March 10, 2022: Not What It Once Was ππ
Do falling stablecoin yields spell trouble for Curve?
Longtime Curve community member Crypto Air Gordo took to Twitter to complain about recent changes to Curve.
Now, before you warn me βnot to feed the trollsβ, Gordo has long been willing to help me when Iβve had questions, so Iβm more than happy to return the favor and offer up my two cents when he raises thoughtful questions. So let me offer up my thoughts on the thread in long-form.
Has Curve changed? Thankfully, yes. In fact, if Curve didnβt change, Iβd be worried.
Since Curve launched, everything else has changed significantly. The world into which Curve launched is almost unrecognizable compared with the world today, both in terms of crypto and everything else. Curve evolving to changing conditions is great.
This first part of the thread laments the state of stablecoin returns on Curve. Indeed, the returns for most stablecoins are lower on Curve than they used to be, although this has always been variable.
Nowadays your only hope for stablecoin yield on Curve is the $UST pool, but this pool is always among the first to risk depegging in every market downturn.
Why have yields slipped? A lot of possible reasons, but a simple one is just the sheer size of Curve at the moment. With $18B worth of liquidity across Curve, the protocol is huge. Itβs a bit of a challenge to earn a 20% effective interest rate on this from token emissions. This requires emitting globally $3.6B worth of the $CRV token each year. $CRVβs market cap is under $1B. At this size, itβs difficult to generate significant yield to every pool. Hence we see Curve has shifted to its present model where yield flows where it is most demanded.
Generally speaking, dollarcoin yield has also plummeted a bit across the DeFi market more broadly. Much of the initial bullishness on stablecoins has plateaued across the sector.
Surely, good yield still exists if you hunt for it. Gordo cites PWRD on Curve.
Outside Curve, one large source of yield is Anchor Protocol, via the Luna ecosystem. If youβve been following this for any length of time, youβll note Luna/Terra faces persistent accusations about its ponzinomics.
To date, the protocolβs aggressive growth has held up impressively, defying naysayers. Weβve not really looked closely into the mechanics of the protocol, so we couldnβt really offer an informed opinion about whether or not growth is likely to continue.
Looking at it from a macro level, weβd suggest the bearish case is that it has grown very large and large protocols eventually have to see their yield drop. Nothing can grow at 20% forever or it will grow so large it subsumes the entire global market. To date much of this growth has been the ecosystem finding more VCs willing to dump piles of money onto the fire, which has to run out eventually.
The bullish case is that Terra has managed to sustain this growth despite a Korean government quite hostile to cryptocurrency. Just this past week, political conditions flipped and became a lot more favorable. Also, all else being equal, always bet on Koreans.
Whatever the outcome, thereβs debate on Terra precisely because itβs unclear. In other words, risky. All investors should be well aware of the correlation between risk and reward. Curve gained a lot of traction early on because it was a relatively safe place to park dollarcoins. Nobody should expect high yield and low risk to persist forever.
Continuing, Gordo cites the highest yields come from the riskier v2 pools.
Itβs another point thatβs inarguably true. Where Gordo brings this up disparagingly, I again see this as the natural evolution of Curveβs growth.
I wasnβt following Curve when they issued the $CRV token, but my understanding (OGs can fact check me here) is that the purpose the $CRV token has always been (1) earning trading fees and (2) governance rights. The purpose was not so much to function as a high yield interest account, I think this was more a happy byproduct. The intention of the massive emissions was to decentralize the token as much as possible, so that governance rights could permeate the entire DeFi ecosystem.
The upshot was to serve as a high yield farming option for a time, which in turn attracted the greatest amount of liquidity. However, looking at the long term vision, thereβs not really any capability to forever attract liquidity just through token emissions. It more or less has to increase trading fees over the 300 year time horizon or it breaks down.
If you think of it in terms of supply and demand, Curveβs supply is offering the best trading capabilities among its liquidity. Now its mission is to realize the demand for such trading.
The long-term success of Curve depends on such trading. Stablecoin to stablecoin trading can only get Curve so far. Curve successfully rode this opportunity to bootstrap a phenomenal amount of liquidity. The objective now is the ability to best move among volatilely priced assets β this appears to be the more substantial market to crack and one well suited to Curveβs algorithms and liquidity.
Only recently have Curve v2 pools started proliferating. Each new v2 pool that gains significant liquidity helps Curve to build more advanced routing capabilities and increase the power of this network. As somebody who is rooting for the long term success of Curve, I would absolutely hope the significant majority of Curve emissions goes to bootstrapping these pools instead of reinforcing existing stablecoin pools. The v2 pools are riskier bets for liquidity providers, so I definitely prefer to overcome any hesitation with the juiciest rewards.
Fortunately, the incentives have been well architected here, as youβd expect from anything architected by Michael Egorov. Most new v2 pools feature a protocol token that wouldnβt have had a natural fit in v1. For v1 pools, the greatest reason a protocol would want to incentivize liquidity was to maintain a peg, ie between cvxCRV/CRV or ETH/stETH. Certainly still useful, but not a problem every protocol needs to solve.
For v2 pools, thereβs more greater need to allow a user to trade to or from this token with a more commonly available token, with the price automagically adjusting to supply and demand. Itβs therefore little surprise we see so many of v2 pools pairing an ERC20 token with ETH. Thereβs sort a natural utility in bootstrapping a large reservoir where users can reliably convert to and from these tokens using native ETH.
To be successful at bootstrapping liquidity, the protocol has to be able to make the case to LPs that their token has a shot at outperforming ETH (with some bonus $CRV emissions to smooth over any uncertainty and impermanent loss). Sure enough, we see most of the successful v2 pools happen to be a cadre of really strong tokens.
Continuingβ¦
I deliberately saved my thoughts on FRAX for this section, because I feel itβs a really good case study. If youβre looking to make a bet on dollarcoins, Iβd study the FRAX empire first. The fractionalized stablecoin has proven very resilient through multiple downturns. They continue to launch amazing products such as their inflation-adjusted stablecoin. 0xAlunara via the Convex Discord sums up the value prop well:
FRAX is killing it. FRAX has several successful Curve pools, including its FRAX-3CRV pool with a whopping $3 billion, as well as newer v2 pools in FXS/cvxFXS and FXS/ETH. FRAX is mopping up Convex voting power to the point where it can direct emissions wherever it likes. At the moment, FXS/ETH is the opportunity.
Certainly this just an effect of small sample size β thereβs just a million in liquidity (versus cvxFXS/FXS at $100MM and $3b in the FRAX/3CRV pool). Still, FRAX has the dry powder to do pretty much whatever it wants around the Curve ecosystem, so Iβd consider this one of the more promising v2 pools for the long term.
Finallyβ¦
Iβd agree that the farmer value proposition has changed. Earning free money is never necessarily going to be easy.
I disagree that the farmer UX is similar to basic DEXβs though. The Uniswap farming experience is rather poor IMHO, and most farmers end up losing money because of the mechanics. Uniswap got a lot of mileage out of their v2 formula simply from first mover advantage, but it has real drawbacks. The v3 experience is even tougher.
In contrast, the Curve v2 farming experience was exquisitely well crafted. Granted, Curve v2 pools are riskier than v1 pools, so if youβre looking for zero risk and high yield then youβll be disappointed.
For those willing to take more risk, the capability of farming passively is wonderful. Iβm poor, so I canβt always afford to rebalance my positions constantly, especially when gas prices spike. Of course, gas prices always tend to spike when prices dump and people most need liquidity. Uniswap pools become unusuable in these conditions, whereas Curveβs v2 pools thrive.
At any rate, these are my humble thoughts. Plenty of other people also weighted in to the thread with well considered opinions, and itβs worth reading through the full thread to offer your comments.
Rocket ETHβs recent launch may solve this (22% boosted at the moment), but the point is well taken. Just as early BTC laptop miners got squeezed out by big farms, so too has the yield farming game changed. Wrapped BTC yield across Ethereum has always been a challenge, and it may turn out it can only be solved natively on Satoshiβs chain if BTC ever launches more native DeFi apps on Lightning.
Iβm not sure whether my two cents (more low yield!) were of any help here. I do know that generally speaking, Iβm always very interested in wrestling with the bear case for Curve.
As a Curve maxi Iβve certainly cultivated an echo chamber that puts me at risk of tunnel vision. I maintain a longstanding offer to turn my newsletter over to any Curve skeptic who wants to articulate the bear case for Curve, because I feel itβs really important for everybody to challenge your perspective.
Notably, this point of view appears to be out of vogue. Lately weβve seen a rash of organizations seeking to win their debate by simply censoring the opinion of their opponents. From my perspective it feels to have the opposite effect. The so-called Streisand Effect suggests censorship just brings greater attention to the object being censored, and from my perspective this appears to be real. Often times the first time I hear about somebody controversial is when I research them after finding they were kicked off social media.
That said, I presume there must be substantial evidence such aggressive censorship efforts must actually be net effective, otherwise it wouldnβt be so common. Itβs possible the most enthusiastic expurgators are simply operating on raw emotion, but I would assume management would be working to tamp down such outbursts if they were truly ineffective. Perhaps that gives leadership too much credit.
At any rate, perhaps itβs ineffective or old school, but when I see arguments against Curve I become quite excited and seek to share them. I feel that engaging in these arguments is the best way to challenge myself and improve my understanding of the subject. In other words, I hope many more of you step up to offer skeptical arguments against the Curve flywheel. Debate of course welcome in the comments.
Disclaimers! Author remains a flywheel maxi