Research Summary: Exploring Price Accuracy on Uniswap V3 in Times of Distress

TLDR

  • While professional market makers play a key role in traditional finance, DeFi has come to rely on liquidity providers. However, liquidity providers may fail to provide this service during price shocks.
  • The price of a cryptocurrency pair can only trade in a range where liquidity has been deposited. In the event of rapid price shocks, the liquidity distribution does not change fast enough. Thus, Uniswap V3 is not able to handle such shocks leading to large mispricings.
  • For the success of Uniswap V3, liquidity providers must become more sophisticated and agile.

Core Research Question

Can Uniswap’s latest market design (V3) cope with external price shocks?

Citation

Lioba Heimbach, Eric Schertenleib, and Roger Wattenhofer. “Exploring Price Accuracy on Uniswap V3 in Times of Distress.” Proceedings of the 2022 ACM CCS Workshop on Decentralized Finance and Security. 2022.
https://www.research-collection.ethz.ch/handle/20.500.11850/579316

Background

  • Decentralized Exchange (DEX): DEXes allow users to exchange cryptocurrencies without giving up custody of their assets. Users interact directly with the smart contracts that build the DEXes. Trading is enabled by liquidity reserved in a smart contract referred to as a liquidity pool. Thus, when a user wishes to exchange 𝑋-tokens for 𝑌-tokens, the user interacts with the respective liquidity pool, deposits 𝑋-tokens in the pool, and receives Y-tokens from the pool’s liquidity.
  • Liquidity Provider (LP): Individual LPs supply the pool’s liquidity by depositing both assets in the respective smart contract. For their service, the LPs receive transaction fees from the trades supported by their liquidity.
  • Concentrated Liquidity Constant Product Market Maker (Concentrated Liquidity CPMM): The concentrated liquidity CPMM, pioneered by Uniswap V3, is an overhaul of the original CPMM design. The CPMM ensures the product between the pool’s reserves remains constant during trading. Concentrated liquidity CPMM is a subtype that allows LPs to allocate the capital in a predefined price range, which leads to greater capital efficiency.

Summary

  • Uniswap V3 does not appear equipped to handle volatile cryptocurrency markets. Its Achilles heel partially lies in its reliance on LPs to react to market changes. While market making has long been handled by professional market makers in traditional finance, we observed that Uniswap V3 LPs as a whole currently lack the sophistication and agility required for complex market making.
  • The LPs’ apparent inexperience calls into question the suitability of the concentrated liquidity CPMM for DeFi. Not only did we observe that Uniswap V3 could not track the price of the primary market, but also saw that both its predecessor and Curve were fit for the task.

Method

  • We analyze Uniswap V3 price data between various stablecoin pairs between 7 May and 12 May 2022. Our analysis focuses on two stablecoins: UST and USDT.
    • The price of the UST began to falter on 7 May and fell to 30 cents by 11 May. UST never fully recovered from the price drop and currently trades at a couple of cents.
    • On 12 May, the price of USDT, the biggest stablecoin in terms of market capitalization, also dramatically dropped and traded 6% below its 1 US$ peg. The price, however, recovered, and USDT has since been trading only slightly below its peg.
  • We launch an Erigon client to collect data from the Ethereum blockchain between block 14’726’564 (first block of 7 May) and block 14’764’083 (last block of 12 May). In particular, we filter event logs for all events related to Uniswap V2 and V3 as well as Curve.
  • Price data from Binance is obtained from their market data API and downloading trade data for the relevant trading pairs.
  • We compare the Uniswap V3 price curve of stablecoin pairs, including UST or USDT, to that of Uniswap V2, Curve, and Binance. USDC’s price was stable in mid-May and is, therefore, used as a price reference for the other two stablecoins.
  • Further, we calculate the liquidity depth of the Uniswap V3 pools during the depegging of the stablecoins.

Results

  • We observe significant price inaccuracies in the Uniswap V3 stablecoin pools that hold either UST or USDT.

  • The price in the UST-USDT (f = 0.05%) Uniswap V3 liquidity pool, almost stayed constant during the UST crash and, thus, was extremely inaccurate. The price on Curve, another DEX, followed the price of the reference market (Binance).

  • The reason that Uniswap V3 was not able to follow the reference price, was that there was no liquidity outside the equal price of the two assets. Thus, there was no liquidity to support trading at the new price.

  • Further, half the liquidity remained in the pool as it was loosing its value.

  • We observe a similar trend in the USDT-USDC (f = 0.01%) Uniswap V3 pool. The pool’s price was also inaccurate during the USDT price drop in May. However, other DEXes, such as Uniswap V2 and Curve, tracked the price on the reference market (Binance).

  • Here, again there was no liquidity to support trading outside a small interval around the equal price of USDT and USDC.

Discussion and Key Takeaways

  • Uniswap V3 cannot be used as a time-weighted average price (TWAP) oracle. To address the problems with using Uniswap V3 as a TWAP oracle, we make two suggestions: either do not use Uniswap V3 as a TWAP oracle or only deem prices from Uniswap V3 trustworthy if there is enough liquidity on both sides of the current price to support trading in either direction.
  • Uniswap V3 LPs (in part) lack the agility for complex market making. To address this observed lack of agility, LPs should be made aware that market making is complex and risky. Alternatively, the protocol could also move the liquidity for LPs.
  • The risk-reward ratio is unfavorable for LPs to deposit liquidity around the new price during market distress. We note that LPs must be compensated for taking on this risk, which is comes from price volatility. One suggestions would be to adopt a fee structure that would increase with volatility.

Implications and Follow-Ups

  • We believe that concentrated liquidity market designs must attract sophisticated and agile LPs in order to ensure price accuracy. These LPs will only join the system if they are adequately compensated for their risks.
  • Thus, we conclude that future market designs should investigate how to properly compensate LPs for their service.

Applicability

  • This work highlights flaws in the Uniswap current market design and can guide future development.
1 Like