I'll analyze QQQ's rolling 5-day returns over the last 6 months. Let me load the daily price data and compute the rolling returns.QQQ Rolling 5-Day Returns — Last 6 Months (Jan 15 – Jul 14, 2026)

Over 118 rolling 5-day observations, the market was broadly constructive but with notable bouts of selling:

Key takeaways:

• QQQ posted a 5-day drop of more than 1.45% on 27 days, or roughly 23% of all observations — about once every 4–5 trading days.
• Severe selloffs (down >3% over 5 days) occurred 14 times (12%) — roughly twice a month.
• Drops in the –1.45% to –3% range hit 13 times (11%).
• The index rose over any given 5-day window 54% of the time, with modest declines of less than 1.45% accounting for another 23%.

In short: about half the time QQQ was up over rolling 5-day windows; when it sold off, roughly half those declines crossed the –1.45% threshold, with severe >3% drops happening about once every two weeks.

QQQ Rolling 5-Day Returns — Last Month (Jun 15 – Jul 14, 2026)

The last month was notably choppier than the 6-month trend. Drops exceeding 1.45% over a 5-day window jumped from ~23% (6-month average) to 35% — occurring 7 out of 20 days. Severe >3% drops also ticked up to 15% vs. 12% over the broader period.

QQQ Rolling 15-Day Returns — Last 6 Months (Jan 15 – Jul 14, 2026)

Over 108 rolling 15-day observations:

Contrast with the 5-day window:

• The win rate is similar — 56% up vs 54% for 5-day — but the distribution of losses shifts notably.
• Deep selloffs become much more prominent: >3% drops over 15 days hit 20% of the time (vs 12% for 5-day), meaning roughly one in every five rolling windows saw a material drawdown.
• Smaller dips (0 to –1.45%) shrink dramatically — just 13% vs 23% for 5-day. Over a 3-week stretch, mild declines naturally compound or reverse, so they're less common.
• Total drops exceeding –1.45% rose to 31% (vs 23% for 5-day) — about one in every three rolling 15-day windows.

In short: the longer window doesn't change the odds of being up, but when QQQ does decline over 3 weeks, it tends to be more serious — large >3% drawdowns happen nearly twice as often.