Kyle's lambda (λ), from Albert Kyle's 1985 paper Continuous Auctions and Insider Trading, is the price impact coefficient in the linear market model:
Δp = λ × (Q_buy − Q_sell)
where Δp is the price change and (Q_buy − Q_sell) is the signed order flow (net buying pressure). Lambda measures how much prices move per unit of net order flow — higher λ means a less liquid market where informed trading moves prices more strongly.
Kyle's model distinguishes three types of market participants:
- Informed traders — trade on private information; their order flow is the source of adverse selection for market makers
- Noise (liquidity) traders — trade for reasons unrelated to information (liquidity needs, hedging, rebalancing)
- Market makers — set prices to break even in expectation, setting the spread wide enough to cover adverse selection losses from informed traders
Empirically, λ is estimated via regression of price changes on signed order flow. It is a building block for Order Flow Imbalance and VPIN measures.