Reference
How Liquidity Affects Pricing
Price isn't a number a protocol stores — it's an emergent result of where liquidity sits. Thin books move on a whisper; deep ones absorb a shout.
TradFi →Market impact & price discovery
Prerequisites
01 · Concept — what problem does it solve?
A common beginner assumption: an asset "has a price." It doesn't. Price is discovered — it is wherever the next trade clears, and that depends entirely on how much liquidity sits near the current level. The same $1M order is a rounding error in a deep ETH/USDC pool and a 30% candle in a thin altcoin. Understanding pricing means understanding that liquidity is the denominator of every price move: impact scales with your size relative to the liquidity behind the price, not your size in dollars.
02 · Mechanics
- : in an , price is the last match; in an AMM, it's the reserve ratio. Either way it moves only when someone trades, and how far depends on depth.
- Depth sets the slope: deep liquidity = a flat local curve = small moves per trade. Thin liquidity = a steep curve = violent moves. The chart below shows it directly.
- Your trade IS price discovery: a market buy walks up the book / along the curve, and the new resting price is higher because of you. There is no passive way to take size.
- Arbitrage propagates price: when one venue moves, arbitrageurs trade the laggards back in line — which is why liquidity anywhere affects price everywhere.
- Oracles inherit this: a price feed is only as trustworthy as the liquidity behind the market it reads — thin liquidity is the root of most oracle manipulation.
03 · Formulas
// constant-product price impact (f = trade as a fraction of the reserve)
impact = f / (1 − f)
// f = 1% → ~1.0% impact (deep relative to trade)
// f = 10% → ~11.1% impact
// f = 33% → ~50% impact (thin relative to trade)
10% of reserve
Price impact ≈ 11.1%
Constant-product (x·y=k): impact = f / (1 − f). The deeper the pool relative to your trade, the smaller f — and impact stays near zero.
04 · Edge cases & risks
- Headline price lies under size: the quoted mid-price applies to an infinitesimal trade. The price you get degrades with every unit — always think in average fill, not top-of-book.
- Liquidity is reflexive: depth thins exactly in a crash, so the same dollar order that barely moved price yesterday gaps it today. Price discovery is least reliable when it matters most.
- Manipulation is cheap where liquidity is thin: skewing a low-liquidity market for one block is the entire basis of flash-loan attacks. See Oracles.
- Concentration changes the math: Uniswap v3 and Curve reshape where liquidity sits, so impact depends on the shape of the curve, not just total TVL.
Connected concepts