The Amihud illiquidity ratio is one of the most widely used measures of how costly a security is to trade. Introduced by Yakov Amihud in his 2002 paper "Illiquidity and Stock Returns: Cross-Section and Time-Series Effects," it answers a deceptively simple question: how much does the price move for each dollar that changes hands? A security whose price lurches on modest volume is illiquid; one that absorbs large trades with barely a ripple is liquid. The ratio puts a number on that intuition using only daily returns and trading volume — data that is freely available, decades deep, and consistent across markets.
What makes the measure valuable is precisely this frugality. The more accurate microstructure measures of price impact — Kyle's lambda, order flow imbalance, and the realized effective spread — require intraday tick or order-book data that is expensive, recent, and patchy across venues. The Amihud ratio recovers most of the same information from end-of-day numbers, which is why it underpins so much empirical work on liquidity and asset pricing. This guide explains how the ratio is constructed, the illiquidity premium it documents, how to turn it into a signal, and the limitations its daily construction imposes. It is a spoke of the broader market-microstructure alpha hub.
Key Takeaways
- The Amihud ratio measures average daily price impact per dollar of volume: the mean over a period of the absolute daily return divided by that day's dollar trading volume. A high value means small trades move the price a lot — the signature of an illiquid asset.
- It is a low-frequency proxy for price impact. It approximates what intraday measures like Kyle's lambda capture, but from daily data that is cheap, long-history, and broadly available.
- Amihud (2002) documented an illiquidity premium: in the cross-section, more illiquid stocks earn higher average returns, and in the time series, expected market illiquidity is positively related to expected market excess return.
- As a signal it is used to build a cross-sectional liquidity factor and to time aggregate liquidity — but it is a slow, noisy measure, sensitive to how volume is scaled and to the price level.
- It is a measure, not an alpha. The illiquidity premium is partly compensation for real trading costs and capacity limits, so a paper return on an illiquid-sorted portfolio overstates what is actually harvestable, and any edge still decays.
What the Amihud Ratio Measures
Liquidity has several distinct facets — the bid-ask spread you pay on a small trade, the depth available at the touch, the resilience with which the book refills, and the price impact a large order leaves behind. The Amihud ratio targets the last of these: price impact, the degree to which trading itself moves the price against you. This is usually the dominant cost for any institution trading in size, which is why a price-impact proxy is so useful even in crude form.
The underlying logic is that on a day when a given dollar amount of trading produces a large price move, the market was unable to absorb that flow without repricing — it was illiquid that day. On a day when the same dollar volume barely moves the price, the market was deep and liquid. Averaging that ratio over many days gives a stable characteristic of the security. Amihud's insight was that you do not need to watch the order book tick by tick to see this; the daily relationship between how much traded and how far the price travelled already contains the signal, if noisily.
The Formula and How to Compute It
For each trading day, take the absolute value of the day's return and divide it by the day's dollar volume (price times shares traded). The Amihud illiquidity ratio for a security over a chosen window is the average of that daily quantity across the days in the window:
ILLIQ = average over days of ( |daily return| / daily dollar volume )
Because raw dollar volume is a large number, the result is a very small decimal, so in practice it is scaled — Amihud multiplied it by ten to the sixth — purely to bring it into a readable range. The scaling is cosmetic; only relative values matter. The window is typically a month, a quarter, or a year, with longer windows giving steadier but more slowly updating estimates.
The table below shows the calculation for an illustrative five-day window (values illustrative, rounded, scaled by one million):
| Day | |Return| | Dollar volume ($m) | |Return| / $vol (×10⁶) |
|---|---|---|---|
| 1 | 0.8% | 120 | 0.067 |
| 2 | 1.5% | 60 | 0.250 |
| 3 | 0.4% | 200 | 0.020 |
| 4 | 2.1% | 45 | 0.467 |
| 5 | 0.6% | 150 | 0.040 |
The five-day Amihud ratio is the average of the last column, roughly 0.169. Notice that days 2 and 4 dominate: a large move on thin volume is exactly what "illiquid" means, and the measure weights those days heavily. A security that posts many such days carries a high Amihud ratio; one whose moves are consistently backed by deep volume scores low.
The Illiquidity Premium
The reason the ratio became a cornerstone of empirical finance is what Amihud found when he sorted stocks by it. Building on the earlier bid-ask-spread result of Amihud and Mendelson (1986), the 2002 paper documented two effects. In the cross-section, stocks with higher illiquidity earned higher average returns, even after controlling for size, risk, and other characteristics — investors demand extra expected return to hold assets that are costly and risky to exit. In the time series, periods when the market as a whole was expected to be more illiquid were associated with higher expected market excess returns, and unexpected jumps in illiquidity coincided with falling prices as that premium repriced.
This illiquidity premium is now a standard factor in asset-pricing research, conceptually adjacent to the size and value factors and to the broader study of cross-asset signals. The important caveat is interpretive: the premium is, at least in part, genuine compensation for bearing real transaction costs and the risk of being unable to sell in a stress event. That makes it partly an artifact of the very illiquidity it measures — a point that matters enormously when you try to trade it.
Using Amihud as a Signal
There are two common uses. The first is a cross-sectional liquidity factor: rank a universe by Amihud ratio and form a long-illiquid, short-liquid spread to harvest the premium, or — more often in practice — use the ratio as a control and a risk characteristic when constructing any other signal, so that you are not unknowingly being paid for illiquidity you cannot actually capture. The second is liquidity timing: track aggregate Amihud illiquidity across the market and treat sharp increases as a risk signal, since rising illiquidity has historically accompanied stress and drawdowns.
A subtler, higher-frequency use is to watch changes in a security's Amihud ratio. A stock whose illiquidity is compressing as volume grows — entering an index, gaining analyst coverage, attracting flow — can experience positive drift as its illiquidity discount narrows. This complements the order-book-level read from order flow imbalance and the toxicity read from VPIN: the Amihud ratio is the slow, structural view of liquidity, while those measures are the fast, tactical view.
Strengths and Limitations
Strengths. The measure's appeal is its data frugality and reach. It needs only daily prices and volume, so it can be computed for almost any security with a trading history, extended back decades, and applied uniformly across thousands of names — none of which is true of intraday measures. Despite its crudeness it correlates well with the finer price-impact measures, which is why it survives as a workhorse.
Limitations. The daily construction is a blunt instrument. It is sensitive to how volume is measured (shares versus dollars, and the well-known double-counting of dealer markets), to the price level, and to days of zero or near-zero return that can distort the average. It says nothing about intraday liquidity dynamics — the spread, depth, and resilience that determine the cost of a specific order — for which you need the realized bid-ask spread, the Roll estimator, or full order-book data. And because the illiquidity premium is partly payment for real market impact and slippage, a backtest that ranks on Amihud and ignores those costs will report a return that evaporates on contact with the market.
The Amihud Ratio in Practice
Treat the Amihud ratio as the cheap, durable, structural gauge of price impact that it is — excellent for characterizing a universe's liquidity, controlling for it in any cross-sectional signal, and timing aggregate liquidity risk, but too coarse to size a single order or to be trusted blind as a standalone alpha. Compute it over a window long enough to be stable, scale it consistently, and pair it with intraday measures when the decision is execution rather than selection. Above all, remember that the illiquidity premium it reveals is partly the market charging you for the very costs the ratio measures: anything you build on it must be validated and backtested with realistic transaction costs before the paper edge is believed.