Do or DAI
Do Kwon's bid for algostable supremacy threatens to sideline DAI
Lots of detailed content has been written summarizing the cold-to-hot Curve War that broke out between Do Kwon (Terra) and Rune Christensen (Maker) with Kwon’s threat —
— and subsequent action, of building an alliance to effectively expel DAI from Curve.
The question is, is this a drama-of-the-month backed mainly by expensive Curve/Convex voting bribes, or a tactical coup that structurally sidelines DAI? Let’s have a look.
The past: A little algostable history
Historically there have been 3 types of stablecoins:
Fiat-collateralized or centralized stablecoins, fully backed by fiat collateral (Tether/USDT, USDC)
Crypto-backed, overcollateralized stablecoins (DAI v1)
Partially-backed algorithmic stablecoins (UST, FRAX, MIM)
Centralized stablecoins, which require a direct line to a US bank in order to function, still make up 80% of the stablecoin market: USDT 45%, USDC 25%, and BUSD 10%.
Within the stablecoin market, however, centralized stables are growing much more slowly than their decentralized algorithmic stablecoin counterparts, primarily UST, DAI, and FRAX. (The former are increasingly called ‘cStablecoins’ and the latter ‘algostables,’ which I will use for the rest of this article.)
The bullish conviction around algostables is driven by a growing belief — turbocharged by the Mareva injunctions of 34 crypto donors to the Canadian Freedom Convoy and an explosion in financial censorship across the West (not to mention more and more governments at odds with the West) — that crypto on cStablecoin rails offers no privacy, no freedom, and no censorship resistance from the fiat world. The same libertarian beliefs that guided us to this new form of money, dictate that increasingly insolvent Western governments will find more and more excuses to refuse to honor more and more liabilities (which is all a currency is).
In short, crypto with cStablecoins is not real crypto, except insofar as cStables are an evolutionary pit stop on the highway out of fiat hell.
DAI, founded by MakerDAO, was the first algostable to truly break out at scale: a smart contract that minted heavily-collateralized DAI as users deposited ETH in its vaults, supported by a system of third-party arbitrageurs who could theoretically make large profits if a poorly-informed public sloppily ran for the exits.
However, if something happens which causes those third parties to lose systemic confidence, the fundamental pegs in the system (for example, DAI-USDC or DAI-USDT) lose support, and the currency ‘loses its peg.’ That happened in March of 2020, when Maker’s vaults slipped into technical default when the price of ETH dropped 30% in one day. As synthetic long ETH / short DAI positions were liquidated en masse, DAI’s value surged above its peg, causing more ETH deposits to slip into default and get liquidated.
Maker and DAI survived the ordeal, but its founder Rune Christensen permanently lost operating confidence in the algostable model. Maker/DAI migrated from a 95% ETH-collateralized algostable to a 50% stablecoin-backed algostable by 4Q20. Today, around 60% of DAI is backed by USDT and USDC.
After surviving the initial brush with death, Maker also reimagined its place in the financial world. Most crypto protocols basically want to build a primitive in the race to usurp the USD’s crown and figure out the cash-on-cash economics later. However, after the March 2020 near-death experience, Christensen seemingly reimagined the Maker DAO as a supremely-efficient neobank running on smart contracts.
Once Maker settled into this new mindset, as a vanilla real-world-asset-based lender running on a superior software and governance backbone, it embraced most of the fiat regulatory process.
A lot of that fiat love-fest was no doubt borne out of an interest in prudent regulatory posturing. It also dovetailed with other un-crypto-like convictions of the Maker DAO, such as a vocal passion about global warming and a broader focus on incorporating ESG frameworks into Maker’s credit allocation.
But as DAI changed from an ethereum-backed algostable to a mostly USDC-backed algostable, and adopted a broader belief set commonly associated with the most coercive forces in modern Western culture, DAI became seen as a sellout to centralization.
Maker’s rejection of the algostable crown, in its defense, had solid grounding in reality. Many algostables have failed. TITAN, a $2 billion algostable backed by Mark Cuban, went from $2 billion to zero in June 2021. LUNA, then just another upstart, very nearly went to zero at around the same time. Literally as I write (4/4/2022), the USDN (Neutrino) algostable, the rail on which the WAVES protocol conducts stablecoin transactions, saw a sustained 20%+ break of its peg, an implosion more severe than UST’s May 2021 depegging death-spiral.
Two algostables in particular (UST, which is basically backed by faith in its LUNA governance token; and FRAX, which started around the same time UST did) stayed true to the algostable vision, and scaled with some success. A third, MIM, rose and fell with the profile of Daniele Sestagalli.
As Terra rose, Rune Christensen increasingly attacked Terra’s “ponzu”1 customer acquisition strategy and highly promotional, personality-driven protocol marketing. Christensen had learned the hard way that in deposit-rich DeFi, borrowers — not depositors, TVL, or other more familiar metrics — are king. In my view, his suspicion of TVL-chasing was not just selfish, it was thoroughly grounded in reality.
But as much as Christensen oozed contempt for Terra, the wider crypto community cheered Kwon’s willingness to risk everything to build a successful algostable where so many others had failed. In Kwon’s pugnacious marketing style, his shadowboxing with US fiat regulators, his single-minded libertarian-ization of the stablecoin industry, and his indifference to a loud chorus of naysayers, many top crypto practitioners are reminded of Elon Musk.
At the same time, as 2022 became an ongoing advertisement for crypto payment rails, DAI’s embrace of USDC — seen by many as a privatized or prototypical Federal Reserve stablecoin — seemed increasingly out of step with the winds of change.
Curve War III: The algostable alliance expels DAI from Curve
(If you find the history and mechanisms of crypto-bribery fascinating and want to learn more, I recommend Luca Prosperi’s analysis of the recent Curve Wars.)
Do Kwon’s “we pretty much own the CVX” declaration-of-war announcement was shorthand for saying the following:
“We” — an alliance of past and future algostable “ponzus” who’re also CRV holders (mainly UST, FRAX, and Redacted Cartel, aka OHM) — own a critical mass of CVX tokens which can control the CVX DAO, and therefore control the CRV DAO.
CVX is basically the dominant political party in the Curve shareholder-democracy. If you control enough of the majority political party (CVX), you control the whole country (Curve).
Kwon and his coalition voted to remove all CRV subsidies from the DAI-3Pool on Curve. This will cause DAI volume to shift further to Uniswap v3 (which is already basically equal to Curve in stable-to-stable market share).
But Kwon & Co only control 12% of CVX tokens. How does that get you to 50% of CVX?
Actually, you don’t need 50% of CVX. Just like IRL democracy, you only need 50% of people who actually vote. However, even if, say, only 40% of hodlers have enough skin in the game to vote on something like this (making 20% of all CVX tokens the breakeven requirement), 12% is still not nearly enough.
Kwon & Co have corralled the additional votes through the Votium dapp, a market for CVX bribery in which major Curve players bribe CVX holders for their biweekly votes.
We can see that FRAX and UST (the leftmost two columns) paid $13m in native-token emissions to corral the additional votes. The $5m for UST is substantial, but probably not too different from what UST usually spend to consistently defend their Curve incentives. The $8m that FRAX pays — for two weeks of subsidizing the new Curve 4pool at the expense of the DAI-3pool — otoh, is enormous. On an annualized basis, FRAX is spending $210m of FRAX emissions; they have a market cap of $2.8B.
The votes occur every two weeks, so FRAX and UST are paying $13m * 26 = $338m annualized for the right to control around $270m in annualized CRV/CVX emissions. Either CVX/CRV need to go up in value, or the amount spent on bribes must come down. Or both. In any case, being bribed in crypto hasn’t felt this rewarding in quite a while.
It’s hard to see how FRAX could spend that kind of money in the real world in a way that would brand themselves as a major player to existing key players in the crypto industry, ‘rent’ market share from a competitor, and create significant additional demand for their stablecoin — although if FRAX don’t ultimately have demand use-cases for their stablecoin, their Curve promotions will indeed be another form of ‘ponzu’ marketing, lining the pockets of CVX holders.
Meanwhile, the large population of indifferent CVX tokenholders can just allocate their CVX to Votium and enjoy being bribed.
While everyone always talks about Kwon’s pugnaciousness, Do Kwon has actually been a booster of competitor algostables for a while now. MIM, for example, would never have existed without leeching off the Terra ecosystem.
While Kwon is happy to hurt DAI, he didn’t want to pay a lot more to dramatically elevate the profile of a direct competitor, FRAX. Kwon seemingly doesn’t view algostables as a winner-take-all market, and probably believes other algostables supporting each other’s liquidity is bullish for systemic stability, especially in times of stress where his own algostable will naturally take share from weaker algostables.
Kwon saw an opportunity to once again play The Guy Who Makes Stuff Happen, giving a smaller competitor a bigger platform to give DAI a black eye, at fairly low marginal cost to himself.
The future: In the age of Uniswap v3, are Curve wars dumb now?
Christensen (DAI)’s public line is that he doesn’t care about Curve ponzu games, and if his competitors want to piss more money away bribing Curve to be a worse platform, fine.
On one level, the numbers agree with him. Paying 35%+ annualized rates for liquidity on Curve is dumb when Uniswap v3 and the Maker PSM provide very good alternatives for virtually nothing.
Curve was historically the runaway leading stablecoin dex. In the last year, Uniswap v3 took substantial share in stablecoin-to-stablecoin trading from Curve. Today, Curve still controls around 54% of the share between the two.
Most DAI liquidity migrated to the Maker PSM, a one-to-one, zero-slippage exchange between DAI and USDC, anyway. Most intrinsic decentralized demand for DAI will migrate to Uniswap v3, which has no governance token and thus lacks the very expensive crypto-lobbying that defines Curve today.
Christensen is probably happier to let the bribeoooors ponzu it out on Curve, Convex and Votium, and quit Curve altogether, than join a bribery bidding war that is lining the pockets of Curve LPs and Convex tokenholders.
However, he probably still has a bitter taste in his mouth. Getting de-platformed from Curve is a loss, no matter how you slice it. Ironically, as much as Christensen has mocked Kwon and Terra for being a ponzu merry-go-round, FRAX played the “ponzu” underwriter role for which Christensen so often criticized Kwon in the past.
FRAX, in doing the heavy lifting behind Kwon’s publicity coup, ultimately made an aggressive growth branding bet: if algostables are going to take more share from cStables, FRAX’s investment will be remembered as a brilliant marketing ploy, for buying two key seals of approval (co-author with Terra of a new Curve primitive, and partnership with Terra) at the cusp of a liftoff in algostables amidst a crypto rebound, which incidentally kicks DAI to the Curve curb, as DAI-3pool emissions were killed off.
On the other hand, if algostables are just the latest flavor of ETH alts (I very strongly disagree with this), this is just very inefficient marketing spend.
But unlike The Average Joe, I don’t think Maker/DAI has a strategic interest in fighting this battle. DAI has left the retail / whale / degen world behind for the much ‘greener’ pastures (pun intended) of institutional deposits; fighting back would be a loser’s game of irrational brand advertising for a market that DAI has left behind.
Kwon and his allies thus presented a lousy set of choices for DAI, and Kwon got a smaller frenemy, FRAX, to foot most of the bill for kicking Kwon’s bigger competitor in the shins: not bad for a day’s work, not bad at all.
“Ponzu” or “ponzinomics” refers to the tendency in DeFi to incentivize users through the emissions of new tokens, often of highly questionable value, in the absence any obvious sustainability of those yields after the emission budget had expired.