Green Jeff's Magical 40% APY ETH Farm Bonanza
Fellas we YIELDING out here !!!
DeFi yield remains incredibly powerful if you know what you are doing. In my opinion, yield is particularly interesting on non-stable assets for bullish holders. Getting APY and accumulating more of a coin you are looking to hoard as much as possible of is invigorating.
Below, I outline steps I take myself to achieve 40%+ APY on my favorite asset, ETH.
Before we get in, my previous article explains how DeFi yield is more than just risk-adjusted ponzi games. Make sure you read through and understand those mechanics before digging into the content below.
40% APY on ETH? In this economy?
The following outlines steps taken (and associated risks) on deploying ETH in a low-risk manner to hit this premium APY. There are other boutique opportunities to replicate high yield that you can figure out yourself. I will tease some of those in future articles.
ETH > wstETH > Euler > ETH > wstETH > Euler
Convert ETH (or stETH if you already hold it) to wstETH. You can use 1inch to source the best conversion rate. Lido Staked ETH currently pays out ~4% APY.
Deposit wstETH to Euler finance as collateral. Euler pays an extra 3.75% on deposits. That takes our total APY to 7.65% on ETH.
Borrow WETH from Euler. Swap through 1inch back to wstETH and redeposit. I am levering up to 2x my deposit. Note that this will require you to do multiple loops of these actions.
Net borrow cost is about -0.3% APY. 1 x 0.3% + 2x 7.65% = 15.6% total APY on this position.
Euler > DAI
Our maximum LTV is ~75% on our 2x initial deposit, but we’ve only borrowed 50%. This leaves another 0.25 * 2 = 50% borrow capacity from our initial deposit.
Borrow DAI from Euler. DAI currently has a net borrow cost of about -1%. Generally speaking, bluechip collateral + stablecoin loan + active human management is more-or-less safe around 70% of maximum LTV. 0.5 * 70% = 0.35x borrow from our initial deposit, bringing our APY on the borrow to about 0.35%
Understanding Risk
Euler is a little bit different from other lending markets. Each token has a collateral factor and a borrow factor that are used to derive effective value of the deposit/borrow.
Essentially, take the collateral * collateral factor to get max LTV, then take the borrow / borrow factor to get the multiplier cost on the borrow. From above, DAI has a borrow factor of 0.88, so every 1 DAI you borrow counts the loan value as 1 / 0.88 = 1.136 DAI.
There are two major risks of liquidation:
stETH:ETH peg deteriorates
USD/ETH rate decreases
There’s also some minor concern to watch out for in the fluctuation of lending/borrow rates at Euler. Just monitor and react if anything insane happens (like ETH borrow cost goes to 100%).
Using each of the numbers above with simplified units, our position is 2 wstETH collateral, 1 WETH borrow, and 0.35 DAI borrow. The “risk-adjusted” collateral and borrow is as follows.
Deposit: 2 * 0.85 = 1.7
Borrow: (1 / 0.91) + (0.35 / 0.88) = 1.5
The position can be liquidated if the net borrow is valued higher than the net supply. From the above scenarios, that could happen in either situation:
wstETH drops 12% vs ETH (1.7 - 1.5/1.7). From current peg, that happens around 0.85 stETH/ETH
ETH dumps so much vs DAI (USD).
That equation is a little bit trickier, because we have an ETH denominated position on both sides of the equation. It looks something like:
1.7x = x(1 / 0.91) + (0.35 / 0.88)
1.7x = 1.0989x + 0.398
0.6x = 0.398
x = 0.663
In other words, our position can be liquidated if ETH drops ~33%.
Both of these rates are pretty modest, but there is a much bigger risk of a turbulent market where each of these risk scenarios happen simultaneously (and history suggests this is plausible).
Let’s say stETH drops 5% vs ETH immediately. Now our deposit is only worth 1.7*0.95 = 1.615.
Rerunning that same equation above, now our liquidation only requires a more modest drop in ETH of 23%.
Fortunately, there are three defenses in our favor to react to this:
Our collateral is autocompounding faster than our loans improve. Snapshot our position values and fastforward a month from now. Our collateral is now worth 1.711, while the borrow is still worth virtually 1.5. We’re slightly better off, with a new stETH liquidation trigger at 0.845, which is a 5% bigger hit.
It’s easy to add more collateral. Note that our position is still earning more yield (mostly outlined below) that can be converted to USDC and added as collateral. It’s also important to have some coin on the side to deploy as needed (for example, if stETH drops to 0.92, not a bad time to convert some ETH or stables into stETH and pad that margin.
One can easily refund the stablecoin portion of the loan. Actively manage the loan to maintain those target benchmarks above. There’s no degen leverage trading going on, just unwind from the stablecoin yield strategy to pad liquidation rates.
Additionally, stETH post-merge will eventually trade around 1 ETH. At some point in the future, holders will be able to swap both ways through LIDO, which will greatly protect the stETH peg by enabling arbitrage both ways, instead of just above peg (if stETH trades higher than ETH, traders can always mint stETH 1:1 and sell. Now if stETH trades below peg, they can swap in from ETH and redeem 1:1).
With all that being said, these are quite advanced mechanics and should not be employed by users who are not already intimately familiar with how these protocols work. If you aren’t there yet, use this bear market to get your defi iq up!
DAI > FRAX > Vesper > Convex
The last part of this equation is earning additional yield on your DAI loan. I believe that the Vesper-Convex FRAX strategy is the best risk-adjusted APY right now. It is a no-IL 2x boosted by Convex FXS gauge that autocompounds FRAX deposits by yield farming through Convex.
First, swap DAI to FRAX using 1inch like before. Again, 1inch is good because it aggregates the best rate for you.
Next, deposit FRAX to Vesper. Right now, the FRAX pool pays out 6.23% auto-compounded.
Take the output pool share token and stake on Convex (similar to how Curve-Convex activity works). They pay out an additional 6.46% in FXS + VSP in addition to that underlying 6.23%
The stablecoin is earning ~13.65% across the negative loan APY, Vesper, and Convex.
Note that the 6.46% is the base APY and you can opt-in to lock the position and multiply APY. For example, if you lock for 6 months, that 6.46% gets boosted an additional 1.35x. I don’t necessarily endorse this, as you may need to actively withdraw and rebalance your position. But if you have significant coin on the side you can post as more collateral instead, this could give you another minor APY bump.
Grand Finale: Yield Assessment
To recap, our total APY derived from our ETH positions is 15.6%. Then the stablecoin earns us 13.65% * 0.35 = 4.78%. Our initial APY is about 20.xx%.
Of course, 20 and change is a lot less than the clickbait 40% I floated. So what’s the catch?
Our stETH position also encompasses hidden high APY under the notion that you enter the position at 0.965 stETH:ETH and exit at 1:1.
The merge is scheduled for next month, and then redemptions begin probably 6-12 months after. In anticipation of redemption, ETH bulls will drive stETH closer to 1:1 while getting their premium staking APY. Let’s assume we can exit our position at 1:1 somewhere in the next 6-12 months. At close, that translates to 1 - 0.965 * 2 = 7% ROI. If we do it in 6 months, that’s 14%. Within that 6-12 month benchmark, our APY is now somewhere in the 28-35% range.
But it doesn’t stop there! ETH staking will also see boosted APY at merge, as validators will now also earn some tips from transaction fees. It’s hard to gauge what this will look like until it happens. Estimations I have seen range anywhere from an extra 2-12% APY. To those calling for 18% native network APY on ETH, I say I’ll believe it when I see it. Let’s use that low end. 2% * 2 gives us another 4% APY.
Additionally, the stETH earned below peg earns the re-peg premium too. It’s a small number in the grand scheme of things, probably an extra 0.25-1%.
Lastly, If you believe in the DeFi supercycle squared infinity thesis, you should appreciate that a lot of your yield comes in the form of good DeFi protocols: FXS, EUL, VSP. If those tokens go up in value before you convert them, so too does your APY increase. And even if they don’t, FXS and VSP (unsure about EUL) earn significant native APY in rev share. You can stake your rewards to get revenue share there.
And as mentioned above, you can lockup your FRAX position in Convex to multiply APY there. Add it all up, and over a 6-12 month timeframe, you can expect somewhere between 30-45% APY on your ETH.
For an ETH bull like me, that’s pretty damn good
.