Square-Root Law for Impact

• The paper demonstrates that the average price impact of large metaorders scales as Q^(1/2), confirmed by extensive empirical data across diverse markets.
• Square-root law for impact is defined by a non-linear scaling where transaction costs and liquidity responses grow with the square root of the executed volume, contrasting with linear models.
• It provides a robust theoretical framework linking microstructural market dynamics to diffusive price behavior and universal liquidity patterns.

The square-root law for impact refers to a family of empirical and theoretical results across multiple disciplines—most prominently financial market microstructure, but also granular physics and voting theory—in which the average response (impact) of a system to a perturbation of size $Q$ scales asymptotically as $Q^{1/2}$. In the context of financial markets, the square-root law governs the average price impact of large "metaorders" executed over a trading horizon, dictating how transaction costs and observable market responses scale with order size. The law's ubiquity, its cross-domain manifestations, its microstructural origins, and the debate over universality have made it both a central theoretical object and a practical benchmark for institutional execution and market design.

1. Formal Statement and Empirical Law

In financial markets, the canonical square-root law for price impact is formulated as follows. For a metaorder of signed traded volume $Q$ executed over a time window $T$, and on an instrument with daily volatility $\sigma$ and daily traded volume $V$, the average (reactional) price impact is given by

$I(Q) = Y\,\sigma\,\sqrt{\frac{Q}{V}},$

where $Y$ is a dimensionless, market-specific prefactor of order unity. This law is observed to hold for "large" metaorders—typically $Q/V \gtrsim 10^{-3}$—in equities, futures, and option markets, across diverse asset classes and geographical regions (Bucci et al., 2019, Sato et al., 2024, Toth et al., 2016). For small $Q$ there is a linear regime ($Q^{1/2}$0), and empirical crossover functions describe intermediate scaling (Bucci et al., 2018).

A broad empirical literature using datasets of millions of institutional trades in US and Japanese equity markets confirms this scaling over two to three orders of magnitude in $Q^{1/2}$1, with exponents $Q^{1/2}$2 in fits of $Q^{1/2}$3 extremely close to $Q^{1/2}$4 and little systematic cross-sectional variation (Sato et al., 2024, Maitrier et al., 22 Feb 2025).

2. Mechanistic and Microstructural Origins

The square-root law emerges from the mechanical interplay between incoming market orders and the liquidity profile of the limit order book ("latent liquidity"). Models that describe the order book as a diffusive reaction–diffusion system or as a set of metaorders with overlapping execution rates predict that, when the density of available liquidity grows linearly with price distance from mid (a "V-shaped" profile), execution cost grows as $Q^{1/2}$5 (Bucci et al., 2018, Toth et al., 2016, Maitrier et al., 22 Feb 2025).

The law is robust to details of child order slicing and to the presence of anonymity in public data. The impact of synthetically reconstructed metaorders—by scrambling trader identities—remains identical, supporting the hypothesis that the law is a mechanical consequence of order flow and book replenishment, not the information content of orders (Maitrier et al., 22 Feb 2025, Maitrier et al., 5 Sep 2025).

On the microscale, single market orders already display square-root scaling when their impact is measured after the book has "digested" them. On the mesoscale, the aggregate impact of a metaorder is built up from a double square-root effect: the individual impact due to the size, and a relaxation in time governed by a $Q^{1/2}$6 decay (Maitrier et al., 22 Feb 2025). This temporal structure explains empirical observations that transient impact is mostly independent of execution time in the square-root regime (Bucci et al., 2019, Bucci et al., 2018).

3. Theoretical Foundations and Diffusive Price Dynamics

The mathematical necessity of the square-root law is demonstrated in generalized models of persistent, correlated order flow such as the Lillo–Mike–Farmer (LMF) framework (Sato et al., 25 Feb 2025). Here, metaorders are Pareto-distributed in size and generate long-range auto-correlations in order flow signs: $Q^{1/2}$7 A linear impact law under persistent order flow would lead to superdiffusive, predictable prices. However, assigning impact $Q^{1/2}$8 and mapping the system to a Lévy-walk yields mean-squared displacement scaling

$Q^{1/2}$9

with $Q$0 the tail exponent of $Q$1. The choice $Q$2 is marginal: for $Q$3, this guarantees diffusive (Brownian) price increments despite enduring predictability in order flow, thus preserving the efficient market hypothesis (Sato et al., 25 Feb 2025). If $Q$4 were significantly above $Q$5, the market would be trend-following; if below, diffusion persists but at the expense of possible liquidity instability.

4. Universality and Competing Theories

The universality of the square-root law's exponent $Q$6 has been the subject of prolonged debate. Universalists hold that $Q$7 is a critical exponent, independent of asset class, instrument, or market (modulo minor statistical errors). High-precision surveys on the Tokyo Stock Exchange confirm $Q$8 to within $Q$9 across thousands of stocks and traders, with the observed dispersion fully attributable to finite-sample effects (Sato et al., 2024).

Competing nonuniversality models attempt to relate $T$0 to metaorder-size or duration distribution exponents (e.g., $T$1 [Gabaix–Gopikrishnan–Plerou–Stanley], $T$2 [Farmer–Gerig–Lillo–Waelbroeck]) but large empirical studies find no correlation between fitted $T$3 and these predicted values (Sato et al., 2024). The strict universality result places significant constraints on admissible microscopic theories of liquidity and impact.

5. Extensions, Crossovers, and Limiting Behavior

The square-root law is valid when metaorder sizes are large in relative terms $T$4, and metaorder execution consumes a modest to high participation rate of daily market volume. For small metaorders and low participation rates, the impact reverts to the linear (Kyle) regime. Transition between regimes is described via universal crossover functions dependent on the participation rate $T$5 or related variables (Bucci et al., 2018, Bucci et al., 2019): $T$6 For sufficiently rapid execution or high $T$7, one observes a second square-root regime: for fixed volume, impact is proportional to $T$8 in the large $T$9 limit, with a linear regime at small $\sigma$0 (Durin et al., 2023). Temporal decay of impact after execution