Gearbox V2 Protocol Update: December 2022
Back from the holiday hangover, we are at it again. Though, in all fairness, it’s only natural to have a Holiday Hangover when you Cider up this heavy. So here it is, once again, time for a protocol update — this intern was rudely yanked out of full-on holiday mode in order to write this particular update, so you’d better enjoy it.
PS: You can check your Cider and other reward claim details through our Github. In case you need, instructions are in our Discord announcements.
This is only the second time protocol update we have published, so here’ a quick summary of what the protocol update will include. The goal here is to update you on how the protocol is doing overall — including how much usage it’s getting, how healthy/safe the protocol is, what strategies are currently the most popular, and how well (or poorly) its users are doing in terms of profitability when using the leverage provided by Gearbox. These will be done monthly to see how the protocol has been doing.
You can read our first ever protocol update below
So without further ado, here’s the Gearbox Protocol Update: December 2022 Edition
Protocol Performance 🧮
Before we get into the nitty-gritty, let’s take a look at how the protocol is doing overall. How much money is being borrowed and lent through the protocol? How many people have gotten liquidated? Are people making money, and if so how? To get a complete picture, we’ll cover three main areas: liquidity pool stats, credit account stats, and integration stats.
You can find all the data here in dune made by apeir99n or here in google data studio made by Amantay. And some more in main charts by Harsh.
TVL & Liquidity Pools 🏊♂️
1. TVL: TVL is currently ~$105.4m, down $10m since the last update. This is to be expected given no Ninjas were added during the holiday week leading to lower utilisation. This lowered utilisation meant a drop in lending rates leading to LPs pulling out slightly. Bit part of the reason also goes to the $GEAR token based APYs dropping post token launch. All in all, is a small dent through a major transition period and we go back to build now.
2. Lending pools: Most of the TVL drop comes from the lending side — total supply on the lender side is down ~$9m to $92m, compared to $101m last month. This is probably attributable to APYs going down durto above reasons for lenders. Currently WETH supplied leads the pack (bolstered by recent price movements) — there is $32m WETH on the lending side, followed by $35m of USDC and $19m of DAI. There continues to be relatively small amounts of WBTC and wstETH in lending TVL, $5m and $1m respectively. Most of the TVL decrease has come from a decrease in USDC deposits on the lending side ($40m → $35m).
3. Credit Account Borrowing: USDC borrowing still leads the pack, with $25m, following by ETH with $21m, DAI with $16m. WBTC and wstETH remain less popular with only $800k of wstETH borrowed and only ~100k of BTC borrowed. Borrowing demand remains healthy overall — usage has remained stable (it actually probably increased a bit since the last update), which demonstrates the robustness in demand for composable leverage.
4. Utilization Ratios: As stated above, due to no Ninja addition during Holidays the utilisation saw a momentary dip to 55%. With week 1 and 2 Ninjas now in, this has stabilised back to 73%. As a reminder for readers — the ideal utilization ratio in Gearbox is 70%-85%, this is basis our current interest rate curve. We are currently at very healthy utilization ratios for DAI and USDC (83% and 84% respectively), and ETH is slightly underutilized at the moment at 65%. wstETH has a good utilization ratio (despite low overall TVL) of 72%, and WBTC has a rather pathetic utilization ratio of 1.8%. Overall utilization is up from 65% to 73%.
5. APYs: Organic APYs (excluding GEAR) are reasonably good for the three most popular assets, at 2.06% for USDC and 1.24% and 1.23% respectively for ETH and DAI. Organic wstETH yields are at 0.61%, but remember this is on top of staking yields. WBTC organic yields are 0%. When GEAR incentives are included, we have APYs of 4.72%, 4.34%, and 4.22% for USDC, ETH, and DAI, followed by 1.84% for WBTC and 0.61% for wstETH (there are currently no token incentives for wstETH).
That’s the larger picture on TVL, to understand the why of these trends let’s dive in to the CA level.
Credit Accounts🥷
A great way to analyse CAs beyond this is to refer the datastudio page directly and figure out alpha regarding users: Here
Make sure you go to Version option > Select Leverage Ninja(V2) for the most relevant data
- Leverage Ninja Stats: As of right now, there are 124 CAs open, with 101 unique borrowers (some borrowers have more than one CA). Cumulatively there have been 1214 CAs opened. Since our last update, we have 3 less CAs open.
- CA Size Stats: The average (mean) size of CAs is down from last month, at $607k. Average borrow is up slightly while collateral has remained largely the same — we’re currently at averages of $500k borrowed and $107k collateral deposited.
- Leverage Stats: Average leverage in the system is 5.71x, slightly higher than last month but still within the healthy level (leverage went up over the holiday period, suggesting that degens are unafraid of being liquidated even if they are distracted by their families/friends). DAI borrowers are the most degen, with 6.32x leverage. The few WBTC and wtSTETH borrowers are also quite degen, with 6.21x and 6.04x respectively. USDC borrowers are (relatively) comfy at 6x leverage and WETH borrowers are downright safety conscious compared to the others, with average leverage at 5.48x
- Health Factor Stats: The weighted average health factor for CAs is 1.085, which is pretty much in line with last month (1.09) and is a reasonably good health factor (you can be liquidated if this number falls below 1).
Read our leverage bible for tips on how to manage risk better in Gearbox
5. Liquidations: Since last month, we have only had 2 liquidations. Both were stablecoin CAs belonging to the same user, one in USDC and one in DAI, and both happened on December 31st… Happy New Year to the poor soul who was liquidated, may she or he comeback stronger and degen harder. You can read more about exactly what happened here “Least Profitable Strategies” section if you scroll below.
Quick sum up: Size Up, Leverage Up, Liquidations Down, Market Up. Seems like the best way to make monies atm… but where? and that’s what we answer in the next segment.
CA Strategies/Usage Breakdown
I suppose this is where the “alpha” is. In this section we’ll highlight some of the most popular strategies that CA users are carrying out, as well as do a brief overview of some of the most and least profitable CAs that we can find.
NOTE: As you read below and realise your degen leverage trading can also make you additional money if you did it on Gearbox instead, at a cheaper fee, remember one thing. To sign up as a Ninja you can either just pop up in our Discord or post directly on our governance forum. We usually add all the addresses on Tuesdays.
1. Convex & Curve: Convex remains the most frequently used protocol in Gearbox CAs. There is currently about $50.3m of assets in Curve via Convex, divided between ~4 strategies:
- Convex sUSD-3CRV Pool: The most popular strategy in Convex (and overall in Gearbox) is the Convex sUSD-3Crv Pool, which is not surprising because this strategy is also the highest APY (27% Max ROI) strategy available for stablecoins. There is currently ~$27m in this strategy.
- Convex gUSD-3CRV Pool: The second most popular strategy in Convex and the third most popular overall is the Convex gUSD-3Crv Pool, with $8.5m in assets. Max ROI on this strategy is 13%, and usage of this pool within Gearbox CAs is up $1.5m compared to last month. Last month, this was the 3rd most popular Convex strategy, so it seems more people have been depositing into the gUSD strategy, probably at the expense of other stable strategies.
- Convex stETHCrv Pool: The third most popular Convex strategy is the stETHCrv pool, with about $7.6m. This is interesting, because as of today the yields on this strategy are basically 0% (Max ROI of 0.54%). Perhaps something has shifted recently to lower the APYs on this strategy — otherwise the popularity of this strategy is quite puzzling.
- Convex LUSD-3CRV Pool: The fourth most popular strategy that uses Convex is the LUSD-3Crv pool. The Max ROI is currently negative… -9.3%, and there is $4.2m in the strategy currently. Compared to last month, the popularity of this strategy is down quite a bit, which is unsurprising given that the APY is currently negative when you take into account the cost of borrow. ~4.8m has left this strategy, and presumably shifted to the positive APY sUSD/gUSD strategies.
- Convex FRAX Pools: Additionally there is a combined ~$4.4m in the two Convex strategies that utilize FRAX — one is a 3crv pool and one is a FRAX-USDC pool. This (perhaps) suggests some sort of intrinsic demand to use FRAX, since the APYs for these two pools are quite low, at 3.1% and 0.5%.
2. Lido: The second most popular strategy overall in Gearbox is Lido Staked ETH. There is $13.1m of assets in this strategy, and the APYs are extremely healthy, with a max ROI of 27.22%. This is up about 2.5m from last month. Overall, the popularity of this strategy is unsurprising, since staked ETH yields are quite hard to beat, and the strategy is also relatively simple, involving just 1 extra protocol beyond Gearbox, which (from a smart contract perspective) means less risk. The enduring popularity of this strategy is also probably buoyed by the fact that staked ETH withdrawals seem likely to be included in the Shanghai update, which is the next update to the Ethereum network. Once withdrawals are live, the risk of a major depeg event with stETH becomes significantly less likely, which in turn probably increases the perceived safety of Lido’s staked ETH product. This might in turn lead to more willingness to use leverage with stETH.
The two other strategies that also involved Lido’s stETH are the Yearn stETHCurve strategy and the Convex stETHCurve strategy (which was discussed above). Both these strategies involve splitting your assets into stETH and WETH and then depositing them into Curve to earn additional yields via LP fees and CRV emissions, but you’re also giving up half of your potential staking yields (because ~half your assets are in WETH). Since the launch of Gearbox v2, straight up holding stETH has mostly been higher ROI than LPing stETH/WETH, but if this changes in the future, it’s possible the popularity of these various strategies will switch around a bit.
3. Yearn: ~8m of assets are in various Yearn strategies via Gearbox at the moment, which is down from last month when it was closer to ~10m. Most of this 8m comes from deposits in the Yearn WETH vault (4m) and the Yearn DAI vault (2.6m). The WETH vault has a max ROI of 0.74% and the DAI vault has a max ROI of 0.84%, so it’s a bit of a question mark why these strategies still have significant assets in them. This might be due to the strength of the Yearn brand and its perceived safety.
Interestingly, at the moment the Yearn stETHCrv vault (that pairs stETH with WETH and deposits into Curve) has the highest APY of any strategy in Gearbox (max ROI at 29.52%). Perhaps this is a temporary APY spike — there is only ~750k of assets in this strategy via Gearbox, and you would expect that if these APYs were available for a longer period, that more people would switch over, especially from the equivalent Convex strategy, which has 7.6m in it.
So… Are CAs winning?
The popularity of strategies is interesting, but popularity doesn’t always mean profitability… Let’s check in with some of our most profitable CAs and see what they’re up to (and see what, if anything, there is to learn from them).
Keep in mind that when using in-house Gearbox numbers, everything is viewed through a pessimistic lens. This is because Gearbox’s primary concern is, in a sense, “what is the minimum these assets could be worth?”. This attitude is important to ensure that we can always successfully liquidate an account for the amount we think it’s worth. In this sense, ‘profitability’ is somewhat understated and losses are somewhat overstated. See this part of our docs for more details.
Top performing CAs
Last month, gains were good but relatively sensible (our biggest winner made 10% in 1 month, other top performers made 2–4% returns. These returns aren’t bad for just 1 month, but they also aren’t mind blowing.
Now we’re two months in, and the returns over 2 months are a little bit more exciting than in the last update.
Of the top 5 most profitable Credit Accounts so far, 2 are USDC accounts, 1 is DAI, 1 is WETH, and 1 is WBTC.
Some of these CAs are doing quite interesting things, and its worth looking into them a bit more deeply along with the lessons that can be learned.
1. Leverage trading on Gearbox enables you to open positions at lowest fee
First off, we’ll say that the number one most profitable CA is the same one as last month… it’s the person who opened their CA, longed BTC with low leverage, and just held until now. The CA is up 14% in about 2 months time, which is not too shabby. Leverage trading(not farming)on Gearbox has an advantage because the cost of borrow can be significantly cheaper than on perp-driven DEXs or CEXs. While you pay 12–15% on DEXs or CEXs, this can be as low as 0.1% on WBTC on Gearbox.
An aside about borrow costs: it’s very important for all CA users to remember that your net APY will always be affected by the cost of borrow that you’re paying. Because of this, shifts in the borrow cost for each borrowable asset in Gearbox can change the profitability of your strategy, and in some cases even turn a winning strategy into a strategy with negative yield. If you are degen farming with high leverage (e.g 7x or more), we highly recommend that you keep an eye on the utilization ratios and interest rates for the assets you have borrowed so that you are aware if/when your strategies become unprofitable.
2. Understand your strategy and when to rotate it to a trade
The second and third most profitable CAs that we currently have share some similarities, and are the most interesting CAs that this humble intern has investigated so far. Both these CAs are leveraging (pun intended) the unique features of Gearbox to do a range of activities. They are definitely leveraged farming, but they are also switching between various farming strategies based on profitability, as well as occasionally taking on leveraged trading positions (with pretty good results). The beauty of the Gearbox CA is that you can do a range of different things with it, and these two winning CAs are taking full advantage of this fact.
The first of the two CAs is USDC denominated, currently at 7.84x leverage, which is quite high. They begin at the start of November by depositing into the Convex LUSD pool. It appears when the APY of that pool went down, they switched to the Convex GUSD pool — this happened in the back half of November. All is quiet until the start of the new year — a couple of days into the new year, this CA used about a third of its assets and bought ETH with USDC. Over the next 5 days, they sell off this ETH, which leaves us where we are today. The CA is fully back in USDC positions (back in the gUSD strategy), but thanks to smart farming and a good trade, it’s up ~11.5% since inception.
I’d describe the above as a CA user with the strategy of ‘leverage farm and trade opportunistically’. In this case it has worked out great for this particular ninja.
3. Or just mix up the above two and create a leverage trade with a partial farm…
Somehow, the 3rd most profitable CA is even more interesting than the last one…
The user starts this CA near the end of November by borrowing WBTC. Already this CA is extremely intriguing. They then trade ~20% of this BTC into ETH, which they then stake into Lido, but then the next day, they sell this staked ETH back into WBTC (???). Over the month or so, they trade between WBTC, ETH, and USDC, including some intraday trading, then a few days ago, they move some of the ETH into Lido. This CA currently 60% in BTC, 25% in staked ETH and 15% in WETH. The account is up 8.1%, and has a leverage ratio of 7.36 at the moment. Keep in mind that this CA was started at the end of November, so it’s only been active for 1.5 months.
This CA user appears to be using Gearbox to trade actively, and is taking advantage of two features of Credit Accounts.
- Firstly, they are very smartly taking advantage of the structural low borrowing cost of BTC (almost free to borrow right now) — levering up on BTC costs them almost nothing.
- Secondly, they are taking advantage of the incredible flexibility of CAs — when their ETH position is large enough, they are depositing into LIDO to earn additional yield from their ETH.
As an aside, I think it’s worth noting that the next 5 most profitable CAs are all WETH denominated accounts that are farming in using Lido stETH. Simple depositing into Lido appears to (thus far) have been the most consistently profitable simple farming strategy available in Gearbox, despite what the various APY numbers might show.
It’s pretty cool that some of these CAs are nearing 5% return since inception, given that 5% yield is (approximately) what you might earn in a whole year if you were just using Lido instead of combining Lido with leverage in Gearbox.
But then, there are CAs who aren’t doing as well…
Least profitable strategies
Rather than highlight the actual ‘worst performing’ currently existing CAs, we’ll do something different here. Since the time between last month’s update and today, we saw two liquidations, and therefore it’s probably fair to say that the ‘least profitable CAs’ are the ones that got liquidated.
The two CAs were opened by the same user, and in short they teach the same lessons that users of leverage around the world learn everyday…
- Don’t use too much leverage
- Don’t trade/move in and out of positions too frequently
- Don’t chase headline APYs without a sense for what the yields might look like longer term
On one of these CAs, denominated in DAI, this user immediately went max leverage (9x+ leverage). At these levels of leverage, you have no room for anything at all — just some very minor and run-of-the-mill slippage on any given tx can basically cause you to get liquidated. It’s not clear exactly what happened — this account went max leverage, started farming Convex gUSD, and got liquidated in 2 weeks. My best guess is the leverage was just too damn high, and between gas costs and slippage getting into the farming position, this CA was destined to be liquidated.
The other liquidated CA from the same user was USDC denominated, and also used high leverage, plus it did a lot of moving funds between different stablecoins and in and out of strategies. On top of using too much leverage, this CA was also entering and exiting farming positions constantly (probably chasing headline APYs with no eye on the longer term?). Needless to say, between trading back and forth between USDC and DAI + moving between the Yearn DAI vault and the Convex gUSD pool, the CA encountered enough slippage to be liquidated.
The goal of this kind of analysis of liquidated/poorly performing CAs is not to make fun of the recently departed or make light of users losing significant amounts of assets. The aim here is to better educate existing and future users so that they better understand the risks of using leverage and so that they can understand what kinds of behaviors are correlated with good and bad performance when its comes to Gearbox’s Credit Accounts (and leverage in general).
Protocol Health
And phew! That sums up our protocol performance bit. The next part is regarding protocol health and safety related developments and more:
Bad Debt
There were two liquidations that happened since the last protocol update, and both liquidations went smoothly with no bad debt occurring.
Risk Assessments & RiskDAO Reports
Another month, another set of juicy reports provided by the RiskDAO team for those of you who are big enough finance nerds to find such things interesting.
- There was some commentary about the exploit on BSC of a MakerDAO fork called Helio and its stablecoin called $HAY, following an earlier hack on ANKR and their staked BNB product. The RiskDAO team comments on why this was possible (bad oracle choice) and why such an attack couldn’t happen on Gearbox.
- The RiskDAO team covered a hack that happened on Arbitrum to the Lodestar protocol. The hack was quite complex, and is broken down in detail here.
- Nearer the end of the year, a range of different topics were touched on, including the Ankr hack (inside job!), some potential risks (at the time) for GMX, the winding down of REN 1.0 assets, and pre-holiday CEX fud.
- In the last risk report of 2022, the RiskDAO team covered the arrest of Avi Eisenberg of “highly profitable trading strategies” fame, and a project called Defrost on Avalanche (allegedly) exit scammed by pretending they got hacked.
You can read all RiskDAO reports in the governance forum here, and the Gearbox Risk dashboard can be found below.
Bug Bounty & Insurance
If you read last month’s update, in November and December, we temporarily doubled the bug bounty program via ImmuneFi until the end of the year as part of a special campaign. Since then, we’ve also permanently raised the maximum bounty in the program by 50k, from 150k originally to 200k now (these updates just went live). You can check the page below.
ANNOUNCEMENT: You can also now buy insurance for your personal positions in Gearbox via Insurace. The below video will show you how.
Protocol Enhancements
This section covers all the details regarding protocol enhancement that took place over the month of December. Dig in!
Gearbox V2.1
Well, we teased you last time. Why not tease you more? Here’s the first round of alpha on the next phase for Gearbox from 0xMikko
Adapter upgrade
There was an minor update to the protocol’s Uniswap adapters as voted on in GIP-38. Basically, this update prevents weird pathing when trades are made through the Uniswap adapters.
Alpha: This update also is one step towards partial withdrawals, which will possibly be implemented in the future.
Enhancement Discussions kick off
There are two key discussions that kicked off regarding possible enhancements
- There is some initial discussion about the potential addition of LUSD as a borrowable asset in Gearbox. You can discuss that here.
- There is also some discussion about token utility — given some of the suggestions, this has bearing on both the protocol and the DAO, so we are including it here for your consumption. You can check it out below.
And that’s all folks…
That’s all we had about last month’s protocol updates. Hope you gathered some alpha from it
If you would like to earn passively, remember to LP into our Lending Pools to earn GEAR. If you want to take leverage as Ninja just ping us on Discord or post on the forum. Otherwise you can always get involved with the DAO — discuss, research, lead and share. Call contributors out on their bullshit and collaborate on making things better.
Here is how you can follow developments:
- Website: https://gearbox.fi/
- dApp: https://app.gearbox.fi/
- User Docs: https://docs.gearbox.finance/
- Developer Docs: https://dev.gearbox.fi/
- Forum: https://gov.gearbox.fi/t/start-here-forum-rules/
- Blog: https://medium.com/@gearboxprotocol
- Github: https://github.com/Gearbox-protocol
- Twitter: https://twitter.com/GearboxProtocol
- Snapshot page: https://snapshot.org/#/gearbox.eth
- And, of course, the DAO working docs:
