Algorithmic Stablecoins Breakdown: Attempting to Fulfill Satoshi’s Dream
When Satoshi first published the Bitcoin whitepaper to the world, he titled it “Bitcoin: A Peer-to-Peer Cash System” and although Bitcoin would evolve to adopt the digital gold narrative instead, the dream of creating a fully functional digital cash system has remained. Stablecoins, defined as a subset of cryptocurrencies that are pegged to a specific …
By: David LiebowitzDeFi 101
When Satoshi first published the Bitcoin whitepaper to the world, he titled it “Bitcoin: A Peer-to-Peer Cash System” and although Bitcoin would evolve to adopt the digital gold narrative instead, the dream of creating a fully functional digital cash system has remained. Stablecoins, defined as a subset of cryptocurrencies that are pegged to a specific asset and have some kind of reserve backing it, is a continuation of that dream.
The total stablecoin supply is over $100 billion and over 4.3 million wallets hold at least USDT or USDC. Stablecoins are primarily used today by traders to safely store value during times of volatility but have seen adoption beyond trading. Many use USDT as remittances and Visa announced it would settle transactions in USDC.
Today multiple different types of stablecoins exist, each with their own advantages and drawbacks. Fully collateralized stablecoins such as USDT and USDC are backed 1:1 with the US dollar and are by far some of the most liquid assets in crypto. According to CoinGecko, USDT and USDC make up nearly 80% of the entire stablecoin market. Users can seamlessly go from USDC to dollar fiat at a 1:1 ratio on platforms such as the juggernaut exchange Coinbase or DeFi-friendly Dharma.
Yet, these types of stablecoins are seen as a step in stablecoin development and not the final goal due to their centralized nature. The organization overseeing each centralized stablecoin has the ability to blacklist addresses or potentially become a victim to an unforeseen attack, making them a point of failure for the entire system. According to DUNE analytics, USDT has banned 468 addresses while USDC has banned 8 addresses.
Multi Collateral DAI
Another type of stablecoin that have proven to be popular are overcollateralized stablecoins such as DAI and sUSD. They offer a crypto-native alternative to centralized stablecoins and have achieved billions of dollars in circulation.
The Maker protocol itself has been around since 2014 with the mission of creating a permissionless credit system via collateralized loans and is arguably one of the first DeFi projects to launch on ETH. Loans are created through smart contracts as users open vaults in order to mint DAI, a stablecoin that is pegged to the US dollar.
At first, ETH was the only collateral accepted in its Single Collateral Dai (SCD) iteration that launched in December 2017. In November 2019, DAI upgraded to Multi Collateral DAI and added BAT as its next accepted collateral. Today, DAI accepts a plethora of collateral types including WBTC, USDC, YFI, LINK, TUSDT, PAXUSD, AAVE, UNI, renBTC, MANA, 0x, GUSD, etc. as well as UNI V2 LP pairs such as DAI/USDC, ETH/USDT, DAI/ETH, and more.
Furthermore, MCD made additional design changes including the introduction of the Dai Savings Rate (DSR) which gives the opportunity for Dai holders to earn yield and collateral auctions that create a competitive market for liquidations via external keepers.
Yet, DAI in its current form is capital inefficient and is prone to drastic price shocks if the system were to face overwhelming pressure. In fact, in March 2020 Maker DAO systematically failed resulting in 5.67 million DAI in losses. This was not a result of lack of collateral, but rather manipulation by malicious keepers who were able to take advantage of a late update to Maker’s Oracle Security Module. By the time the price in the protocol was updated, a cornucopia of opportunities to liquidate the positions had appeared.
SUSD is a slightly different system than DAI with different objectives. It is the official stablecoin of the Synthetix ecosystem. Anyone can acquire SUSD by minting it via staking SNX or buying it on the market on Curve or Uniswap. In turn, SUSD can be used to perform actions such as buy synthetic assets on Kwenta or invest in trader portfolios on dHedge. When staking SUSD, the user must be mindful of both their collateral ratio and obligations to pay back the loan. SUSD is debt and as the value of all circulating Syntheix Synths such as sETH, sBTC, sTSLA, etc. increase, so does the debt obligation and vice versa.
Similar to DAI, SUSD can only expand so much due to its overcollateralization requirements. Users must stake their SNX at a 500% ratio in order to mint SUSD. The supply of SUSD can only meet so much demand.
With the faults of fully collateralized and overcollateralized stablecoins in mind, how do algorithmic stablecoins present themselves as not only a viable alternative but to eventually supersede other stablecoins as the de facto pegged asset standard?
Algorithmic stablecoins have offered a way to achieve Satoshi’s dream of digital cash, as well as the intellectual framework needed to incrementally make it a fixture in financial infrastructure. One such part of this framework was Robert Sams’ groundbreaking paper on seigniorage shares describing a two token system consisting of one pegged asset and one speculative asset. The goal of the seigniorage shares is for the pegged asset to maintain stability even through times of duress and does this through the speculative asset not only absorbing the volatility of the system, but profiting from it as well.
One of the earliest and most infamous attempts at creating a seigniorage share-like model was Basis which would raise over $130 million in 2018, only to shut down due to “regulatory concerns.” Two years later, a Cambrian explosion of stablecoin experiments such as Basis Cash, Dynamic Set Dollar, and ESD v1, yet all failed to maintain its peg due to lack of demand for their peg assets after their initial launches.
Two interesting stablecoin models to look at are Terra and Frax, both are two token systems but achieve stability in different ways.
Terra is a Proof of Stake blockchain built using Tendermint that focuses on global payments and settlement. Terra has built a burgeoning ecosystem of DeFi protocols including MIRROR which offers synthetic assets, Anchor which acts as the “Federal Funds Rate” for the blockchain world. Although Terra hosts a number of pegged assets including those denominated in South Korean won, Mongolian tugrik and the International Monetary Fund’s Special Drawing Rights basket of currencies, by far the most widely used one is its Dollar-pegged asset, TerraUSD.
Consensus and security is achieved through the LUNA token and absorbs the volatility of the stablecoins. When TerraUSD is priced at $1.01, the system expands to print more TerraUSD through burning LUNA. The idea is that LUNA holders can exchange their for $1 worth of TerraUSD, then sell it for $1.01, and profit from the difference. Part of the LUNA exchanged is burned while the rest of it compounds in a community pool. When TerraUSD is $0.99, the opposite happens, and TerraUSD is exchanged for $1 worth of LUNA.
Terra’s growth is hard to ignore. The ecosystem has a market capitalization of $4.7 billion and TerraUSD has over $2 billion circulating across Terra, Ethereum, and Binance Smart Chain. The native MIRROR protocol has over $1.9 billion in TVL and regularly achieves tens of millions of dollars in volume with users trading synthetic assets such as mBTC, mTSLA, mCOIN, and more. Terra’s lending protocol Anchor has been able to garner $294 million in UST deposits and has lent $95 million of it to borrowers.
Yet in May 2021, LUNA faces its biggest test due to TerraUSD falling below its peg for several days due to a combination of the LUNA price crashing and a number of collateral liquidations in Anchor. At one point, the stablecoin reached $0.92, but has since recovered. In a post-mortem tweet thread, Terra stated that it is “designed with explicit, real-time levers to combat the negative effects of endogenous collateral models (increasing tax rate on txs + cashflows to stakers) that traditional banking models cannot match.”
Frax which started at the same time as other seigniorage share tokens has created a hybrid approach to the concept. At its core, it’s a seigniorage model yet in its infancy decided to be partially backed by USDC. Currently, FRAX is 85% backed by USDC with the remaining 15% being algorithmically stabilized by Frax Shares (FXS). FRAX is continually minted and burned by arbitrageurs in order to maintain its $1 peg which to date has never shifted off of it. According to Page 26 on CoinGecko’s Q1 2021 report, the average deviation has been 0.1%.
Just as central banks perform actions to help stabilize their currency, Frax has developed a system of AMOs (automated money operator) controllers to help maintain FRAX’s peg. These AMOs are an autonomous contract(s) that enact arbitrary monetary policy so long as it does not change the FRAX price off its peg. This abstracts Frax’s stability operation at a protocol level rather than dealing with tokens themselves.
An example of this can be found in the differences between how DAI and FRAX can utilize Curve. MakerDAO can decide to approve CRV or CRV LP tokens as collateral which can be used to stabilize and overcollateralize DAI. Frax takes it a step further and utilizes Curve’s protocol functionality via the FRAX-3pool to stabilize FRAX and maintain its peg.
FRAX is going with the multi-chain approach for adoption. Currently, FRAX is circulating on Ethereum, Binance Smart Chain, Polygon, Fantom, Avalanche, and plans to expand to Moonbeam. In addition, FinNexus, a cross-chain decentralized options platform, has a FRAX pool where users can utilize FRAX as collateral. FRAX’s hybrid collateralization model has so far proved successful. Since its launch, FRAX has yet to divert from its $1 peg on Uniswap or Curve.
Some criticisms of FRAX’s model include that since it’s backed 85% by USDC is that it is not actually as algorithmic as it touts it to be and leaves it vulnerable to a central point of failure. In response, Kazemian has said that FRAX is the only stablecoin to become less dependent on USDC while other stablecoins such as DAI have become more dependent on USDC, adding it as collateral in the aftermath of March 2020’s Black Thursday. FRAX has lowered its collateralization ratio from 100% to around 85% since its launch in December. In a series of tweets from January, Benjamin Simon theorized that FRAX could ditch its USDC collateral altogether in favor of using hedged collateral such as a 50/50 combination of Synthetix’s iETH/ETH or on-chain options/futures.
So what is the future for stablecoins and its algorithmic variants? If there is one thing clear, stablecoins will continue to grow in market share. The need for a stable medium of exchange and unit of account is not going away and in the present moment fully-backed stablecoins fulfill this market need. DAI and SUSD serve their respective ecosystems efficiently, but it remains to be seen what their plan is to become more capital efficient.
Algorithmic stablecoins are still in it’s nascent stages of development and have been ripe for experimentation. Although many seasoned crypto-natives doubt the success of algo-stables and failure is an all-to-close of a shadow, those algo-stables which make it past the stress-tests of massive downturns will prove themselves to be lindy and resilient.
As the founder and architect of Ethereum, Vitalik Buterin, stated in his UNI should become an oracle token post, “The goal of algorithmic stablecoins is to try to be maximally censorship-resistant and robust by being free of dependencies to the ‘fiat world’” and used providing data to algorithmic stablecoins as the primary reason why UNI should become an oracle token.
Algo-stablecoins need a robust price feed that is expensive to attack and Vitalik understands how important it is to protect the integrity of them. Furthermore, Mark Cuban suggested in a tweet storm that DAI should be used to pay for gas in the future which makes sense because people, when given the opportunity, would prefer not to spend their ETH.
There are a lot of different elements that go into how to make the perfect stablecoin. How is it designed? What is backing it? How can it be used? What are its attacked vectors? These questions are something that every protocol economic designer needs to ask and decide where to put its priorities.
Different stablecoins serve different markets at different times and have a wide range of goals and objectives over a given time horizon. USDT and USDC are the most accessible here and now, DAI, SUSD, and UST serve their respective ecosystems, and Terra and FRAX are trying to push how far the concept of seigniorage can go. If there is one thing for sure, Satoshi’s dream of a global peer-to-peer cash system is alive and well and when it comes to fruition, the ramifications could very well mean as Erik Voorhees puts it, the separation of money and state.
David Liebowitz is the Head of Growth at Gelato, a protocol that automates smart contracts on Ethereum and Web3. He was previously the Vice President of Business Development at Everipedia, a blockchain-based encyclopedia. Disclosure: David does hold $FXS tokens.