Some Numbers That Show Why Yield Farming COMP Is So Seductive
COMP has created a sudden craze for "yield farming." Here are three scenarios illustrating the risks and rewards of DeFi's latest trend.
It's hard to figure out what the new governance token from Compound, COMP, costs.
Not its price, but its cost: How much will a user pay to earn freshly minted COMP? This is made especially complicated because the cost isn't what a user deposits or borrows, it is how much net interest they ultimately pay.
Insanely, right now, it is possible to earn COMP on extremely risky trades for effectively no cost, as we'll show below. This is not a situation that is likely to end well for many.
Compound rewards investors with COMP both for supplying capital and borrowing. To maximize returns, most users do both. They deposit and borrow against that deposit. There are even ways to spin this into loops that eke out even more yield (at higher risks).
This is how Compound has suddenly become the world’s top decentralized finance (DeFi) platform in terms of total value locked (TVL), according to DeFi Pulse. With a small supply and lots of pent-up demand, crypto users are rushing in to earn a return – one that is very strong right now, but not much more likely to last than 2017's initial coin offering boom.
Let’s do the numbers
To estimate cost, a website called Predictions Exchange gives a reasonable idea of how much an investor will spend to earn new COMP, and it helps to show why this asset is so attractive at current prices. The Compound team confirmed to CoinDesk the site's estimates are accurate enough to provide useful guidelines, in a market where factors are changing all the time.
COMP currently sits at $222, according to CoinGecko. The last week has seen wild swings for the new governance token, rising to $338 on June 23 and briefly dipping below $200 on Wednesday. Total value locked (TVL) has been falling in a staggered fashion since June 21, down to about $570 million as of this writing, from a high this weekend of over $600 million.
Below, we game out three scenarios – from conservative to very risky – using Predictions Exchange, assuming everything stays the same for a year (which is a very bad assumption). All these scenarios will make the assumption of a modest investment of $10,000 in capital, with a COMP price of $200.
The point of this exercise is to give some sense of what investors will pay for each newly minted COMP under different scenarios. It's important to note these numbers change very fast and this post is only meant to explain the current frenzy.
The safe-by-crypto-standards way
The lowest-yield stablecoin that can be supplied as collateral on Compound is USDC. It earns an APY of only 0.12% as of this writing.
The safe move here is to borrow another stablecoin, so let's go with DAI. At a collateralization rate of 75% on Compound, this means the user could borrow 7,500 DAI. Then, there's nothing stopping the user from turning around and depositing that DAI again, increasing their COMP earnings on the supply side.
This earns 2.29 COMP at the end of the year, or $458 at the assumed token price of $200.
Over that time, the user would pay $107.25 in interest and earn $76.50 on the two deposits, a net loss on the deposits of $30.75. So, the cost per COMP over that time would be $13.43.
If the investor sold the COMP right away, it would net $427.25.
In fact, if an investor only put the 10,000 USDC in and did nothing else, they would earn 1.06 COMP and $12 in interest, for a net of $224.
By taking on just ever so slightly more risk, the far better move for the conservative investor is to do it with USDT. That would earn 3.21 COMP and $450 in yield on the deposit, for a net of $1,092.
The moderately risky way
This is crypto so the low-risk, low-return move above was never the one driving the action.
Just after COMP began dispensing on June 15, the optimal trade was actually on stablecoins, which meant buyers were fairly protected from swings in the underlying assets.
Users were playing with USDC and tether (USDT), two stablecoins. If someone did basically the same trade now (that is, deposit USDC, max out their borrow for USDT and then deposit it again), they’d get more COMP but it also costs more.
The trade earns 8.0 COMP in a year. The deposits earn $349.50. The loan costs $866.25, though, for a loss of $516.75.
So COMP costs $64.60 in this scenario, and if it were all sold at the end of the year for $1,600, the user would net $1,083.25.
However, yield farmers have now shifted away from trading stablecoins. We saw an unprecedented uptick on the stablecoin DEX Curve last week but there was a giant falloff in volume there Monday, dropping from $110 million on Sunday to around $30 million.
The very risky way
Since last week, Brave’s basic attention token (BAT) and 0x’s ZRX have spiked in yields on Compound, so they earn much more COMP.
Unlike playing with stablecoins, this exposes investors to enormous underlying volatility, and the free money is much too good to last.
The supply of BAT on Compound has skyrocketed. One week ago, it was $1.89 million. It has risen to $237.71 million on Thursday (dipping a little since Wednesday). That means 63.5% of BAT’s total market cap is locked into Compound as we speak, according to CoinMarketCap.
Brave, as the creator of BAT and a major holder, has confirmed to CoinDesk it did not move its reserves into Compound in order to earn a return, as some on Crypto Twitter and elsewhere have speculated.
Meanwhile, ZRX is the next most expensive token to borrow. Its supply has also spiked on the application, going from $5.63 million a week ago to $41.38 million on Thursday (again with a dip since Wednesday). That means 17.5% of the ZRX market cap is on Compound.
So, if an investor ran the same trade (deposit BAT, borrow ZRX at its 60% collateralization rate and then deposit what's borrowed), the deal looks too good to be true.
First, they would earn 33.6 COMP. Note that this is the fastest strategy listed and it's still only 0.65 COMP each week.
Then they would also earn $2,538 on the BAT deposit plus $367.20 on their ZRX deposit. Total earnings of $2,905.20, against a borrowing cost of only $978.60. Amazing! That's a profit just on the deposits of $1,926.60.
If they sold the COMP earned, that would be $6,720. Total profit: $8,646.60.
Cost of COMP? Risk. A lot.
ZRX has been as low as $0.13 this year and as high as $0.43. BAT has had similar swings, as low as $0.11 and as high as $0.31. All it takes for these coins is to move against each other for a user's collateral to get slashed by liquidators and make the price of running this trade very uncomfortable.
Robert Leshner, Compound founder, offered a note of caution on Twitter, writing:
There's already one proposal to further lower the amount of augur (REP), BAT and ZRX that can be borrowed per dollar of assets, and the stakeholders in the community have entered into a broader discussion about adding more ways to tame this boom.
COMPlexity
It's important to note all the estimates above are just that – estimates – and probably not very reliable. This is a brand-new market evolving at the speed of crypto. If nothing else, COMP returns depend heavily on participation levels.
In fact, we ran these numbers last night and again this morning, and many of them had already shifted. There's very little doubt that COMP earnings will change a lot over the course of a year.
Compound Labs made a fairly simple formula for distributing COMP tokens, but in a weird way that makes it somewhat complex to estimate what users might expect each day.
Each day, the software distributes 2,880 COMP tokens to borrowers and lenders on the platform. The amount doesn't change, so obviously the more activity there is the less each participant gets (and vice versa).
This is further complicated by the fact that COMP yield accrues the quickest to markets that have the most demand, and this can change on a dime.
STORY CONTINUES BELOW
As the supply grows people will sell. This is likely to bring the price down into equilibrium with the actual market demand for borrowing crypto for uses besides yield farming.
The question is how many retail investors will get caught up in the frenzy and lose their savings before that happens.