Drawdown recovery math comes down to one formula: a loss of L needs a gain of L ÷ (1 − L) to get back to even. A 10% loss needs +11.1%, a 20% loss needs +25%, and a 50% loss needs +100%. In the November 2016 to November 2026 data, the deepest closing drawdowns were −42.3% for $IWM, −37.1% for $DIA, −35.6% for $QQQ and −34.1% for $SPY. At the November 4, 2026 close, $QQQ sat 8.9% below its highest close in the data set.
That asymmetry is the most important arithmetic in trading. Losses come off a bigger base than the gains that repair them. The deeper the hole, the steeper the climb, and the curve bends hard after about −30%.
This piece puts the formula next to ten years of real ETF closes. It covers how deep the big drawdowns went, how long the climb back took, and where each index fund stands now.
The Formula: Why a Loss Needs a Bigger Gain
Start with $100. Lose 20% and $80 is left. To get back to $100, that $80 has to grow by $20, and $20 ÷ $80 = 25%. The loss was measured against $100, but the repair is measured against $80. That is the whole trick.
In general, required gain = 1 ÷ (1 − L) − 1, which simplifies to L ÷ (1 − L). StockCharts' write-up on bear markets and drawdowns uses the same expression, Percent Drawdown ÷ (1 − Percent Drawdown), and notes that a 50% loss takes a 100% gain to get back to even.
| Loss | $100 becomes | Multiplier to get back to $100 | Required gain |
|---|---|---|---|
| −5% | $95.00 | 100 ÷ 95 = 1.0526 | +5.3% |
| −10% | $90.00 | 100 ÷ 90 = 1.1111 | +11.1% |
| −15% | $85.00 | 100 ÷ 85 = 1.1765 | +17.6% |
| −20% | $80.00 | 100 ÷ 80 = 1.2500 | +25.0% |
| −25% | $75.00 | 100 ÷ 75 = 1.3333 | +33.3% |
| −30% | $70.00 | 100 ÷ 70 = 1.4286 | +42.9% |
| −40% | $60.00 | 100 ÷ 60 = 1.6667 | +66.7% |
| −50% | $50.00 | 100 ÷ 50 = 2.0000 | +100.0% |
| −60% | $40.00 | 100 ÷ 40 = 2.5000 | +150.0% |
| −75% | $25.00 | 100 ÷ 25 = 4.0000 | +300.0% |
Read the right-hand column. Up to about −15%, the required gain is only a few points above the loss. By −50%, it is double the loss, and by −75% it is four times the loss. Small losses are cheap to repair, but large ones change the math entirely.
The Real Drawdowns: $SPY, $QQQ, $IWM and $DIA, 2016–2026
The definition here follows the standard one. Fidelity's research glossary describes maximum drawdown as the peak-to-trough percentage decline over a period. A drawdown starts at a closing high. It ends on the first close at or above that high. The trough is the lowest close in between. The table lists every drawdown of 15% or more in the data, measured on daily closes.
| ETF | Peak close | Trough close | Depth | Gain needed from trough | Sessions peak to trough | Sessions trough to recovery | Recovered |
|---|---|---|---|---|---|---|---|
| $SPY | Sep 20, 2018 ($293.58) | Dec 24, 2018 ($234.34) | −20.2% | +25.3% | 65 | 85 | Apr 29, 2019 |
| $SPY | Feb 19, 2020 ($338.34) | Mar 23, 2020 ($222.95) | −34.1% | +51.8% | 23 | 103 | Aug 18, 2020 |
| $SPY | Jan 3, 2022 ($477.71) | Oct 12, 2022 ($356.56) | −25.4% | +34.0% | 195 | 318 | Jan 19, 2024 |
| $SPY | Feb 19, 2025 ($612.93) | Apr 8, 2025 ($496.48) | −19.0% | +23.5% | 34 | 55 | Jun 27, 2025 |
| $QQQ | Aug 29, 2018 ($186.74) | Dec 24, 2018 ($143.50) | −23.2% | +30.1% | 80 | 78 | Apr 17, 2019 |
| $QQQ | Feb 19, 2020 ($236.98) | Mar 16, 2020 ($169.30) | −28.6% | +40.0% | 18 | 57 | Jun 5, 2020 |
| $QQQ | Nov 19, 2021 ($403.99) | Dec 28, 2022 ($260.10) | −35.6% | +55.3% | 277 | 243 | Dec 15, 2023 |
| $QQQ | Feb 19, 2025 ($539.52) | Apr 8, 2025 ($416.06) | −22.9% | +29.7% | 34 | 52 | Jun 24, 2025 |
| $IWM | Aug 31, 2018 ($173.02) | Mar 23, 2020 ($99.90) | −42.3% | +73.2% | 390 | 165 | Nov 13, 2020 |
| $IWM | Nov 8, 2021 ($242.56) | Oct 27, 2023 ($162.21) | −33.1% | +49.5% | 495 | 473 | Sep 18, 2025 |
| $DIA | Oct 3, 2018 ($267.95) | Dec 24, 2018 ($218.10) | −18.6% | +22.9% | 56 | 122 | Jun 20, 2019 |
| $DIA | Feb 12, 2020 ($295.72) | Mar 23, 2020 ($186.13) | −37.1% | +58.9% | 27 | 166 | Nov 16, 2020 |
| $DIA | Jan 4, 2022 ($367.87) | Sep 30, 2022 ($287.30) | −21.9% | +28.0% | 186 | 302 | Dec 13, 2023 |
| $DIA | Dec 4, 2024 ($450.94) | Apr 8, 2025 ($376.48) | −16.5% | +19.8% | 84 | 94 | Aug 22, 2025 |
Every one of these 14 drawdowns had recovered by the November 4, 2026 close. The "gain needed" column is the formula applied to real prices. For example, $SPY needed $338.34 ÷ $222.95 − 1 = +51.8% to climb out of March 2020.
What the Recovery Clock Actually Shows
Depth does not predict duration. The data makes that point bluntly.
- The deepest $SPY drop was the fastest round trip. The 2020 crash took $SPY down 34.1% in 23 sessions and back to its high 103 sessions after the trough. The shallower 2022 decline of 25.4% took 318 sessions from trough to recovery, more than three times as long.
- Small caps were stuck underwater the longest. $IWM peaked on November 8, 2021 and did not close at or above that level again until September 18, 2025. That is 968 sessions from peak to recovery.
- Time at the top is rare. Only 14.1% of $SPY sessions closed at a new running high in the data, against 14.8% for $QQQ, 11.9% for $DIA and 5.5% for $IWM.
- Time deep underwater varies widely. $IWM closed 10% or more below its running high in 44.8% of sessions. The figure was 21.3% for $QQQ, 17.3% for $SPY and 12.5% for $DIA.
The formula tells a trader how far the climb is. Only the tape decides how long it takes. A drawdown that needs +34% can take longer than one that needs +52% if the rally behind it is slow and choppy.
Where the Four ETFs Stand at the November 4, 2026 Close
"Highest close" below means the highest close in this data set, which starts November 4, 2016.
| ETF | Highest close in data | Nov 4, 2026 close | Distance from high | Gain needed to reclaim it | Lowest close since that high |
|---|---|---|---|---|---|
| $SPY | $779.09 (Oct 6, 2026) | $766.81 | −1.58% | +1.60% | $752.10 (Nov 2), −3.46% |
| $QQQ | $759.66 (Oct 6, 2026) | $691.95 | −8.91% | +9.79% | $681.09 (Nov 2), −10.34% |
| $IWM | $305.09 (Aug 14, 2026) | $284.02 | −6.91% | +7.42% | $270.44 (Oct 16), −11.36% |
| $DIA | $542.81 (Aug 5, 2026) | $533.11 | −1.79% | +1.82% | $508.55 (Sep 30), −6.31% |
Two of the four are in live double-digit drawdowns that have not recovered. $QQQ fell 10.3% from its October 6 high to its November 2 close. $IWM fell 11.4% from its August 14 high to its October 16 close. $SPY and $DIA sit within 2% of their highs. At this depth the asymmetry is still mild, and $QQQ's 8.91% gap needs a 9.79% gain to close.
How Traders Use Drawdown Math
Here is a hypothetical $50,000 account. It drops 20% to $40,000. Getting back to $50,000 means making $10,000 on $40,000, which is +25%. The same 20% loss on a single position is just as hard to repair.
- Cap the hole before it opens. Traders use stops and position size to keep any single loss in the cheap zone of the table. The position sizing formula turns a stop distance into a share count, and an ATR trailing stop sets that distance from volatility.
- Judge strategies by drawdown, not just return. Two systems with the same average return are not equal if one needs a +55% rally to recover its worst drop.
- Respect the time cost. The $IWM case shows a recovery can take years, and capital tied up underwater is capital not compounding.
Method and Limits
Prices are daily closes from Nasdaq.com historical data for $SPY, $QQQ, $IWM and $DIA, from November 4, 2016 to November 4, 2026 (2,513 sessions each). Prices are not dividend-adjusted, so every figure is price-only; with dividends counted, drawdowns would typically be somewhat shallower and shorter. "Sessions" counts trading days, the running high resets only on a close at or above the prior high, and intraday highs and lows are ignored.
This is one 10-year window on four surviving ETFs, with no costs, taxes or slippage. Past drawdowns describe what happened, not what comes next.
Bull/Bear Verdict
Bull Case: Where it helps: the formula is exact and turns a vague fear into a number. A −10% hole needs +11.1%, a −34.1% hole needs +51.8%, and every 15%-plus drawdown in these four ETFs since 2016 has been repaired. That gives traders a hard reason to keep single losses small.
Bear Case: Where it misleads: the formula says nothing about time. $SPY needed 318 sessions to repair a 25.4% drop but only 103 to repair a 34.1% drop, and $IWM spent 968 sessions underwater. Price-only data on surviving ETFs over one decade also flatters how reliably recoveries arrive.
FAQ
How much do you need to gain to recover a 30% loss?
+42.9%. The arithmetic is 100 ÷ 70 = 1.4286, so the remaining capital has to grow 42.9% to return to the starting value.
What was the worst $SPY drawdown from 2016 to 2026?
On closing prices, it was −34.1%, from $338.34 on February 19, 2020 to $222.95 on March 23, 2020. $SPY closed at or above that high again on August 18, 2020, 103 sessions after the trough.
Is drawdown measured on closing prices or intraday prices?
Both are used. This analysis uses daily closes, which give smaller figures than intraday highs and lows would, because the intraday extremes are skipped.
How far is $QQQ from its high right now?
At the November 4, 2026 close of $691.95, $QQQ was 8.91% below its October 6, 2026 close of $759.66, so it needs +9.79% to reclaim that level. More market breakdowns are on the Markets page and the Traders On Trend homepage.