Here Is What Followed the Last 47 Times
QQQ closed at −11.3% from its 52-week high on Tuesday 29 July, which puts it in correction territory. It climbed back out the next session. That round trip is more typical than you think.
We backtested every time QQQ has first closed 10% below its 52-week high since it began trading in March 1999, then measured what the ETF did one week, one month and six months later. Signals within one month of an earlier signal were dropped, so each distinct selloff is represented only by its first crossing. That left 48 unique occasions out of 100 crossings, across 6,890 trading sessions.
Every time frame beat the market’s own baseline. Six months out the median gain was +15.7% against +8.6% starting from any other day, and QQQ was higher 87% of the time against 73%.
| Time Frame | Times | Median | Average | Red / Green Frequency | Worst | Best |
|---|---|---|---|---|---|---|
| 1 week later | 47 | +2.7% | +2.9% | ▲74.5% / ▼25.5% | −6.1% | +14.3% |
| Any Random 1 week Period | 6,885 | +0.5% | +0.2% | ▲57.4% / ▼42.4% | −24.6% | +20.7% |
| 1 month later | 47 | +4.7% | +3.4% | ▲68.1% / ▼31.9% | −14.5% | +23.0% |
| Any Random 1 month Period | 6,869 | +1.6% | +1.0% | ▲62.4% / ▼37.5% | −31.1% | +41.6% |
| 6 months later | 46 | +15.7% | +14.4% | ▲87.0% / ▼13.0% | −42.4% | +68.7% |
| Any Random 6 month Period | 6,764 | +8.6% | +6.3% | ▲72.7% / ▼27.3% | −60.0% | +95.7% |
The typical path after these selloffs shows an initial bounce that often settles. One month later, QQQ has closed higher only 68% of the time, which is only a modest edge over the 62% baseline from any random day. By six months, the recovery becomes more consistent, with the index more reliably higher.
40 of the 46 completed entries were higher six months on. The six that were not tell you where the risk lives.
| Entered | 1 Week | 1 Month | 6 Months | Deepest Within 6 Months |
|---|---|---|---|---|
| 15 Mar 2000 | +10.8% | −14.5% | −9.1% | −36.5% |
| 4 Jan 2008 | −2.8% | −9.8% | −7.2% | −25.0% |
| 20 May 2008 | −0.0% | −1.0% | −42.4% | −47.9% |
| 21 Aug 2015 | +3.1% | +1.7% | −0.8% | −16.4% |
| 20 Jan 2022 | −5.7% | −5.6% | −16.5% | −32.8% |
| 31 Mar 2022 | −2.3% | −12.2% | −26.3% | −33.8% |
Most of them announced themselves early. Four of the six were already down a week in, and five of the six were down at a month. May 2008 is the exception that should worry you: flat after a week, barely down after a month, and then 42.4% lower over the following six.
That is a thin sample to lean on, but it does mean the entries that failed mostly failed early rather than quietly.
The data also shows how briefly QQQ tends to stay below the 10% line. The median episode lasted 14 sessions. Of the 47 instances, 11 were resolved within just two sessions. Only one stretched into a long decline: the March 2000 episode lasted 801 sessions, a little over three years.
This is why the average results look strong. Most of the time the 10% threshold is touched and quickly abandoned within the same week. The sample is therefore dominated by shallow dips inside ongoing uptrends. Genuine, prolonged bear markets are rare, and those are the only ones that create meaningful downside risk.
Buying QQQ after these corrections has outperformed a random entry at every time frame measured. Six months later, the index was higher 87% of the time with a median gain of 15.7%, compared with 73% and 8.6% for a random entry.
The risk sits in the shape of the losses. The six failures clustered in 2000, 2008, and 2022, and they were not small. May 2008 lost 42.4% over six months, and March 2022 lost 26.3%. A signal that works 87% of the time but loses heavily the other 13% has to be sized for the 13%.
Testing popular ideas without cherry-picking
Browse Reality Check
Comments
Loading comments…