Arthur Hayes talks about the UST death spiral and reviews the market outlook of various stablecoin pairs

As humans, our existence in this universe has shockingly little control over what we can actually do. Even so, we, as a civilized society, spend far too much energy bringing calm and stability to a troubled planet. The mere act of adjusting the temperature up and down in our homes and workplaces takes a lot of energy.

Because we instinctively understand that we are nothing more than reeds blown at will in the wind, humans place enormous power on those individuals and institutions that provide the siren call of calm. Our politicians tell us they have a plan, our corporate chiefs chart a path to future profitability, and we expect the same qualities to manifest yesterday, today and tomorrow. But time and time again, the universe drops unexpected bombshells, and our leaders' best plans are often woefully inadequate. But what can we do, but try and try again?

Just as our civil society exhibits calm, the currencies that power civilization must appear stable. Fiat currencies are designed to depreciate slowly over time. Humans cannot comprehend the loss of purchasing power over decades or centuries, when purchasing power is stable over the course of days, weeks and months. We are conditioned to believe that dollars, euros, yen, etc. today will buy the same amount of energy tomorrow.

The behavior of Bitcoin and the cryptocurrency movement it spawned is utterly pathetic, as is the rest of the universe. Satoshi is a tamperer at heart, their love and rage generating externalities of price fluctuations relative to fiat currencies and pure energy itself. While coin holders claim to accept this volatility with conviction, we are only human. Sometimes, we stray from the path of gold. In times of trouble, the golden bull of stablecoins looms before us, and we can easily find ourselves swayed by its sweet melodies. But what many fail to appreciate is that they are fundamentally incompatible with the financial world we hope to create.

Many people ask me my opinion on this or that stablecoin. The recent volatility surrounding the $1 peg of Terra’s USD stablecoin UST prompted me to start a series on stablecoins and central bank digital currencies (CDBCs). These two concepts are related to the fundamental nature of the debt-based banking system that dominates the global financial system.

This article will explore several broad categories of blockchain-backed stablecoins, including fiat-backed, cryptocurrency-overcollateralized, algorithmic, and bitcoin-backed stablecoins. While there is no perfect solution yet, the final section of this article will address my thoughts on the best way we can currently combine these two incompatible systems: a Bitcoin-backed, dollar-pegged A stablecoin is also an ERC-20 compatible asset.

In this period of heightened downward volatility in markets, the only consolation is that we are alive. Our actions and thoughts are not entirely in control, but we must inhale and exhale slowly, methodically, and mindfully. Only then can we continue the preaching of Bitcoin.

Stablecoins anchored to fiat currency

As I explained earlier, a bank is a public utility that operates a fiat value transfer network. They help individuals and organizations conduct business. Before the Bitcoin blockchain, banks were the only trusted intermediaries that could also perform these functions. But even with Bitcoin, banks are still far and away the most popular intermediaries, enabling certain banks to engage in reckless behavior because they believe governments can print money to nominally bail them out for their indiscretions.

Banks charge a very substantial tax in terms of time and fees to transfer value between participants. Given the near-instant and free encrypted means of communication we now have access to, there is no reason why we have to pay so much money and waste so much time paying each other.

The Bitcoin blockchain creates a competitive peer-to-peer payment system with low time and monetary costs. The problem for many is that Bitcoin's native asset is very volatile when benchmarked against fiat currencies and energy (i.e. a barrel of oil). To solve this problem, the folks at Tether created the first dollar-pegged stablecoin using the Omni smart contract protocol built on top of bitcoin.

Tether created a new digital asset class that is backed 1:1 by fiat assets in banking institutions on a public blockchain, what we now call fiat-backed stablecoins. After Tether (also known as USDT) and USDC, various other fiat-backed stablecoins have sprung up, and the fiat assets held by each project have also expanded as interest in these coins has grown. Currently, USDT and USDC together have more than 100 billion US dollars in fiat currency assets.

Because there is no real farm-to-table bitcoin economy, we still pay for most things in dollars or other fiat currencies. Since the traditional way of sending and receiving fiat currency is very expensive and complicated, the ability to instantly send the other party's fiat currency value at a lower cost bypassing the expensive bank payment system is very valuable. I would rather send USDT or USDC to others than try to fumble through the cumbersome and expensive global fiat currency bank payment system.

The fundamental problem with this type of stablecoin is that it requires a willing bank to hold the fiat assets backing the token. None of the stablecoin transaction fees end up in the pockets of bankers, but there are costs for banks to hold these huge fiat assets. The destruction of the time value of money by central bankers, as we know it, completely shattered the lending business model of commercial banks, which led them to agree to hold billions of dollars for agreements designed to unravel their relationship without identifiable For the benefit, it's an odd policy.

Fiat-backed stablecoins will not be the payment solution that powers Web3 or a truly decentralized global economy. They will not be allowed to grow large enough to properly serve an internet-connected brick-and-mortar world that requires fast, cheap and secure digital payments. We saw this uncoordinated action when the Federal Reserve banned Silvergate Bank from being the custodian for Facebook’s Diem stablecoin. If Diem launches, it will immediately become one of the largest circulating currencies in the world due to Facebook's user base. It will compete directly with various fiat currencies that maintain bank equity, which is not allowed to happen.

This fundamental problem has not disappeared in the cryptocurrency industry. The next iteration of stablecoins is a family of projects that overcollateralize major cryptocurrencies in order to peg their value to fiat currencies.

Cryptocurrency-backed over-collateralized stablecoin

Simply put, these stablecoins allow participants to mint fiat-pegged stablecoins against cryptocurrency as collateral. The most successful of these stablecoins is MakerDAO.

MakerDAO has two currencies. Maker (MKR) is the token that governs the system. This is similar to a stake in a bank, but the bank aims to have more assets than liabilities. These assets are various major cryptocurrencies such as bitcoin and ether, which promise to create a dollar-pegged token called DAI.

1 dai = 1 usd

You can borrow DAI from MakerDAO with a certain amount of cryptocurrency as collateral. As the price of cryptocurrency collateral may drop, Maker will systematically liquidate collateral to satisfy loans in DAI. This is done transparently on the Ethereum blockchain. Therefore, it is possible to calculate the level at which Maker must liquidate the position.

Here is a graph of the percentage deviation of DAI from its peg to $1. A reading of 0% means that DAI has held its peg perfectly. As you can see, Maker has done a good job of keeping its hooks.

The system is robust because it has withstood various price crashes in Bitcoin and Ethereum, and its DAI token still maintains a value close to $1 on the open market. The downside of this system is that it is over-collateralized. It effectively removes liquidity from crypto capital markets in exchange for stability pegged to fiat assets.

MakerDAO and other over-collateralized stablecoins completely drain liquidity and collateral from the wider ecosystem when they reach their f(x) maximum. Maker token holders can choose to introduce risk into the business model by lending out idle collateral in exchange for greater income. However, this introduces credit risk into the system. Who are reliable borrowers who can reliably pay positive interest rates on cryptocurrencies, and what collateral do they pledge? Are these collaterals reliable?

The benefit of fiat proportional banking is that the system can grow exponentially without draining the economy of all good collateral for money. These over-collateralized stablecoins fill a very important niche, but they will always be a niche for the fundamental reasons above.

The next iteration of stablecoins aims to completely remove ties to any "hard" collateral and be backed only by fancy algorithmic minting and burning schemes. In theory, these algorithmic stablecoins could scale to meet the needs of a global decentralized economy.

Algorithmic Stablecoins

The stated goal of these stablecoins is to create a pegged asset backed by less than 1:1 crypto or fiat. Typically, the goal is to back the pegged stablecoin with zero external "hard" collateral.

Given that Terra is a hot topic, I will explain the mechanics of algorithmic stablecoins using LUNA and UST as examples.

LUNA is the governance token of the Terra ecosystem.

UST is a USD 1-pegged stablecoin whose "asset" is simply the LUNA tokens in circulation.

Here's how the hook works.

Expansion: If 1 UST = $1.01, then UST is overvalued relative to its peg. In this case, the protocol allows LUNA holders to exchange 1 UST for 1 USD worth of LUNA. LUNA is burned or withdrawn from circulation, and UST is minted or enters circulation. Given that 1UST = $1.01, the trader makes a profit of $0.01. This has a pro-cyclical effect on the price of LUNA as its supply decreases.

Contraction (where we are now): If 1 UST = $0.99, then UST is undervalued relative to its peg. In this case, the protocol allows UST holders to exchange 1 UST for 1 USD worth of LUNA. Given that you can buy 1 UST for $0.99 and exchange it for $1 worth of LUNA, you make a profit of $0.01. UST was burned, and LUNA was minted. This acts as a pro-cyclical amplifying effect on LUNA’s price decline, as its supply increases along the way.

The big problem is that investors who now own newly minted LUNA will decide to sell it immediately rather than hold it and hope the price goes up. This is why there is constant selling pressure on LUNA when UST is trading at a significant discount to its pegged price.

LUNA is considered more valuable as UST is used more in commerce throughout the Web3 decentralized economy. This minting and burning mechanism is great during the up period, or when UST is getting more and more popular, but if UST cannot maintain its peg during the down period, then a death spiral will start and LUNA will be minted infinitely , in an attempt to bring UST back to its pegged parity.

All algorithmic stablecoins have some mint/burn interplay between the governance token and the pegged stablecoin. All of these protocols have the same problem of how to attract people to support the pegged transaction when the pegged stablecoin is pegged to a fiat currency.

Almost all algorithmic stablecoins have failed miserably due to the death spiral phenomenon. If the price of the governance token falls, then the governance token assets backing the pegged token will not be considered credible by the market. At this point, participants start dumping their pegged tokens and governance tokens. Once the upward spiral begins, it is very expensive and difficult to restore market confidence. Then there is a pool of stagnant water.

Death spirals are no joke. This is a total confidence game. This is a confidence game similar to the current partial debt-based banking system; however, this game has no governments who can coerce the system with deadly threats of violence.

Theoretically, profit-seeking humans should be willing to forego good collateral in order to save algorithmic protocols for huge profits denominated in governance tokens recently created out of thin air. That's at least a hypothesis.

Here is a chart of the percent deviation of UST from its peg to $1. Like MakerDAO, 0% means the peg is rock solid. As you can see, everything was fine until it changed. UST continues to trade at a steep discount relative to its peg.

Again, many have tried, but most have failed or are in the process of failing. That's not to say the model won't work, at least for a while. I happen to own some governance tokens of a specific algorithmic stablecoin project. They are currently profitable at the protocol level, which makes them attractive. The protocol is structured similarly to Terra, but accepts other “harder” collateral to back its governance token.

In theory, this model, similar to proportional banking, could scale to meet the needs of a decentralized Web3 economy, but it requires near-perfect design and execution.

Bitcoin-backed stablecoins

The only laudable goal of stablecoins is to allow fiat-pegged tokens to run on public blockchains. This has practical implications until a true farm-to-table bitcoin economy arrives. So let's try to make the most of this fundamentally flawed premise.

The most primitive cryptocurrency collateral is Bitcoin. How do we convert a Bitcoin to USD 1:1 value ratio into a hard-to-break USD stablecoin?

Various top cryptocurrency derivatives exchanges offer inverse-style perpetual swaps and futures contracts. The underlying object of these derivatives contracts is BTC/USD, but BTC is used as margin. This means that profits, losses and margins are denominated in Bitcoin, while quotes are denominated in USD.

Each derivatives contract is worth $1 in Bitcoin at any price.

Contract value Bitcoin value = [1 USD/BTC price] * number of contracts

If BTC/USD is $1, then the contract is worth 1 BTC. If BTC/USD is $10, then the value of the contract is 0.1BTC.

Now let's synthesize $100 using BTC and a short derivatives contract.

Suppose BTC/USD=100 USD.

How much is 100 contracts or $100 worth of BTC at $100 BTC/USD?

[$1/$100] * $100 = 1 BTC

Intuitively, this should make sense.

100 Synthetic USD: 1 BTC + 100 derivative short contracts

If the price of Bitcoin rises to infinity, the value of the short derivatives contract calculated in Bitcoin will approach the limit of 0. Let's demonstrate this with a BTC/USD price that is larger but less than infinity.

Suppose the price of Bitcoin rises to $200.

What is the value of our derivatives contracts?

[$1/$200] * $100 = 0.5 BTC

Therefore, our unrealized loss is 0.5BTC. If we subtract our unrealized loss of 0.5 BTC from our 1 BTC of collateral, we now have a net balance of 0.5 BTC. But at the new BTC/USD price of $200, 0.5 BTC is still equal to $100. So we still have $100 of synthetic dollars, even if the price of Bitcoin rises and causes an unrealized loss on our derivatives position. In fact, mathematically speaking, it is impossible for this position to be liquidated on the rise.

The first fundamental flaw in this system occurs when the price of BTC/USD approaches zero. As the price approaches 0, the contract value becomes larger than all the bitcoins that will ever exist, making it impossible for the short side to pay you back in bitcoins.

Here's the math.

Suppose the price of bitcoin falls to $1.

What is the value of our derivatives contracts?

[$1/$1] * $100 = 100 BTC

Our unrealized gain is 99 BTC. If we add this unrealized gain to the original 1 BTC collateral, we come up with a total balance of 100 BTC. At a price of $1, 100 BTC equals $100. So our 100 synthetic dollar peg is still in effect. However, notice how a 99% drop in the price of Bitcoin produced a gain of 100 times the Bitcoin value of the contract. This is the definition of negative convexity, showing how this peg is broken when the price of Bitcoin approaches 0.

The reason I don't consider this scenario is that if Bitcoin goes to zero, the whole system doesn't exist. At this point, no public blockchain capable of transferring value exists anymore, because miners are not expending pure energy maintaining a system where the native token has no value. If you're worried that this might happen, just stick to fiat currency.

Now, we have to introduce some centralization, which brings a whole host of other problems to this design. The only places where these inverse contracts are large enough to accommodate a bitcoin-backed stablecoin and serve the current ecosystem are on centralized exchanges (CEXs).

The first point of centralization is the creation and redemption process.

Creation process:

1. Send BTC to the Foundation
2. The foundation pledges BTC on CEXs and sells reverse derivatives contracts to create sUSD, the synthetic dollar
3. The Foundation issues sUSD on the public chain. For ease of use, I recommend creating it as an ERC20 asset

Foundations must create an account at CEXs to trade these derivatives. The BTC collateral is not kept in the foundation, but in the CEX itself.

Redemption process:

1. Send sUSD to the Foundation
2. The foundation buys back the reverse short contract in CEXs, and then destroys sUSD
3. The foundation withdraws the remaining BTC collateral and returns it to the redeemer
   

There are two problems with this process. First, the CEX (for whatever reason) may not be able to return all the BTC collateral entrusted to it. Second, CEX must collect margin from losers. As far as this project is concerned, when the price of BTC falls, the derivatives of the project are in a state of profit. If the price falls too much, too fast, CEX will not have enough long margin to pay. This is where various socialized loss mechanisms come into play. TL;DR, we cannot assume that if the BTC price falls, the project will get all its fair share of BTC profits.

further steps

The foundation needs to raise funds for the development of the project. The biggest funding requirement is a general sinking fund to cover the exchange's counterparty risk. Governance tokens must initially be sold in exchange for Bitcoin. This bitcoin is designated to deal with the situation that CEX cannot pay as scheduled. Obviously this sinking fund is not inexhaustible, but it will create some confidence that if a CEX returns less than it deserves, the $1 peg can be maintained.

The next step is to determine how the protocol will generate revenue. There are two sources of income.

1. The protocol will charge a fee on every creation and redemption.
2. The protocol will earn a natural positive basis of the value of the derivatives contract and the underlying spot. let me explain.

The stated policy of the Federal Reserve (and most other major central banks) is to inflate their money by 2% per year. In fact, since 1913, the year the Fed was founded, the dollar has lost over 90% of its purchasing power when pegged to the CPI basket.

BTC has a fixed supply. As the value of the denominator (USD) grows, the numerator (BTC) stays the same. This means that our attribution of the future value of the BTC/USD exchange rate should always be higher than the spot value. So at a fundamental level, contango (futures price > spot) or funding rate (on perpetual swaps) should be positive, meaning those shorting these inverse derivative contracts have income.

One might counter that U.S. Treasuries have positive nominal yields, while nominally there is no risk-free instrument priced in Bitcoin. So it is not correct to assume that the US dollar will depreciate against Bitcoin in the long run. While this is true, as I and many others have written, negative real interest rates (i.e. when the nominal risk-free Treasury rate is below GDP growth rate) are the only way the US can mathematically repay debt holders in nominal terms .

Another option is to push the population growth rate far beyond 2% per year, which would require couples to abandon contraception and other methods of family planning on a large scale. According to the U.S. Census Bureau, the population growth rate in 2021 is 0.1%. If immigrants were not included, the ratio would be negative.

The final option is to discover some new amazing energy conversion technology that drastically reduces the cost of energy per dollar of economic activity. Neither of these alternative solutions seems likely to materialize anytime soon.

Long Bitcoin vs short inverse derivatives contracts should have positive returns year after year. Therefore, the greater the float in sUSD, the more Bitcoin is managed against short derivatives contracts, resulting in substantial compounding interest income. This provides a pool of funds over which governance token holders have proxy rights.

perfection is impossible

The act of creating a fiat-pegged stablecoin running on a public blockchain is not possible without many compromises. It is up to the users of the relevant solutions to decide whether these compromises are worth making fiat currencies run faster and cheaper on public blockchains than on centralized payment networks controlled by bankers.

Of the four proposed, I like bitcoin and derivatives-backed stablecoins the most, followed by overcollateralized cryptocurrency-backed stablecoins. However, each of these solutions takes cryptocurrencies as collateral. I mentioned earlier that these public networks require assets to flow between parties in order to generate transaction fees and pay for the maintenance of the network. Simply holding is toxic in the long run. So let’s not be complacent, but keep working hard to create a farm-to-table Bitcoin economy.

Will Terra/UST survive?

Terra is currently in the deepest dark depths of a death spiral.

When the market cap of UST equals the market cap of LUNA, the upward spiral will stop. If left unchecked, the protocol will find an equilibrium of market capacity. So the question is, what is the final static market value. Most importantly, when new LUNAs are created by arbitrageurs buying cheap UST, who will buy these LUNAs? Why would you buy from people who are selling LUNA when you know there will be billions of dollars of pressure to sell LUNA as long as UST is < $1?

Even if LUNA and UST survive this incident, in the long run, there must be some genius protocol changes to strengthen market confidence, and the market value of LUNA will always exceed the float of UST. I don't know how to achieve this.

This is a chart of [UST Market Cap - LUNA Market Cap]. When this value is < $0, the system is healthy. Spiking upwards means UST must be burned and LUNA issued to bring UST back to the state it was pegged to.

Algorithmic stablecoins are not much different from fiat debt-backed currencies, except for one key factor. Terra and others like it cannot force anyone to use UST at any price. They have to convince the market with their fancy design that the governance tokens that support the protocol will have a non-zero value that rises faster over time than the number of fiat-pegged tokens issued. However, a government can always force, eventually at gunpoint, its citizens to use its currency. Therefore, there will always be an inherent demand for fiat currency, even though everyone knows that the "assets" backing this currency are worth less than the currency in circulation.

epilogue

In a real crash, the market looks for indiscriminate sellers and forces them to sell. This week's plunge was exacerbated by being forced to sell all Luna Foundation Bitcoins to defend the UST:USD peg. As usual, they still failed to hold the anchor price.

I sold my Bitcoin $30,000 and Ethereum $2,500 June puts. I trade mostly because I like it. I haven't changed my structured long positions in cryptocurrencies even though they are losing "value" in terms of fiat currencies. However I am evaluating the various altcoins I own and increasing my exposure.

I did not expect the market to break through these levels so quickly. This crash came less than a week after the Fed raised interest rates to an expected 50bps. Let me repeat a related fact: the market expected a 50bps rate hike and still threw up afterwards. This market cannot handle rising nominal interest rates. It amazes me that anyone would believe that historically high long-term risky assets would not succumb to rising nominal interest rates.

The U.S. April CPI annual rate was 8.3%, lower than the previous 8.5% annual rate. 8.3% is still too hot for the Fed in firefighter mode to give up their pursuit of inflation. A 50bps rate hike in June is expected, which will continue to destroy long-term risk assets.

Now, the crypto capital markets must determine who has excessive exposure to any Terra-related assets. Any service that offers above-average yields and is deemed to be at any risk from the drama will experience rapid outflows. Given that most people have never read how these protocols work in distress, it will be a sell-before-read process. This will continue to drag down all cryptoassets as all investors lose faith in fiat holdings.

After the bloodletting is over, the crypto capital markets must be allowed time to heal. Therefore, it would be foolish to try to fumble for legitimate price targets. But I will say that given my macro view of the inevitability of eventual more money printing machines, I remain a firm believer in Bitcoin.

So, I'd buy Bitcoin at $20,000 and Ethereum at $1,300. These levels roughly correspond to their all-time highs during the 2017/18 bull market.