Skip to main content
Featured

Good Pools and Bad Pools on Uniswap V3

Brandon Ly

Why are some pools good 🐶 and other pools bad 😈?

The answer comes from breaking down LP profits into:

  • Price changes 📈
  • Fees collected 🎟️

By comparing LPs to options, we discover parallel insights — let's explore!

Price changes
​

  • ⬆️ Price up: positive return
  • ⬇️ Price down: negative return
  • ⤵️ Payoff determined by delta (Δ) & gamma (Γ) of LP position

img-1

Why use options terminology (Δ & Γ) for LPs? Hint: that payoff looks awfully like a short put option!

Fees collected
​

Fees collected are determined by the theta (Θ) of the LP position.

  • 🕒 Θ: Rate of time decay (dV/dτ)

  • 💰 dV = fees collected

  • 🧊 dτ = 1 block → Θ = fees per block 🤯

  • ✅ Near the money: Θ > 0

  • ❌ Far the money: Θ = 0

img-2

Implied Volatility vs. Realized Volatility
​

In TradFi, options selling is more profitable when Implied Volatility (IV) > Realized Volatility (RV). Can we compare IV-RV for LPs?

Yes! But let's use fees instead of IVs since:

  • Easier calculation 🧮
  • Fees collected ⇔ options premia 👇
  • ⬆️ options premia ⇔ ⬆️ IV

img-3

Results match TradFi! 👇

  • 🐶 Good pools (green dots): lie below the line, compensated by high fees given volatility ("IV > RV")
  • 😈 Bad pools (pink dots): lie above the line, not compensated enough ("IV < RV")

(Dot values are summed returns from LPing)

img-4

How do price changes and fees affect returns?
​

  • ⬆️ Price → ⬆️ LP returns (since fees are always positive)
  • ⬇️ Price → ⬆️ LP returns if Θ dominates
  • ⬇️ Price → ⬇️ LP returns if Δ & Γ dominate

Let's define "dominance" so we can analyze pool returns! 👇

We define a metric to measure how much fees dominated LP returns:

Θ dominance=feesfees + |payoff|Θ\text{ dominance} = \frac{\text{fees}}{\text{fees } + \text{ |payoff|}}

(fees & payoff expressed as percentages)

Meaning:

  • 💪 100% Θ dominance → fees drove 100% of LP returns
  • 🤕 0% Θ dominance → price movement drove 100% of LP returns

img-5

Previously, we found that LPing on ENS was highly profitable (+124%), but UNI was not (-28%). By graphing Θ dominance next to cumulative returns, we find:

  • 😔 Bad days (negative returns) driven by price movement
  • 🥳 Good days (positive returns) driven by fees

img-6

Breakdown of positive & negative returns confirms that good pool Θ dominance > bad pool Θ dominance:

  • 😔Bad days: 28% (ENS) > 22% (UNI)
  • 😊Good days: 59% (ENS) > 50% (UNI)

The good pool also had a higher proportion of good days:

  • 🤩ENS: 63% (272/433)
  • ☹️UNI: 55% (335/608)

img-7

The good pool's fees made up for its bad payoffs (ENS):

  • Fees: 466%
  • Payoff: -371%
  • Return: 95%

The bad pool's fees weren't enough to compensate (UNI):

  • Fees: 309%
  • Payoff: -332%
  • Return: -23%

(All values are summed)

img-8

Summary
​

📣 Key Insights:

  • LP = short option payoff
  • Δ, Γ, and Θ affect LP returns
  • LPs compensated when IV > RV
  • Good days/pools driven more by fees than by price changes

Disclaimer:

  • 📢 None of this should be taken as financial advice.
  • ⚠️ Past performance is no guarantee of future results!
Table of Contents