THORChain Data Digest: Pool Depths and Composition of Liquidity
Welcome back my fellow Thorchads and chadettes! In this Data Digest I will be diving into the pool depths and composition of liquidity on THORChain. High liquidity is essential for THORChain to succeed, and THORChain aims to capture this by making LPs the “first class citizens”, drawing liquidity through game theory and economic incentives. In this analysis, I look into how the game theory is playing out in practice. Two major trends stand out in the data. Firstly, we see LPs are observant, and employ evolving LP strategies over time. Secondly, we see that while BTC consistently draws in significant liquidity on every cap raise, ETH and ERC-20 pools are falling behind over time.
1.1 Pool Depth Overview: Liquidity and Unique Deposits
Here we have the pre-hack depth of all pools (USD), the count of unique participants in each pool and the average deposit size (USD). The average pool has 256 unique LPs, and an average deposit size of $44,787. Unsurprisingly, BTC and ETH have most unique LPs, at 1386 and 1141 respectively, and relatively small average deposits of $21, 870 and $17, 665 USD. This suggests they see widespread demand from LPs, and are some of the least whale dominated pools on all of THORChain.
Many of the ERC-20 pools have a vastly different liquidity composition. DODO, Cream, ALCX among others all have very few unique LPs, but very large average deposits, suggesting they are whale dominated. As I explain in section 2.2, this is because the pools are aggressively bootstrapped to get listed on THORChain, but fail to capture additional liquidity in later cap raises. The high opportunity cost of pooling ERC-20s on THORChain (ERC-20s can be highly productive elsewhere in Defi), coupled with poor fee earnings relative to BEP2 pools, make them unattractive for LPs.
LTC stands out to me an interesting anomaly. Of all pools to capture over a million USD in liquidity, it has by far the lowest average deposit of just $5,103, and a high number of unique LPs. My feeling here is there are many participants who have a small amount of unproductive LTC lying around, and want to put it to work. The Defi opportunities for LTC are minimal, so THORChain becomes the best place to deposit their holdings, however small they may be.
2.1 LPs are observant, and evolve their LP strategies over time:
Game theory is an important part of how Thorchain works. Regarding pool depth, we of course want pools to find sufficient liquidity for their demand. In theory, this should play out on THORChain for two reasons. Firstly, small pools allow LPs to get a larger share of the earnings, incentivising liquidity to come in. Secondly, THORChain’s slip fee model reinforces this, generating additional fee revenues for small pools that are easily put out of balance. Combined, these mechanisms should allow pools to find an equilibrium to meet demand. Encouragingly, we can see this play out in the data, even in THORChain’s limited cap raise environment. LPs are observant, and actively move their liquidity to optimise returns.
Here we have the weekly Net Liquidity Changes (deposits — withdrawals) for USDT since launch of MCCN. The peaks correspond with caps raises, and the dips with the weeks in-between. Not only is the amount of liquidity added to the USDT pool in each caps raise decreasing over time, but the amount being withdrawn on in-between weeks is increasing too. On the week 2 caps raise, 141,856 Rune was added to the USDT pool. By the final caps raise, which was the largest raise caps raise of all across all pools, just 62,675 Rune was added. Similarly, week 3 between caps raises saw 0 liquidity withdrawn from the pool, but by week 10 there was a -17,852 Rune net change in liquidity.
This mirrors the earnings of the USDT pool (figure 3). For a more detailed breakdown of earnings on THORChain, you can see my report here. The earnings of the USDT pool have declined over time, with other stable coin pools such as BUSD rising in profitability instead. Importantly, THORChain’s game theory is playing out. LPs are not simply dumping their liquidity and forgetting about it. They are actively paying attention to APYs and depositing with evolving strategies over time.
2.2. All Pools
Looking more generally across all pools, figure 4 shows us the breakdown of which pools are seeing increasing growth rates in liquidity over time, and those which are seeing a decline. To analyse this, I look at net liquidity changes before and after June 18th. In figure 5 we can see the growth rate in Total Value Pooled (TVP) on THORChain over time. The halfway point of TVP, just above 10 million Rune, is reached on June 18th. All things equal, we should expect to see an equal amount of liquidity added to each pool before and after this date. Any deviation from this tells us how the priorities of LPs are changing. Pools highlighted green in the “before June 18th” column saw most of their growth in the first half of MCCN caps raises, and pools highlighted in red saw most of their growth in the second half of caps raises.
BTC and ETH have remained consistent as the two most popular pools for LPs to add liquidity to, only showing a slight decline in growth of liquidity of 10% and 12% respectively across the two time periods. BEP2 pools however have seen a significant increase in popularity, with BUSD liquidity growth up by 75%, and both BEP2 BTC and BEP2 ETH up by over 100%. This follows what we would expect from the earnings. BEP2 pools have performed increasingly well as arbitrage drives up fee revenues. LPs noticed this and adjusted their strategies to capitalise.
By contrast, USDC, USDT, LTC, Bitcoin Cash, as well as many of the smaller ERC-20 pools all saw decreasing growth. For many of the ERC-20 pools, such as HOT, Sushi, and SNX, there was a large initial spike of liquidity early on, but little to no change in later cap raises. They saw significant seed funding to get the pool listed on THORChain, but failed to capture additional liquidity from LPs. ERC-20s are some of the most productive assets in all of Defi, so the opportunity cost of pooling them on THORChain is high. This, coupled with the poor APY of ERC-20 tokens relative to BEP2 pools means they fail to capture much liquidity, and are increasingly less important to THORChain as time goes on.
3.1 Zooming in on the latest Caps Raise:
Zooming in on the final caps raise before THORChain was halted, we get our best glimpse of what the most popular pools will be in a more mature THORChain ecosystem. A total of 3,598,907 RUNE was added on the last caps raise on 2nd July. A breakdown of this is in the pie chart below. Unsurprisingly, XRUNE was the big winner, capturing a whopping 66% of all liquidity due to its IDO. While we cannot expect XRUNE to maintain such a share of growth in liquidity across further caps raises, @TehSlaw’s tweets have made me more bullish than ever on Thorstarter and XRUNE, so my feeling is it will capture a similar amount of liquidity to BTC and ETH once ERC-20s are eventually restarted on Thorchain.
One big surprise is that ETH fell from its consistent number 2 spot behind BTC, to 5th in liquidity growth. It was overtaken by XRUNE, BEP2 BTC and BUSD. In an environment of limited cap raises, the rise of BEP2 pools has eaten into ETH’s share of THORChain liquidity. BTC’s share of growth by contrast is untouched. This speaks to what LPs value on Thorchain. There are great Defi alternatives for your ETH, since Ethereum has such a broad Defi ecosystem. You pay a high opportunity cost when pooling your ETH on Thorchain. By contrast, the opportunity cost of pooling your BTC is tiny. BTC is one of the least productive assets in the whole Defi ecosystem, so THORChain’s unique opportunity to earn 15%+ APY offers immense value to LPs. As the THORChain ecosystem grows, I expect BTC’s important role on the platform to stay untouched.