compound. Lenders can deposit UST and earn an APY of 19,5%. Borrowers can borrow UST by depositing collateral. Staking regards of the posted collateral are used to pay the lender, the remaining needed to pay the lender the promised 19,5% APY is paid by the borrower. Right now, that amount is close to 13%. Note that the borrower also has the implicit cost of the staking rewards he stops earning by depositing the collateral. Since the inception of Anchor, borrowers have been subsidised to borrow by rewarding them with ANC tokens, the governance tokens of the Anchor protocol. Right Anchor is paying around 10% in ANC token for borrowing.
In addition, there is a deficit between the amount deposited and the amount borrowed. Because payments from borrowers are not enough to pay lenders, the difference is paid by the yield reserve. Right now the yield reserve amounts to around 1̶1̶M 363M UST and is decreasing rapidly. In July 2021 Terraform Labs (TFL) had to inject 70M UST into the yield reserve to avoid its depletion. In February 2022, the Luna Foundation Guard (LFG) had to inject an additional 450M UST. That injection caused a sharp increase in UST deposits, as shown in the graph. That has exacerbated the unsustainability of the UST, when the 19,5 APY on deposits inevitably comes down, the chances of a run on UST are very high. What do borrowers do with the borrowed UST? They are mostly used to buy LUNA with leverage.
To sum up. Anchor is the main driver of the demand for UST. However, Anchor mechanics are clearly not sustainable. In the near future they will have to reduce the APY of deposits. When that happens, the yield offered by Anchor will no longer be competitive with the yield offered by other decentralised lending protocols on safer stablecoins like USDC or DAI. That will destroy the demand for UST and will greatly increase UST downward reflexivity. An Anchor governance proposal has recently passed that will lower the yield by 1,5% per month in case the rate of depletion of the yield reserve is high. However, there is a 15% cap below which the APY can not go. That 15% yield will still cause the yield reserve to be depleted fast.
I’ve seen some arguments from people admitting that the current yield is unsustainable but they justify the incentives by pointing out that they are needed to create awareness of Anchor and ‘network effects’. The network effect argument is complete nonsense, lenders and borrowers want a platform with a great variety of assets and the greatest risk-adjusted APY. As we have seen Anchor does not offer that. Even with the current ANC incentives, it is very expensive for borrowers to borrow in Anchor. By refusing to lower the yield on deposits, the unsustainability grows and grows over time, as the gap between deposits and borrowings show.
Other sources of demand for UST
Here I will talk about two use-cases of UST and Anchor frequently mentioned in the Terra community when someone points out that there is no real demand.
NFT
There are numerous NFT marketplaces built on the Terra ecosystem. But NFT’s marketplaces are by no way a relevant source of demand. First, a good chunk of the transactions in NFT’s markets are wash trades. Second, most real volume in NFT is a product of the current crypto bull market and won’t last. Third, who is going to pay thousand of dollars for pictures on the Internet? I could understand some whales might, or even some retail investors might spend a few dollars one a NFT. But a big sustainable demand for NFTs in Terra in the future? No way, especially when you are not the prime market for NFTs (e.g Opensea). Btw do you like my profile pic? It costs around $3000 in LUNA and I got it for free.
Chai and payment systems.
Another common argument in favour of the real demand on the Terra ecosystem is Chai, an e-commerce payment application operating mainly in South Korea. First, they do not use UST but TerraKRW, a much smaller stablecoin representing 1 southkorean won. Second, according the chaiscan, daily volume is laughable (around $96.000) [ Apparently data from chaiscan has not been updated since September 2021 ]. I also find highly unethical that multiple crypto venture capital firms like this and Terra shillers like this allude to Chai in a completely misleading way, by avoiding to mention in their ‘analysis’ that just a tiny fraction of the payments in Chai are done with TerraKRW or any other currency in the Tera ecosystem. They constantly share vague press releases where it is by no means stated that Chai actually uses Terra for their whole payment infrastructure. Please if you have an official statement from Chai saying that they use Terra for their whole payment system send it to me and I will happily remove this!
In addition, thinking that UST is going to be used as a payment system is a complete delusion. No regulator will ever allow that, nor will e-commerce sites want to hold such a risky asset. If they were to receive payments in UST they will most probably sell it immediately. Here is how well TerraKRW has worked as a stablecoin so far.
In addition to the projects and use-cases mentioned, there are dozen of other projects (most of them probably financed by Terraform labs and other VC firms) working on use-cases in the Terra ecosystem but the truth is that they haven’t generated and most likely won’t generate any significant demand for UST nor LUNA. Most of these project have the only purpose of creating additional leverage within the ecosystem.
The bitcoin reserve
The LFG recently announced its intention to purchase $10B in bitcoin to be used as collateral to back UST. Where does all that money come from? TFL kept for themselves more than 40% of the supply of LUNA and they have been giving those LUNA to the LFG to build up the reserves. You can track the LFG address here. But how are they transforming those LUNA to bitcoin? If they were to sell LUNA for USD in the market to buy Bitcoin, LUNA price would crash. Here is where the LUNA-UST swap mechanism comes into play. Terra allows you to burn $1 worth of fluctuating LUNA in exchange for 1 ‘fixed value’ UST. By using the LUNA burn mechanism you can transform LUNA into ‘dollars’ without negatively affecting LUNA supply and demand! When LUNA marketcap is very high, as it is right now, you can transform billions of LUNA into ‘dollars’ without affecting the price. Anchor’s yield reserve has been refilled using the same mechanism. Nonetheless, the problem of changing those UST for bitcoin remains, if you dump too many UST in the market you risk causing a depeg. Here is where Anchor comes into play. By offering a 19,5% APY they have generated a huge demand for UST. For example, the UST liquidity pool in curve has been unbalanced for many weeks, allowing LFG to dump UST there. However, the pool is now balanced so I highly doubt they will be able to dump much more UST to the market. So far they have only managed to buy around $1,5B in bitcoin. You can see the bitcoin address here. That makes UST ~10% backed. I personally think the $10B bitcoin buy is unattainable.
So does this backing make UST safer? It is certainly a step in the right direction, but I think that a 10% backing for a $16,3B asset is peanuts. In addition the value of bitcoin is not fixed, while the value of UST is fixed(in theory). The strategy of Do Kwon is that bitcoin will keep going up until UST is fully backed. The average price at which LFG has bought bitcoin is 41.000$. The risk of bitcoin going to 30.000$ or even lower is real, regardless of the probability you assign to it. We can say Do Kwon and the LFG are playing roulette with UST. Let’s not forget that UST is supposedly a ‘safe’ asset. To sum up, I think that regardless of the recent bitcoin backing, the fate of UST remains the same and all the conclusion of this article are still valid.
But… Terraform labs made simulations!
Indeed, they did simulations. The problem is that no economist or risk manager would take those simulations seriously. In the simulation section in the whitepaper there is not a single mention of the effects of variation in the price of UST and LUNA in the system! They just model transaction volume.
Here is a paragraph of the simulation section. Note the sentence “Transaction volume can be thought of as the GDP of the Terra economy”. That sentence would make anyone who have taken economics 101 cringe. Why? GDP is the amount of goods and services produced in an economy and financial transactions are not included in the GDP calculation. What has been produced in the Terra economy? Absolutely nothing. Terra volume is the result of only speculation. The only reason for using this GDP ‘analogy’ is to form the impression in the mind of readers that Terra is used for something else apart from speculation.
Good thing of this simulations is that they clearly show one of these two things about the creators of Terra:
- How ignorant they are on basic economic principles.
- Their intention to mislead potential investors about the inherent risks of the Terra ecosystem.
I personally don’t think they are stupid.
To sum up, the unavoidable fate of UST and LUNA is their collapse. Right now UST and LUNA holders collectively think they own around $4̶0̶B $50B in assets. All of that is fake, as the real demand for UST is 0. Let me end the article with this quote from Do Kwon, founder of Terra:
“The first goal of Terra is not to create a peg that never breaks, is to create an economy that is vibrant”
The vibrant economy is nowhere to be seen.
Last edition: 08–05–2022
You can see a follow-up article on Terra here. In it, I try to put together all the information in this article to present a simple picture of the Terra ecosystem. I also talk about how you can monitor the stress in the system and provide further evidence that proofs that Do Kwon knows the Terra ecosystem is completely unsustainable.
I want to thank @FreddieRaynolds (make sure to follow him on twitter). I learnt about Terra thanks to him and he has been kind enough to answer some questions I had during the research process.
