From December 5, 2016 to November 2, 2026, a 3 x ATR chandelier stop on $QQQ was hit 47 times and a 2 x stop 70 times, about 4.7 and 7.1 times a year. The average trade lasted 30.5 sessions at 3 x and 17.2 at 2 x. Both cut the worst closing drawdown from 35.6% for buy-and-hold to 18.2% (3 x) and 14.0% (2 x). The cost was steep: price gains of 204% and 151% against 484% for simply holding.
Most of that gap is whipsaw. In 87% of 2 x stop-outs and 85% of 3 x stop-outs, the rule bought $QQQ back at a higher price than it sold. The rules, the data checks, the full results and the limits follow. For how ATR and chandelier stops work, start with the ATR trailing stop guide. More explainers are in the Analysis section.
The test rules
The test isolates one variable, the ATR multiple. Everything else is identical for 2 x and 3 x.
- Data: daily open, high, low and close for $QQQ from Nasdaq.com historical data, November 2, 2016 to November 2, 2026, 2,513 sessions.
- ATR: Wilder's 14-session Average True Range. True range is the largest of high minus low, high minus the prior close, and low minus the prior close. ATR starts as the plain average of the first 14 true ranges and then updates as (prior ATR x 13 + true range) / 14, as described by StockCharts ChartSchool.
- Stop: the highest high of the last 22 sessions minus 2 or 3 x ATR(14), a chandelier-style stop. While long, the stop only ratchets up.
- Exit: sell at the close on the first session that closes below the stop.
- Re-entry: buy at the close on the first session that closes above the highest close of the previous 22 sessions.
- Start: long at the December 5, 2016 close of 116.60, the first session with 22 sessions of history. Buy-and-hold starts on the same day.
Data checks before the test. A split would show as a one-day price change of roughly 50%. The largest close-to-close moves in the series were -11.98% on March 16, 2020 and +12.00% on April 9, 2025. $SPY moved -10.94% and +10.50% on the same two days, so both were broad market sessions, not split artifacts. No gap between consecutive dates exceeded four calendar days, and every open and close sits inside its day's high-low range. The Nasdaq closes also matched Yahoo Finance's unadjusted closes on all 2,512 shared dates. The prices are not adjusted for dividends.
2x vs 3x ATR on QQQ: the results
| Dec 5, 2016 to Nov 2, 2026 | 2 x ATR | 3 x ATR | Buy-and-hold |
|---|---|---|---|
| Times the stop was hit | 70 | 47 | n/a |
| Stop-outs per year | 7.1 | 4.7 | n/a |
| Average holding period (sessions) | 17.2 | 30.5 | whole period |
| Median holding period (sessions) | 13 | 22 | whole period |
| Time invested | 48.3% | 57.6% | 100% |
| Trades closed at a gain | 38 of 70 (54%) | 27 of 47 (57%) | n/a |
| Price return | 150.6% | 204.4% | 484.1% |
| Annualized | 9.72% | 11.89% | 19.50% |
| Worst closing drawdown | -14.0% | -18.2% | -35.6% |
The 3 x stop held trades almost twice as long and finished with a price gain about 54 percentage points higher. The 2 x stop had the shallower worst drawdown, but it was stopped out 23 more times to get there. Both versions were in cash for 42% to 52% of the period, and that cash earned nothing in this test.
Stop-outs clustered in choppy years. The 2 x rule fired 11 times in 2024. The 3 x rule never fired more than 6 times in a calendar year.
Drawdown avoided: the four big declines
The case for a trailing stop rests on the large declines. Here is where each rule first exited during four major $QQQ declines in the window, measured close to close.
| Decline | Peak to trough (close) | 2 x first exit | 3 x first exit |
|---|---|---|---|
| 2018 Q4 | 186.74 (Aug 29) to 143.50 (Dec 24): -23.2% | 183.45 (Sep 5) | 181.81 (Sep 6) |
| 2020 Covid | 236.98 (Feb 19) to 169.30 (Mar 16): -28.6% | 230.27 (Feb 21) | 221.39 (Feb 24) |
| 2021-22 bear | 403.99 (Nov 19, 2021) to 260.10 (Dec 28, 2022): -35.6% | 397.48 (Nov 23, 2021) | 391.20 (Nov 26, 2021) |
| 2025 tariff drop | 539.52 (Feb 19) to 416.06 (Apr 8): -22.9% | 526.08 (Feb 21) | 513.32 (Feb 25) |
In three of the four, both rules exited within 2.8% (2 x) or 6.6% (3 x) of the peak close and stayed out until after the trough. The 2021-22 bear market was different. Both rules got out early, then re-entered five times on new 22-session closing highs before the December 2022 low. Each re-entry was stopped out again, and the worst single trade for each rule came in that stretch. At 2 x, a trade bought on October 25, 2022 at 284.21 was sold on November 2 at 265.68, a 6.5% loss. At 3 x, the same entry was sold a day later at 260.49, an 8.3% loss.
The whipsaw bill
Whipsaw is the cost of the exits that turned out to be noise. The data shows a lot of it:
- Immediate rebounds: the very next session closed above the exit price after 61% of 2 x exits and 57% of 3 x exits.
- Quick recoveries: $QQQ closed back above the exit price within 10 sessions after 89% of 2 x exits and 87% of 3 x exits.
- Buying back higher: 60 of 69 re-entries at 2 x (87%) and 39 of 46 at 3 x (85%) came above the prior exit price. The median re-entry was 2.53% above the exit at 2 x and 2.70% at 3 x.
- Real saves were rare: only 9 re-entries at 2 x and 7 at 3 x came at or below the exit price, but those were a median 8.4% and 8.3% lower. The February 2020 exits were followed by further closing declines of 26.5% (2 x) and 23.5% (3 x) before $QQQ closed back above the exit price. The February 21, 2025 exit at 2 x was followed by a further 20.9% decline.
That is the trade-off in numbers. A handful of exits did most of the drawdown work, and they were paid for by dozens of small round trips. The 3 x stop made fewer of those round trips.
Execution: closes land below the stop
A closing-basis rule never sells exactly at the stop. The median exit close was 0.62% below the stop level at 2 x and 0.59% at 3 x, and the worst cases were 3.26% and 3.52% below. Moving every exit to the next session's open did not change the picture. The stop count stayed at 70 and 47. Price returns were 165.0% (2 x) and 230.3% (3 x), and worst drawdowns were 15.0% and 15.5%.
Where the latest trade stands: both versions bought on September 18, 2026 at 721.45. The 2 x rule sold on October 13 at 743.57 and the 3 x rule on October 14 at 728.95. Under the re-entry rule, $QQQ would need to close above 759.66, the October 6 closing high. That is 11.54% above the November 2, 2026 close of 681.09.
Limits of this backtest
- Price-only data. No dividends, which understates buy-and-hold more than the stop rules, since buy-and-hold was always invested.
- No costs, taxes or interest. Commissions, spreads and taxes on 70 or 47 round trips are excluded. So is any yield on cash while out of the market.
- One ticker, one window, one re-entry rule. A different re-entry rule, such as one that waits longer, would change the stop counts and returns.
- Closing basis. Intraday stop orders would trigger on different days.
- History only. The 2016 to 2026 window was a strong decade for $QQQ, which favors staying invested.
For daily $QQQ trend coverage, see the Traders On Trend homepage.
Bull/Bear Verdict
Bull Case: Both multiples roughly halved the worst closing drawdown, from 35.6% to 18.2% (3 x) or 14.0% (2 x), and stepped aside early in the 2020 and 2025 declines. The 3 x version did it with 23 fewer stop-outs and an average holding period almost twice as long.
Bear Case: In a decade that rewarded holding, both rules left most of the 484% price gain on the table. At least 85% of re-entries came back in above the prior exit price, and the 2021-22 bear market produced repeated losing re-entries.
FAQ
Is 2x or 3x ATR better for QQQ?
Over December 2016 to November 2026 on closing prices, 3 x was stopped out 47 times against 70, held trades about 30 sessions against 17, and kept more of the price gain (204% against 151%). 2 x had the smaller worst drawdown, 14.0% against 18.2%.
How often does a 3x ATR stop get hit on QQQ?
About 4.7 times a year in this test, ranging from 3 to 6 times per calendar year from 2017 to 2025. A 2 x stop averaged 7.1 times a year.
Did the ATR stop beat buying and holding QQQ?
Not on return. Buy-and-hold gained 484.1% in price over the period, against 204.4% and 150.6% for the stop rules. The stops did cut the worst closing drawdown by about half or more.
Why did the ATR stop lose money in 2022?
The rule re-entered five times on new 22-session closing highs during the 2021-22 decline, and each bounce faded. The worst trades for both multiples came in late October and early November 2022.