AI Research MARA

MARA overnight gaps: Monday (Fri→Mon) vs Tue–Fri over the last ~3 years

MARA’s Monday opens are meaningfully larger. Over the past ~3 years the average absolute Friday→Monday gap was 3.11% versus 2.17% for Tue–Fri overnights — the weekend appears to compress larger, unhedgeable Bitcoin moves into Monday’s opening price. The difference is robust; the full analysis below quantifies magnitude, distribution, and statistical confidence.

I tested this by computing absolute overnight gaps from regular‑hours minute bars across the most recent 36 months and comparing the Friday→Monday distribution to other weeknights with a Welch two‑sample t‑test. Scroll down for the charts, detailed statistics, and distributional evidence that explain where the lift shows up (center, tails, and per‑weekday breakdown).

The research question

For MARA over the past ~3 years, are Monday opening gaps systematically wider than other weekdays — does a full weekend of unhedgeable Bitcoin moves get crammed into Monday's open while the stock sits closed? Thesis: the average absolute Friday-close-to-Monday-open gap runs materially larger than Tuesday-through-Friday overnight gaps, so MARA's biggest overnight risk is the weekend it can't trade.

How this was measured

Minute bars were filtered to US regular trading hours (Mon–Fri, 09:30–16:00 ET). For each session, the day’s RTH open is the first 09:30 bar’s open and the RTH close is the last pre‑16:00 bar’s close. Overnight gap for day t is open_t ÷ close_{t−1} − 1; we analyze absolute gaps. The sample is restricted to the most recent 36 months of available data. We compare the Monday (Friday→Monday) absolute gap distribution to the aggregate Tue–Fri distribution using a Welch two-sample t-test and summarize per-weekday statistics.

The key numbers

Coverage window (trading days)
751
2023-07-03 to 2026-06-30
Monday mean absolute gap
3.1149%
N=143 Mondays
Tue–Fri mean absolute gap
2.1675%
N=608 Tue–Fri days
Mean difference (Mon − Tue–Fri)
0.9474%
Mean ratio (Mon / Tue–Fri)
1.437
Welch t-stat (Mon vs Tue–Fri)
3.815
Positive favors larger Monday gaps
Welch p-value (two-sided)
0.0002
p=0.0002 < 0.05 → Monday gaps differ from Tue–Fri

Reading the numbers

MARA's mean absolute Monday overnight gap is 0.031149 (N=143) versus 0.021675 for Tue–Fri (N=608) — about a 1.437× larger weekend jump, and the difference is statistically significant (p=0.000187).

The charts

MARA absolute overnight gap by weekday
What this chart says

The box plot lays out the whole distribution by weekday: Monday's mean absolute gap is 0.031149 (n=143) while the other weekdays cluster around 0.0206–0.0229. Notice the much larger upper tail on Mondays — a Monday max of 0.1896 compared with next-highest Friday max 0.149 — which shows more extreme weekend-driven openings. In short, Mondays not only sit higher on average but also produce the biggest outliers, consistent with concentrated weekend risk.

Mean absolute overnight gap — Monday vs Tue–Fri
What this chart says

This bar chart highlights the central comparison: mean absolute gap of 0.031149 on Monday versus 0.0216749 for Tue–Fri, a mean difference of 0.009474457 and a ratio of 1.437116. The Welch t-statistic is 3.8152496 with a two-sided p-value of 0.00018697, meaning the larger Monday mean is unlikely to be random. That supports the thesis that weekend Bitcoin moves tend to be compressed into Monday's open more than a typical weekday overnight.

Weekday summary of absolute overnight gaps

WeekdayNMeanMedianStd
Monday1430.03110.02580.028
Tuesday1560.02150.01610.0206
Wednesday1530.02170.01790.0196
Thursday1480.02290.01820.0228
Friday1510.02060.01660.019

The takeaway

Yes — Monday (Friday→Monday) opens are meaningfully larger. The average absolute Friday→Monday gap is 3.11% versus 2.17% for Tue–Fri overnights. That’s a mean difference of +0.95 percentage points and roughly a 1.44× lift (Mon/Tue–Fri = 1.437), so a full weekend of unhedgeable Bitcoin moves tends to be compressed into Monday’s open. This result is statistically strong: 143 Mondays vs 608 other overnights across the 751 trading‑day window, Welch t = 3.82 with p ≈ 0.00019 — only about a 2‑in‑10,000 chance the gap difference is pure luck. The shift shows up in the center of the distribution too (Monday median 2.58% vs midweek medians around 1.6–1.8%), so weekend risk raises both typical and tail overnight moves. Practical takeaway: treat the weekend as MARA’s largest overnight exposure — expect roughly 3% absolute moves into Monday opens versus ~2% on regular weeknights.

The fine print