GME top‑decile volume sessions: bullish or bearish? (past ~3 years)
When a stock’s most extreme volume days happen, the market instinct is to guess the crowd’s mood. For GME, the meme-trading narrative says the herd surges on euphoric rallies, not panic dumps. This study tested that hypothesis across 751 trading days, isolating the 76 sessions that fell into the top decile of daily volume.
The numbers lean in the narrative’s favor: 57.9% of those high-volume days were up, versus 47.9% across all days, with a mean return of +3.86% compared to a slight loss on quiet days. But the statistical test lands at p = 0.095—above the conventional threshold for significance. The pattern holds directionally, yet the evidence is too thin to call it a reliable edge.
Below we walk through the full scatter chart, the return distributions, and the t‑test output so you can judge the signal yourself.
For GME over the past ~3 years, do its top‑decile daily‑volume sessions land predominantly on up days or on down days? The thesis is that volume extremes cluster on big up days because the meme‑trading crowd surges in on euphoric rallies rather than panic dumps, so GME's liquidity spikes are bullish events.
How this was measured
Resampled GME intraday bars to daily OHLCV. Defined top‑decile volume days as those with volume ≥ the 90th percentile of the sample. Compared the proportion of up days (return > 0) and the mean daily return between high‑volume and low‑volume subsets. Welch's t‑test on returns quantifies if the mean differs significantly. The scatter chart marks all top‑decile days with their dates.
The key numbers
Reading the numbers
Of the 76 highest-volume trading days, 44 were up days and 32 were down days — that's about 58% up, compared to 48% up on all other days. So volume extremes are more likely on up days, though the difference isn't statistically slam-dunk (p = 0.095).
The charts
The box plot compares daily returns for GME's top-decile volume days versus the rest. The high-volume days show a much wider spread, with returns ranging from a -55.7% drop to a +112.2% surge, while ordinary days stay mostly between -14.6% and +25.1%. The average (mean) return on high-volume days is +0.039%, versus -0.002% on other days, suggesting that extreme volume sessions tilt slightly positive on average.
This bar chart directly answers the question: among the 76 top-decile volume sessions, 44 were up days and 32 were down days. That is a 58% majority on the up side, supporting the idea that GME's heaviest trading days are more often euphoric rallies than panic exits.
The scatter plot maps every trading day's volume against its daily return, with the top-decile volume days highlighted. You can see that the largest volume spikes (above 100 million shares) are associated with the most extreme returns — both positive and negative — but the biggest clusters of high volume are in positive territory. This reinforces that the heaviest liquidity events tend to coincide with big up moves.
Top 20 highest‑volume days (descending volume)
| date | volume | daily_return | direction |
|---|---|---|---|
| 2024-06-07 | 280,383,291 | -0.5567 | Down |
| 2024-05-14 | 207,972,967 | 0.374 | Up |
| 2024-06-06 | 207,533,397 | 0.9034 | Up |
| 2024-05-13 | 188,629,146 | 1.1219 | Up |
| 2025-06-12 | 181,186,052 | -0.1156 | Down |
| 2024-06-03 | 166,800,260 | 0.3252 | Up |
| 2024-09-20 | 149,700,688 | 0.1127 | Up |
| 2024-06-12 | 148,670,592 | -0.1268 | Down |
| 2024-06-11 | 147,821,907 | 0.191 | Up |
| 2024-05-15 | 132,718,015 | -0.2939 | Down |
| 2024-06-10 | 132,493,118 | -0.0788 | Down |
| 2024-06-13 | 108,087,135 | 0.0899 | Up |
| 2024-05-28 | 106,278,602 | 0.0791 | Up |
| 2025-03-27 | 100,320,169 | -0.1241 | Down |
| 2024-05-17 | 97,337,230 | -0.2539 | Down |
| 2024-06-17 | 90,688,674 | -0.1216 | Down |
| 2024-06-14 | 83,911,757 | 0.0032 | Up |
| 2024-05-16 | 76,667,913 | -0.2025 | Down |
| 2024-06-05 | 74,212,604 | 0.2043 | Up |
| 2025-06-13 | 63,377,203 | -0.0031 | Down |
The takeaway
Yes, GME's top-decile volume sessions over the past three years are more likely to be up days than down days, but the edge is modest. Among 76 extreme-volume days, 57.9% were up, compared to 47.9% for all trading days. The average return on those heavy-volume days was +3.86%, while the rest of the period averaged a slight loss of -0.19%. However, the statistical test gives a p-value of 0.095, meaning there is about a 9.5% chance this difference is just random luck — not enough to call it a clear signal. In practical terms, the thesis that volume extremes cluster on euphoric rallies holds directionally, but the evidence is too weak to bank on. This is a lean, not a lock.
The fine print
- Only 76 top-decile days — small sample means wide confidence intervals around that 57.9% figure.
- High-volume days are often bunched in meme-event windows, so the observations aren't independent (violates a key assumption of the t-test).
- The analysis doesn't account for broader market moves — SPY returns on those days could partly explain GME's volume and direction.