General · EdgepediaOctober 10, 20267 min read

Yakov Amihud

Yakov Amihud is a financial economist, the Ira Rennert Professor of Entrepreneurial Finance at New York University's Stern School of Business, which he joined in 1990, and the author of what is described as the most widely used daily-data measure of stock market illiquidity, introduced in his 2002 paper "Illiquidity and stock returns: cross-section and time-series effects."¹^•² He is among the top 5% of authors in the RePEc registry by citation criteria, with the RePEc Short-ID pam182.³

Key fact	Detail
Position	Ira Rennert Professor of Entrepreneurial Finance, NYU Stern; joined in 1990¹
Signature measure	ILLIQ, the average daily ratio of absolute return to dollar trading volume, $\mathrm{ILLIQ} = \overline{\|R_{iyd}\| / \mathrm{VOL}^{D}_{iyd}}$ ²
2002 finding	Across NYSE stocks 1964–1997, illiquidity has a positive, highly significant effect on expected returns, cross-sectionally and over time²
Earlier work	Amihud & Mendelson (1986), "Asset pricing and the bid-ask spread," Journal of Financial Economics 17, 223–249⁴
Adoption	Over 120 papers in the Journal of Finance, Journal of Financial Economics, and Review of Financial Studies used the measure during 2009–2015⁵
Premium size	Risk-adjusted return on the illiquid-minus-liquid (IML) factor about 4% annually over 1950–2012⁷

Career and institutional roles

Amihud earned a BA in Economics and Political Science from the Hebrew University and a PhD in Economics and Quantitative Analysis from New York University in 1975.¹ At Stern he teaches Mergers and Acquisitions and Restructuring Firms and Industries, and is affiliated with the Volatility and Risk Institute and the Glucksman Institute for Research in Securities Markets.¹ He has published more than seventy research articles and edited or co-edited five books on topics including leveraged buyouts, bank mergers and acquisitions, international finance, and securities market design, and has consulted for the NYSE, AMEX, CBOE, CBOT, and other securities markets.¹ With Haim Mendelson and Lasse Heje Pedersen he wrote Liquidity and Asset Prices (2006) and the Cambridge University Press book Market Liquidity (2013).³

The Amihud illiquidity measure

The measure is deliberately simple. For stock $i$ on day $d$ of year $y$ , daily illiquidity is the ratio of the absolute return to the dollar trading volume, $|R_{iyd}| / \mathrm{VOL}^{D}_{iyd}$ , and the stock's illiquidity is the average of this ratio over a period. Amihud interprets it as the daily price response per dollar of trading volume, in the spirit of Kyle's (1985) concept of illiquidity as the price impact of order flow, the discount a seller concedes or premium a buyer pays because of adverse selection and inventory costs.²^•⁸

Why daily data. High-frequency microstructure measures, such as bid-ask spreads and Kyle's lambda, generally require intraday data that many markets do not provide and that, even where available, do not cover long periods. ILLIQ can be computed from daily prices and volumes over decades, which is what made long-horizon asset pricing tests possible.² NYU Stern's V-Lab documents the measure as $\mathrm{ILLIQ}_t = |R_t| / \mathrm{VOLD}_t$ , forecasts it with a Multiplicative Error Model, and reports a historical measure as a 22-trading-day moving average.⁹

The 2002 results. Across NYSE stocks during 1964–1997, ILLIQ has a positive and highly significant effect on expected return, both in cross-section and over time. Expected market illiquidity raises the ex ante stock excess return (the risk premium), and stock returns are negatively related over time to contemporaneous unexpected illiquidity. Illiquidity affects small-firm stocks more strongly, helping explain time-series variation in the small-firm premium, and market illiquidity's effect helps address the equity premium puzzle.²^•⁸

In his 2020 revisit, Amihud specifies the computation: annual ILLIQ is averaged over the 12 months ending in November of year $y$ and used to analyze returns in year $y+1$ , with filters of price between $5 and $1,000, more than 200 valid trading days, and exclusion of the top 1% of ILLIQ values as outliers.¹⁰

From Amihud–Mendelson to modern liquidity pricing

The 2002 measure grew out of a research program Amihud began with Haim Mendelson in the 1980s. Their 1986 paper modeled investors with different expected holding periods trading assets with different relative bid-ask spreads, and derived the testable hypothesis that expected return is an increasing and concave function of the spread. Testing on NYSE stocks from 1960 to 1980, they found average return rises with the spread after controlling for systematic and unsystematic risk, and that the illiquidity premium exceeds expected illiquidity costs.⁴^•⁷

The shift from spreads to the return-to-volume ratio was a data decision, not a change in theory. Spread-based tests require spread data; ILLIQ requires only daily prices and volumes, available for most markets over long histories. Hasbrouck (2009) found ILLIQ highly correlated with Kyle's lambda and with the bid-ask spread, and Goyenko, Holden, and Trzcinka (2009) found in horseraces against TAQ and Rule 605 benchmarks that the Amihud measure is a good proxy for price impact, while Pastor–Stambaugh's Gamma and the Amivest liquidity ratio are never in the winning group.⁹^•¹¹ The program also went international: Amihud, Hameed, Kang, and Zhang (2015) found stock illiquidity positively affects expected returns across 45 developed and emerging countries.¹⁰

By the numbers

The 2002 paper had garnered over 5,000 Google Scholar citations by the time of Harris's 2019 replication and is among the most cited papers in asset pricing.⁶ During 2009–2015, over 120 papers in the Journal of Finance, the Journal of Financial Economics, and the Review of Financial Studies used the measure.⁵

Premium magnitudes. The risk-adjusted average return on Amihud's IML factor, long illiquid stocks and short liquid ones, is about 4% annually over 1950–2012 using Fama–French–Carhart control factors.⁷ Barardehi and colleagues report that a one-standard-deviation increase in their modified illiquidity measure is associated with 15.9 basis points of excess return, versus 8.3 basis points for the standard close-to-close version.¹²

Criticisms and refinements

The replication. Harris's 2019 replication, using the current CRSP dataset, obtains essentially the same results as the 2002 paper for the original sample, but applying the same methods to 1998–2015 shows a much weaker relation between illiquidity and asset pricing; only the unexpected component of illiquidity still strongly affects returns. Harris also finds the Amihud measure is no better than substantially simpler measures computed from the same data.⁶ Amihud's 2020 revisit agrees the cross-sectional effect has diminished recently but reports the IML premium remains positive and significant over the last 63 years.¹⁰

The volume debate. Lou and Shu (2017) decompose ILLIQ and find its pricing is driven by the trading volume component rather than the return-to-volume ratio intended to capture price impact, suggesting the volume effect reflects mispricing rather than compensation for illiquidity.⁵ Amihud and Noh (2020) rebut that Lou and Shu's decomposition omits the covariance between daily absolute return and the inverse of daily dollar volume, which is priced; the return-relevant component significantly affects returns in cross-section and time-series, and the ILLIQ premium remains significantly positive after controlling for mispricing, sentiment, and seasonality.¹³ A 2023 study in the Journal of Economic Dynamics & Control adds a third position: decomposing prices into permanent and transitory components, it finds the transitory half-Amihud measure on days of negative permanent price returns plays an important role in pricing, so both the trading volume component and transitory price impact drive the premium, in contrast to Lou and Shu.¹⁴

The night-and-day mismatch. The standard measure averages ratios of close-to-close absolute returns to dollar volume, but close-to-close returns include overnight, information-driven movements while volumes come from regular trading hours. Barardehi and colleagues' open-to-close modification (OCAM) produces 50–120% larger liquidity premium estimates and correlations with trading-cost benchmarks higher by 8 to 37 percentage points; the modified measure remains priced post-2001 after adjusting for microstructure-noise biases, especially for small and mid-cap stocks, and the modified series are publicly available for 1964–2019.¹²

Alternative daily measures. Harris and Amato (2018) test alternatives including the inverse Amivest measure and the Kyle–Obizhaeva (2016) invariance measure, the third root of the ratio of return variance to average dollar volume, and find they significantly predict expected returns.¹⁰

References

Topic: Encyclopedia › Society and history › Social and behavioral scientists › Financial economists

Initially written Oct 10, 2026 · Edited: Oct 11, 2026; Oct 15, 2026; Oct 25, 2026

Notice something wrong?