Maker in a Dilemma, and the Canto Blockchain

Tornado Cash sanctioned, Tornado Cash-linked USDC frozen - Implications for MakerDAO

DeFi’s top protocol by TVL, MakerDAO, has been taking some heat, with the debate on DAI’s decentralization back in the spotlight considering that about a third of its collateral is in USDC, and slightly more than half of all DAI in circulation are generated by USDC collateral.  

DAI collateralization & DAI generated by collateral. Source: https://daistats.com/#/overview 

A widely circulated tweet by Yearn Finance developer @Bantg claimed that MakerDAO was considering a market buy of ETH with the $3.5B USDC from the peg stability module, which was shortly after countered by Vitalik to be a bad idea as it could result in turning MakerDAO into a fractional reserve if ETH market price fell. He further added that he thinks no single type of non-ETH collateral should be allowed to exceed 20% of collateral reserves, and to limit DAI’s growth if necessary (e.g. implementing negative interest rates) until the former is achievable.

For context / a recap, in December 2020 MakerDAO introduced the peg stability module (PSM) - a type of vault implementation which allows users to swap an asset for DAI at a fixed rate (and vice versa), instead of depositing collateral to borrow DAI. The PSM was designed specifically for using other stablecoins as collateral to increase the stability of the $1 DAI peg, as instant arbitrage opportunities would arise if DAI diverged from the price of the PSM stablecoin collateral given the 1:1 swap ratio. PSMs backed by volatile assets, such as ETH, and not pegged to the same asset as DAI would not really work as then DAI would become undercollateralized as soon as the price of the backing asset declines significantly.

Rune Christensen, co-founder of MakerDAO, clarified that what he said was “yoloing all the stablecoin collateral into ETH would be a bad idea”, but suggested that using some of the USDC collateral to execute gradual dollar-cost averaged buys of ETH can be considered depending on the perceived risk of Maker getting blacklisted by the US Treasury Department.

Our Thoughts

In a recent interview, Rune posited that the recent regulatory actions on Tornado Cash could push the market to finally reward true decentralization, and be more tolerant of the tradeoffs incurred (i.e.  less peg stability) in order to have a censorship-resistant on-chain stablecoin.  However this may be somewhat optimistic. As we have witnessed, it is easy for users to fall into complacency and prioritize convenience and perceived ‘safety’ in day to day usage. Even long-standing concerns about USDT reserves had been largely brushed off since its earlier days, and to date centralized stablecoins USDT and USDC remain by far the largest stablecoins on the market. 

However, the fact remains that a sanction on MakerDAO, and inevitably the subsequent freezing of USDC reserves by Circle would immediately make DAI undercollateralized. While people have differing opinions on how likely a sanction on MakerDAO would be, that DAI could suddenly become undercollateralized based on the decision of a single entity seems to be a significant risk for the project. At the very least, immediate steps should be taken to reduce the reliance on USDC or any other centrally-issued collateral which carry similar risks.

1y Price chart of FRAX, LUSD, DAI, MIMATIC, and RAI (on secondary axis). Source: CoinGecko

In light of regulatory concerns, could some of DAI’s competitors step up? If all of DAI’s USDC backing were to be converted to ETH, that would make the stablecoin at least ~64% collateralized by ETH, nudging it closer to Liquity’s LUSD model, which is exclusively collateralized by ETH at a 110% collateral ratio. QiDAO, issuer of the MIMATIC stablecoin, also accepts various crypto collateral similar to Maker but eschews centralized stables (with the exception of some small exposure to USDT/USDC via DAI/3CRV collateral types). Meanwhile, RAI has chosen the path of pure crypto-nativity, being ETH collateralized, but unpegged to any state-backed fiat currencies (similar to how USD is not pegged to anything, but considered stable).  At one glance, it’s clear that all these alternatives are significantly more volatile compared to DAI, with their value frequently oscillating between ±2% range (and some days even more). Such volatility arguably may not be acceptable for a so-called “stablecoin” whose only objective is to preserve stable value.  

Meanwhile, Frax and its fractional reserve model sits quietly at #5 by stablecoin market cap ($1.4B). It is relatively closer to DAI in stability compared to the rest, but is exposed to the same sanction risk as DAI due to the fact that FRAX is essentially 90.5% collateralized by USDC. Safe to say, an acceptable solution that doesn’t rely on a centralized stablecoin being a significant part of the collateral currently doesn’t exist yet. 

It’s also not sufficient to choose just any group of random assets (say 5 evenly distributed) and use that to collateralize a stablecoin. Ideally, the price of these assets should obviously be relatively stable, but their prices should also be uncorrelated to each other. Otherwise, a general downturn will cause multiple collateral types to deteriorate in value collectively. This is obviously very difficult for the current crypto market, where prices of assets generally trend in the same direction as BTC and ETH. That’s why centralized stables are so attractive as collateral for decentralized stables - they don’t move according to the whims of the market. 

Stablecoin models (over-collateralized, centralized, algorithmic) and the stablecoin trilemma (peg stability, capital efficiency, decentralization). Similar to the blockchain trilemma of decentralization, scalability, and security, stablecoin designs have typically had to pick two out of three, and make substantial tradeoffs on the third. Source: https://cepr.org/voxeu/columns/algorithmic-stablecoins-and-devaluation-risk

Perhaps, the decentralized stablecoin scene will pivot towards nascent and more experimental models? UXD is a recently launched USD stablecoin backed 1:1 with delta-neutral derivatives positions, while the more radical Beanstalk claims to overcome the stablecoin trilemma by issuing its BEAN USD-pegged stablecoin with credit and forgoing collateral completely, though it has previously suffered a major exploit last April which it is still recovering from. 

Time will tell if we inevitably converge on some form or other of the three widely adopted models (i.e centralized & issuer-backed, over-collateralized, or uncollateralized/algorithmic/fractional reserve), or manage to design a novel stablecoin mechanism which manages to overcome the intractable stablecoin trilemma of decentralization, scalability, and stability. 

New Chain on the Block - Canto 

Even in the depths of the bear market, innovation and building has continued. Multimillion dollar raises have made headlines over the past months, be it for protocols or new blockchains. However, this time around, the blockchain we’ll be covering is slightly different from the rest, built by a group of developers, and not having raised capital from big venture funds.

Canto is an upcoming layer-1 blockchain which intends to be a hub for decentralized finance (DeFi). It is aiming to achieve these few aspects: accessibility, transparency, and freedom for new systems. To achieve these aims, Canto will be providing free public infrastructure (FPI).

There isn’t much information on the team as it is rather fragmented. Development is mainly done by a community of contributors. Notably though, Scott Lewis is one of the main contributors. Known for starting one of the earliest TVL aggregators in the form of DeFi Pulse, and also having founded the DEX aggregator Slingshot Crypto.

What is Canto?

To put it simply, Canto is similar to Evmos in the sense that it is built using the Cosmos SDK, and utilizes Ethermint. This makes it EVM compatible, and developers are able to deploy code / protocols from Ethereum onto it.

The main feature of the Canto network is its Free Public Infrastructure (FPI) which Canto draws parallels to “free parking on a city street”

By being a blockchain focused on FPI, it will be launching a number of core DeFi protocols with features that will avoid any rent-seeking behavior in the future. For example, the default decentralized exchange of Canto will be launched without a token (none will be released) and there will not be fees. On top of that, it will be ungoverned and cannot be upgraded, ensuring that any adverse features aren’t added. However, in turn this means that Canto will have to figure out other ways to attract user supplied liquidity in order to compensate them in lieu of fees, and any beneficial features in the future cannot be implemented. It should be noted that Canto will be having a Liquidity Provider (LP) rewards program once the DEX launches, in order to reward users for providing liquidity. This comes in the form of CANTO, the network’s native token (more on this later).

As such, it is likely that the DEX will be a rather simple one come launch. At the time of writing, the DEX contracts are live, but a dedicated UI doesn’t exist. Users are able to interact with the contracts via Slingshot, a DEX aggregator platform with support for Canto. Meanwhile, to provide liquidity, users can head here and LP tokens received can be staked in the lending market to participate in the liquidity mining program.  

The tokens: $CANTO & $NOTE

CANTO

CANTO is the native token of the Canto network, and is used to stake with validators in order to secure the network, and also for governance votes. There are a total of 1 billion CANTO, though only a rather small portion has been distributed thus far. Most of this (2% of the supply) has been awarded to early users of the Canto testnet under the “Settlers of Canto” program. 

The vast majority of the supply is being dedicated to medium-term and long-term liquidity mining (80% in total) in order to compensate LP’s on Canto’s zero LP fee exchange. The remaining supply is set to be awarded in grants, as well as to contributors. For a full breakdown of the distribution, check it out here.

NOTE

Described as an over-collateralized currency with a value that rebalances toward $1 through an algorithmic interest rate policy. At first glance, it may be mistaken as an algorithmic stablecoin, but the team at Canto insist it isn't, considering it isn’t pegged to $1. 

NOTE cannot be created / minted, instead it can be borrowed from a smart contract called the “Accountant” from the Canto Lending Market (CLM). In order to borrow some, users will need to post collateral in the form of major crypto assets such as USDC, USDT, ETH, ATOM, and also CANTO.

As a unit of account on Canto, price has to be kept somewhat stable, and that is achieved with varying interest rates. If NOTE is trading under $1, the interest rate to borrow NOTE is raised. This is done to incentivize users to purchase NOTE from the market instead of borrowing from the CLM. On the flipside, if NOTE is trading above $1, interest rates are lowered to encourage borrowing from the CLM, and it being sold on secondary markets. Currently, the interest rate is adjusted in epochs of 6 hours. 

Governance: The EVMOS Model

Canto will be introducing “unified governance” whereby the governance of the entire ecosystem is controlled by holders of the network token. On the majority of other networks, governance of each protocol is determined by holders of protocol-issued tokens. For example, Uniswap governance is decided by UNI holders, while Sushiswap governance is voted upon by SUSHI holders. 

In the case of Canto, governance will solely be in the hands of CANTO holders. This model is replicated from the EVMOS network, though Canto will be taking it a step further and introducing the GovShuttle Module. This is needed to ensure that any proposals which pass governance on the Cosmos SDK can be “shuttled” and verified to the EVM part of Canto. 

What’s next?

Currently, the CANTO token is already available for trading via the Slingshot interface, and it is possible to bridge between mainnet and the Canto EVM. However, the team insists that Canto has not launched yet as there are still a number of minor issues to fix, and not all features are live. 

At the time of writing, these are the latest updates shared by Scott Lewis in the Canto Discord channel. 

The Canto liquidity mining program will likely go live soon in order to encourage users to provide liquidity to the Canto DEX. 

Our Thoughts

1. Canto is a rather interesting layer-1 blockchain, seeing that it did not raise any outside capital. Instead, it intends to be Free Public Infrastructure. However, without a large treasury, it remains to be seen how the project can incentivize developers and users to utilize the chain over a sustained period of time. Particularly for its DEX, it remains to be seen why LP will retain liquidity in Canto if there are no share of fees distributed. 

2. Being rather similar to Evmos, Canto could prove to be a close competitor. There are already many out there drawing comparisons between the two, with some dubbing Canto as “undervalued” given its current market cap.  It’s important to note of course that other than a rektdrop that was chaotic to say the least, Evmos haven’t had much to show for either since launching, especially since the main Nomad bridge was exploited. 

3. Aiming to be a DeFi-centric blockchain also puts Canto in direct competition with Osmosis (the current hub for DeFi in the Cosmos ecosystem). This might be tough in the short-term as Osmosis has been a longstanding pillar in Cosmos, while Canto is just starting out and is heavily dependent on its community for success. Osmosis also follows the more well-followed incentivized-DeFi model, which despite its flaws has proven ability to attract liquidity and users in the short term. How Canto aims to wrest users away from Osmosis will be interesting.

This article was written in collaboration with Shaun Paul Lee. You can follow him on twitter at @ShaunPaulLee.