Does trading volume or liquidity predict stock returns?
Opulence Alpha Research · Published Sep 25, 2026 · Data through Aug 21, 2026
A little. Across US stocks since 1995, the fifth trading on the heaviest volume for its own past year beat the median stock over the next week 51 times in 100, the quietest fifth 49. Over three months the least liquid fifth beat it 52 times in 100, the most liquid 48: a liquidity premium, before the higher cost of trading illiquid stocks. Small tilts, not forecasts.
Keywords: trading volume and stock returns, does volume predict stock price, high volume return premium, liquidity premium, Amihud illiquidity, Kyle's lambda·JEL classification: G11, G12, G14, C12, C58·Concepts: Multiple testing, Point-in-time data, Backtest overfitting
- Universe
- The same for every study: 1,767 US common stocks in the 11 GICS sectors, 422 of them since delisted; S&P 500 members since 1996 plus large and mid-sized companies outside the index
- Period
- Every Wednesday from Jan 4, 1995 to Aug 19, 2026: 1,634 Wednesdays across 7,959 trading sessions
- Sample
- n = 1,763 stocks of the 1,767-stock universe; 1,889,758 stock-weeks, a median of 1,148 stocks per Wednesday
- Outcome
- Return over the next 1, 5, 21 and 63 trading sessions against the same day's median stock
- Inference
- t on non-overlapping dates; proven only when |t| ≥ 3 and the same sign in at least 60% of years
1Introduction
Volume is widely read as confirmation: a move on heavy trading is taken to mean more than one on light trading, and thinly traded stocks are thought to pay more for the trouble of holding them. The literature gives each idea its own name. Gervais, Kaniel and Mingelgrin found that stocks with unusually high volume tend to do better over the following month, the high-volume return premium; Amihud found that less liquid stocks earn higher returns, as pay for the cost of trading them; and Kyle's model defines price impact, the measure of a thin market used here. This study measures all three on one universe, every Wednesday since 1995, against the same day's median stock.
2Data and method
2.1Sample design
Population and frame. The population is US common stocks listed on the NYSE and Nasdaq; funds, ETFs, trusts, preferred shares, warrants and units are excluded. The sampling frame is a fixed universe of 1,767 companies, drawn once when the platform was built and not re-sampled since, in two strata. Stratum 1 is a census of the S&P 500: every company in the index at any time since 1996 whose price history could be recovered, 1,027 companies of which 413 have since delisted; it holds 76% of the index's members in 1996 and at least 96% in every year from 2010. Stratum 2 is 740 large and mid-sized companies outside the index, selected in proportion to the market's sector weights from the stocks that passed a minimum share price of $15 and a minimum average daily trading value of $25 million; 9 of them have since delisted.
Sample. The unit of observation is a stock-week: one stock on one Wednesday. A stock enters a Wednesday's cross-section when it has a valid close that day, a value of the signal and a measured outcome; bars flagged as bad data and returns that cross a change of issuer are left out. This study's sample is n = 1,763 stocks of the 1,767: 1,889,758 stock-weeks on 1,634 Wednesdays from Jan 4, 1995 to Aug 19, 2026, a median of 1,148 stocks per Wednesday (range 820 to 1,371). 331,954 stock-weeks (17.6%) come from the 422 companies that have since delisted. Every stock and every Wednesday carries equal weight.
Representativeness. Table 1 gives the sample by GICS sector beside the S&P Composite 1500: 7.9% of companies would have to change sector for the two to match exactly. By latest market value, 55% of the active companies are large (at least $10bn), 38% mid ($2–10bn) and 6% small. Because Stratum 2 was chosen from companies listed at construction, its history carries survivorship bias. Section 4 repeats the lead result among the stocks that were S&P 500 members on each date, the part of the sample largely free of that bias.
Companies in this study's sample, of them those since delisted, the size of the active companies, the sample's share of stock-weeks and the sector's share of the S&P Composite 1500.
| Sector | Sample | Delisted | Large | Mid | Small | Stock-weeks | S&P 1500 |
|---|---|---|---|---|---|---|---|
| Information Technology | 282 | 67 | 122 | 76 | 18 | 13.9% | 12.7% |
| Financials | 256 | 60 | 114 | 74 | 8 | 14.6% | 17.2% |
| Industrials | 255 | 53 | 126 | 67 | 11 | 15.9% | 17.5% |
| Health Care | 237 | 47 | 92 | 81 | 16 | 12.4% | 10.9% |
| Consumer Discretionary | 209 | 39 | 69 | 87 | 12 | 12.3% | 12.9% |
| Energy | 105 | 30 | 39 | 31 | 5 | 5.9% | 4.7% |
| Consumer Staples | 101 | 38 | 40 | 17 | 5 | 5.9% | 4.9% |
| Materials | 94 | 34 | 35 | 24 | 2 | 5.6% | 5.1% |
| Real Estate | 84 | 10 | 40 | 32 | 2 | 5.7% | 6.9% |
| Communication Services | 80 | 31 | 28 | 13 | 5 | 3.9% | 3.3% |
| Utilities | 60 | 13 | 35 | 12 | 0 | 4% | 4% |
| All sectors | 1,763 | 422 | 740 | 514 | 84 | 100% | 100% |
Sample: n = 1,763 of the 1,767 companies. Size by latest market value for the 1,338 active companies with one: large ≥ $10bn, mid $2–10bn, small < $2bn. S&P 1500 shares count the constituents of the S&P 500, MidCap 400 and SmallCap 600 (1,506 companies, lists read Sep 25, 2026). Sectors are each company's current GICS sector.
2.2Signals
| Signal | Definition | Since | Wednesdays |
|---|---|---|---|
| Volume against its own past year | The day's volume against its 20-session average, scored against the stock's own past year | Jan 4, 1995 | 1,634 |
| Volume jump, compared across stocks that day | The day's volume against its last month's average, compared with every other stock the same day | Jan 4, 1995 | 1,634 |
| Volume against its last month | The day's volume divided by its 20-session average | Jan 4, 1995 | 1,634 |
| Price impact, last five sessions | How far the price moved per unit of volume traded over the last five sessions (Kyle's lambda) | Jan 4, 1995 | 1,634 |
| Illiquidity, last month | Absolute daily return per dollar traded, averaged over the last 21 sessions (the Amihud measure) | Jan 4, 1995 | 1,634 |
| Bid-ask spread estimate, last three months | The spread implied by how daily prices bounce back and forth | Jan 4, 1995 | 1,634 |
2.3Measurement
Outcome. The return from the close on the Wednesday to the close 1, 5, 21 and 63 trading sessions later, on closes adjusted for splits and dividends, compared with the same day's median stock; half of all stocks beat the median by construction, so chance is 50 in 100 on every date. No delisting return is added.
Each Wednesday the stocks are ranked on the signal. The study reports how often each fifth of that ranking beat the median stock, and the rank correlation (IC) between the signal and the return that followed.
The t-statistic uses non-overlapping dates only. A result is called proven when |t| is at least 3, it held in at least 60% of years and it is large enough to matter; with 1,470 tests across the studies, a looser bar would pass dozens by luck.
Full data and methods3Results
Unusual volume: a slightly better week
Every week we rank US stocks by how heavy their trading volume is against their own past year, and split them into fifths. The fifth on the heaviest volume beat the median stock over the next week 51 times in 100; the quietest fifth 49. Across the whole ranking the relation pointed the same way in 28 of 32 years, and it fades: over three months the heaviest-volume fifth beat the median stock 50 times in 100 and the quietest 50, no measurable edge. The effect has a name in the literature: the high-volume return premium.
Left, the quietest fifth; right, the heaviest volume against the stock's own past year. Out of 100; the dashed line is chance.
Left, the quietest fifth; right, the heaviest volume against the stock's own past year. Out of 100; the dashed line is chance.
Illiquidity: more over three months, at a higher cost to trade
Illiquidity here is how far a stock's price moves per dollar traded, averaged over the last month: the Amihud measure. Over the next three months the least liquid fifth beat the median stock 52 times in 100 and the most liquid fifth 48; across the whole ranking the relation pointed the same way in 25 of 32 years. On average over those three months the least liquid fifth returned 3.30 percentage points more than the average stock, and the most liquid 1.78 points less. Over a week there was no measurable edge. These returns are before trading costs, and costs are highest for exactly these stocks.
Left, the most liquid fifth; right, the least liquid. Out of 100; the dashed line is chance.
Left, the most liquid fifth; right, the least liquid. Out of 100; the dashed line is chance.
Price impact: a slightly weaker week
Price impact, or Kyle's lambda, is how far a stock's price moved for each share traded over the last five sessions; a high reading means a thin market for that stock. The fifth with the highest impact beat the median stock over the next week 49 times in 100 and the lowest-impact fifth 50; across the whole ranking the relation pointed the same way in 27 of 32 years. A measured tendency, and a small one that does not last: over three months the highest- and lowest-impact fifths sat at 50 and 50 in 100, no measurable edge.
Left, the lowest price impact; right, the highest. Out of 100; the dashed line is chance.
Left, the lowest price impact; right, the highest. Out of 100; the dashed line is chance.
Each measure at each horizon
Every volume and liquidity measure on this page, at each horizon we followed. Proven means the relation cleared the bar set out under the method; faint is a tilt that is not reliable on its own; the rest showed no measurable edge. The three ways of measuring unusual volume agree with each other.
| Measure | A higher reading | the next session | the next week | the next month | the next three months | Same way, at its best horizon |
|---|---|---|---|---|---|---|
| Volume against its own past year | did better | Proven | Proven | Faint | No measurable edge | 28 of 32 yearsthe next week |
| Volume jump, compared across stocks that day | did better | Proven | Proven | Faint | No measurable edge | 29 of 32 yearsthe next week |
| Volume against its last month | did better | Faint | Proven | Faint | No measurable edge | 26 of 32 yearsthe next week |
| Price impact, last five sessions | did worse | Faint | Proven | Faint | No measurable edge | 27 of 32 yearsthe next week |
| Illiquidity, last month | did better | No measurable edge | No measurable edge | Faint | Proven | 25 of 32 yearsthe next three months |
| Bid-ask spread estimate, last three months | did worse | Faint | No measurable edge | No measurable edge | No measurable edge | 22 of 32 yearsthe next session |
For the curious: the rank correlation (IC) between the reading and the next week's return was 0.011 for unusual volume (t-statistic 8.6) and -0.011 for price impact (-6.3); with the next three months' return it was 0.044 for illiquidity (3.9).
What this means for any one stock
Every relation here is a tilt across hundreds of stocks at once. For any one stock it is a nudge, not a forecast of its price: in the least liquid fifth, 52 stocks in 100 beat the median stock over the next three months, against 48 in 100 in the most liquid fifth. Research, not advice.
The spread itself, and the cost of trading
An estimate of the bid-ask spread, read from how daily prices bounce back and forth, came with a slightly weaker next session: a faint tilt, not reliable on its own, and no measurable edge over a week or longer. The spread is better read as a cost. Wide spreads and high price impact are what an illiquid stock charges to trade, and none of the returns on this page deduct them.
Short-term reversal: the strongest relation we measured4Robustness
In Table 3, unusual volume held up best over the next week: 0.011 (t 8.6) overall and 0.009 (t 4.8) among S&P 500 members on the date, with the same sign in 11 of 11 sectors and 3 of 3 decades, but it shrank decade by decade and is too small to measure in the latest. Price impact over a month kept its sign in 11 of 11 sectors but stayed faint overall (-0.009, t -2.8) and was not measurable among S&P 500 members (-0.007, t -1.1). The three-month illiquidity premium, 0.044 (t 3.9) with the same sign in 10 of 11 sectors, is not measurable among S&P 500 members (0.008, t 0.8): it rests on the smaller, less traded companies outside the index, where trading costs are highest.
Rank IC and its t-statistic on non-overlapping dates.
| Sample | IC | t | Wednesdays |
|---|---|---|---|
| Volume against its own past year, next week | |||
| All stocks, whole period | 0.0113 | 8.6 | 1,633 |
| By decade: 1995–2004 | 0.0208 | 9.2 | 516 |
| By decade: 2005–2014 | 0.0110 | 5.0 | 518 |
| By decade: 2015– | 0.0033 | 1.4 | 599 |
| S&P 500 members on the date | 0.0089 | 4.8 | 1,581 |
| Sectors with the overall sign | 11 / 11 | ||
| Price impact, last five sessions, next month | |||
| All stocks, whole period | -0.0093 | -2.8 | 1,630 |
| By decade: 1995–2004 | -0.0137 | -2.4 | 516 |
| By decade: 2005–2014 | -0.0067 | -2.4 | 518 |
| By decade: 2015– | -0.0077 | -1.2 | 596 |
| S&P 500 members on the date | -0.0071 | -1.1 | 1,578 |
| Sectors with the overall sign | 11 / 11 | ||
| Illiquidity, last month, next three months | |||
| All stocks, whole period | 0.0437 | 3.9 | 1,621 |
| By decade: 1995–2004 | 0.0634 | 3.2 | 516 |
| By decade: 2005–2014 | 0.0367 | 2.1 | 518 |
| By decade: 2015– | 0.0326 | 1.4 | 587 |
| S&P 500 members on the date | 0.0078 | 0.8 | 1,569 |
| Sectors with the overall sign | 10 / 11 |
5Limitations
- The liquidity premium is measured before costs, and it sits where costs are largest: the least liquid fifth is the most expensive to trade, so its measured edge overstates what could be kept.
- Volume is measured without asking who traded or why; news-driven volume, index rebalancing and other mechanical flows are not told apart.
- Before costs. Averages exclude trading costs, taxes and market impact.
- Same-close timing. Returns start at the close the signal is computed from; a real trade would start later.
- Survivors among smaller companies. The non-index names were chosen from companies listed when the universe was built; no delisting returns are added.
6Conclusion
Volume and liquidity carry small tilts, not forecasts. Stocks trading on unusually heavy volume for their own past year did slightly better over the next week, as the high-volume return premium describes, and the effect faded within three months. Less liquid stocks did slightly better over three months, but that premium was not measurable among S&P 500 members and is measured before the higher cost of trading the stocks that carry it. High price impact came with a slightly weaker week and little after it. None of this says where any one stock's price goes next.
References
- Amihud, Y. (2002). Illiquidity and Stock Returns: Cross-Section and Time-Series Effects. Journal of Financial Markets 5(1).
- Kyle, A. S. (1985). Continuous Auctions and Insider Trading. Econometrica 53(6).
- Gervais, S., Kaniel, R. and Mingelgrin, D. H. (2001). The High-Volume Return Premium. Journal of Finance 56(3).
- Benjamini, Y. and Hochberg, Y. (1995). Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. Journal of the Royal Statistical Society, Series B 57(1).
- Grinold, R. C. and Kahn, R. N. (2000). Active Portfolio Management, 2nd ed. McGraw-Hill.
- Harvey, C. R., Liu, Y. and Zhu, H. (2016). … and the Cross-Section of Expected Returns. Review of Financial Studies 29(1).
- Shumway, T. (1997). The Delisting Bias in CRSP Data. Journal of Finance 52(1).
Appendix A. Questions readers ask
- Does high trading volume predict stock returns?
- A little, and briefly. On US stocks since 1995, the fifth trading on the heaviest volume for its own past year beat the median stock over the next week 51 times in 100, against 49 for the quietest fifth; across the whole ranking the relation pointed the same way in 28 of 32 years. By three months there was no measurable edge.
- Is there a liquidity premium in stocks?
- In this data, over three months, yes, and it is small. The least liquid fifth of US stocks beat the median stock 52 times in 100 over the next three months and the most liquid fifth 48; since 1995 the relation pointed the same way in 25 of 32 years. The returns are before trading costs, which are highest for exactly those stocks.
- What is the Amihud illiquidity measure?
- A stock's absolute daily return divided by the dollar volume traded that day, averaged here over the last month: how far a dollar of trading moves the price. Higher means less liquid. Its least liquid fifth beat the median stock over three months 52 times in 100, its most liquid fifth 48.
- What is Kyle's lambda?
- Kyle's lambda is price impact: how far a stock's price moves for each share traded, here over the last five sessions. A high lambda means a thin market. The highest-impact fifth beat the median stock over the next week 49 times in 100 and the lowest-impact fifth 50: a measured tendency, and a small one.
- Does the bid-ask spread predict stock returns?
- Only faintly. A spread estimated from how daily prices bounce back and forth came with a slightly weaker next session, a tilt that is not reliable on its own, and showed no measurable edge over a week, a month or three months. It is better read as a cost of trading.
- Can one stock's volume tell you where its price goes next?
- No. The tilt is measured across hundreds of stocks at once: 51 in 100 beat the median stock over the next week in the heaviest-volume fifth, 49 in 100 in the quietest. For any single stock that is a nudge, not a forecast of its price.
Data availability and citation
Every figure in this paper is quoted from one frozen snapshot, published as JSON with the sample description, the robustness results and the test counts. The same snapshot feeds the research console, so the two cannot disagree.
Opulence Alpha Research (2026). Does trading volume or liquidity predict stock returns? Opulence Alpha Studies, Sep 25, 2026. https://opulencealpha.ai/studies/volume-and-liquidity
Research, not advice. A measured tendency across hundreds of stocks is a nudge for any one of them, never a forecast of its price.