Decentralized Exchanges

Alfred Lehar and Christine Parlour

Discussion by Katya Malinova

DeGroote School of Business

McMaster University

SNB-CIF Conference on Cryptoassets and Financial Innovation 20–21 May 2021

 

  • asset class
  • common resource

Decentralized finance (DeFi): why exciting?

 

DeFi: provision of financial services without the necessary involvement of a traditional financial intermediary​
​

blockchain

Decentralized Exchanges

Idea:

  • use blockchain to exchange items
  • fully decentralized
  • \(\to\) no single controlling entity, or location, everything runs with smart contracts

 

But how?

Q: how do you set the price?

  1. (Use an oracle)
  2. Use a hard-coded function

Basic idea is great! *Share* the common resource. Pool liquidity. 

Price mechanism:

  • \(T_0=\) contract balance of "Token"
  • \(E_0=\) contract balance of ETH
  • \(k=\) invariance factor
  • key relation \(k=T\times\ E\)

automated market maker

  • send \(T_1-T_0=\) of "Token"
  • \(\to\) receive \(E_0-E_1\) of ETH s.t.
  • \(k=T_0\times\ E_0=T_1\times\ E_1\)

Does/can this pricing function work?

  • "Unbundle" liquidity provision & price discovery 
    • very "non-standard" in finance/econ 
    • Capponi & Jia (2021):
      • volatile exchange rates
        • \(\to\) exploitable arbitrage opportunities = losses to liq. providers
      • \(\to\) AMM are more likely to be used for "stable" pairs of coins
         
  • Mechanical pricing + "tech details" \(\to\) malicious behavior?
    • The price computed using contract balances at settlement
      • \(\to\) opportunities for front-running while in mempool
      • Buterin (2018 blog post); Park (2021)

Possible issues with AMM? 

a

b

c

d

e

f

g

Possible issues with AMM? 

mempool \(\to\) possibly profitable front-running

Figure 1 in Park (2021)

  • Theory: LOB vs. Bonding Curve AMM DEX
    • pooled liquidity & shared price impact \(\to\) may be more liquidity on DEX
  • Empirics: 
    • 36,958 liq pools on Uniswap; 43+ million transactions; 2.5 yrs
    • Many cool empirical observations
      • Re: does the pricing function work?
        • Prices on Uniswap & Binance are very close!
        • But: price impact always higher on Binance (and "spiky" on Binance?)

This paper

  • Stronger connections between the many facts
  • Theory \(\to\)  predictions/inform the empirical analysis?
    • Maybe on returns to liquidity providers?
    • Possibly from other theoretical works on AMM
      • Aoyagi & Ito (2021); Capponi & Jia (2021); Park (2021) 
  • Pick & choose the empirical observations to zoom onto?
    • Drop some or have a clearer "why"?
    • E.g., the "core-periphery" structure on Uniswap
      • ETH = numeraire similar to USD = the reserve currency? 

This paper: suggestions

@katyamalinova

malinovk@mcmaster.ca

slides.com/kmalinova

https://sites.google.com/site/katyamalinova/