CASE STUDIES: Providing Liquidity on Uniswap V3
I wanted to answer the following questions about Uniswap V3 Liquidity Providing:
Uniswap V3 introduced "concentrated liquidity" where users can supply liquidity in a narrow band to increase their reward generation. But how narrow of a band do we want to make? How often do we want to rebalance?
To answer these questions, I ran several experiements as well as looked at competitors running Uniswap V3 optimizers as well. Note: I don't want to name and shame them so they'll be listed as "Competitors #1/#2", but if you look around you could see who I'm referring to.
How often to rebalance: daily vs weekly vs monthly+?
We only earn rewards on the position when prices moves within the range. If price doesn't move, we don't get rewards. If price moves out of the range right away, we pay swap fees as well as not generating rewards (ick!). We need to pick a timeframe that's long enough to generate rewards, but short enough to get max rewards.
Results of WETH/stable pairs rebalancing frequency:
- Daily: -5.6% APR
- Weekly: 40 to 69% APR
- Monthly+: 6% APR and 13% APR
Daily rebalancing was an absolute disaster: we LOST money by providing liquidity! I deposited $26,659 and withdrew $25,741 after 223 days (~7.5 months). Gross! WETH price was almost identical for both deposit and withdraw, which should've been perfect to avoid impermanent loss.
How could APR go negative when the ranges were tight and we were earning all those rewards? Swap fees, slippage, and impermanent loss. Even though the pool was the lowest fee pool possible at 0.05%, the rebalancing was too active for the rewards generated. Impermanent loss was likely the real killer. Remember that offering liquidity means you take the BAD side of every trade! The rewards generated have to be good enough to overcome this.
Weekly rebalancing seems to be the sweet spot: 40% to 69% APRs! Both WETH/USDT and WETH/GMX tested confidently positive when weekly rebalances.
Competitor #1 and Competitor #2 both rebalance on a monthly schedule (or greater). The rewards generated are available on their applications and they look correct to me: 13% APR for #1, and 6% APR for #2. Frankly, those rewards are too low when considering the risks of impermanent loss.
Conclusion: weekly or bi-weekly rebalancing works best.
How wide to keep the ranges?
The distance between the lower and upper limits of liquidity providing is related to how often you need to rebalance, but it's a little different. If the range is too narrow for the time period, you need to rebalance more frequently or sit out of range. If the range is too wide, there's too little rewards generated.
TidePools is unique because it adjusts the range based on weekly volatility. If the pair is volatile, the range increases. If it's a quiet pair, the range decreases. It does this all on its own. You can read more about
How TidePools Works.
Current WETH / stable price: $1,910 (5/7/23)
- TidePools.io WETH/USDT 0.05% fee. range: $1,722 - $2,051, 66% APR (17.5% price range)
- Competitor #1 WETH/USDC 0.3% fee. range: $1,350 - $2,790, 13% APR, (206% price range)
- Competitor #2 WETH/DAI 0.05% fee. range: $483 - $4,902, 6% APR, (1,014% price range)
Because TidePools keeps a tighter range, we can generate more reward APR. Both competitors have much wider ranges, which means lower reward generation. But as a positive, a wider range means less impermanent loss! You can read the pros and cons of wide vs narrow ranges here.
Competitor #1 sets their ranges manually, so I don't know what their thought process is behind this range.
Competitor #2 looks absolutely lazy to me, uninterested in making money for their users. WETH/stable hasn't been at $483 in years, and has never been as high as $4,902. They have never even run a rebalance on their pool in over a year. They're not even trying.
Conclusion: sub 100% price ranges are necessary to generate enough rewards. For example, if price is $100, the range should be narrower than $50 to $200.
What are the returns like for WETH/ERC20 pairs?
We've talked a lot about WETH/stable pairs because they're the most traded. But WETH/X pairs (or MATIC/X, OP/X, BNB/X) are often used as bridge pairs if a direct pair doesn't exist. For example, if a user wants to swap token A for token B but no A/B pool exists, the swap will often go through the WETH/A and WETH/B pools so the swap still happens.
I started by depositing $25,063 worth of WETH/GMX, and withdrew $29,952 worth of WETH/GMX just 42 days later! Price did increase on both WETH and GMX over USD by 14% and 8% respectively, so we can say price increase on the pair rose by a blended 11%. This gives us value of $27,920, which means $2,031 worth of value came from generated rewards over this time period which is 69% APR! (nice)
WETH/X pairs are also great because crypto tends to move together. If ETH is going up against fiat, BTC probably is too. This means we can have tighter ranges and generate more rewards. The current TidePool for WETH/GMX has a 15% wide price range vs the 17.5% price range on WETH/USDT.
Conclusion: WETH/X pairs have been the highest performers so far, provided the pair has enough volume.
Case study #2 on WETH/USDT: the bull and the bear
We deposited $32,851 worth of WETH on 3/28/23, and have $34,505 worth in the position [as of this writing] (5/7/23). WETH/USDT had some ups and downs during this time period, starting at a price of $1,719, spiking to a high of $2,097, and settling at the current [as of this writing] $1,910. The position made a nominal $1,654 from the start, for a 40% APR and a 5% gain on the position.
However, didn't those realized 5% nominal gains underperform the 11% price increase of WETH over USDT? Yes. Liquidity providing will underperform price in bull markets, and overperform price in bear markets. The rebalance on 4/14/2023 had a total value of $34,563 when WETH/USDT was $2,097. Price decreased from $2,097 to the current $1,910, a -9% drop. However, the total value of the position only lost -2.3%: we outperformed the market during the dump.
If you're really bullish on an asset, it's probably best to hold that asset directly. If you're unsure about price or want to protect your downside, providing liquidity is an excellent choice.
Conclusion: liquidity providing will underperform price in bull markets, and overperform price in bear markets.
Summary
- Daily rebalancing loses money due to fees, slippage, and impermanent loss
- Weekly or Bi-weekly rebalancing looks best
- Monthly+ rebalancing doesn't generate any fees
- Wide price ranges from half to double current price are unlikely to generate meaningful rewards
- WETH/X pairs can generate equal or more rewards than just WETH/stable
- Providing liquidity underperforms bull markets, but overperforms bear markets
With these studies, we can show returns of 40-69% on TidePools managed liquidity positions!
TidePools has a "Live APR" on each pool with an active position to show you the real returns. Of course, actual returns may vary due to
impermanent loss and other factors such as
swap volume, but we've done many studies and tests to show your we're profitable and safe for your crypto.
Farming with TidePools is easy.
Just pick your pool, deposit, and generate rewards on your crypto. If you don't see the pool you want, you can
create your own in seconds!
Have a question that wasn't answered here? Let me know on
Twitter!