Drawdown Recovery Calculator
Michael Whitman
·
Trading since 2017, full-time since 2020
·
Published 2026-09-08
·
Updated
·
Reviewed
Drawdown recovery is the gain needed to return a portfolio to a previous high after it has fallen, and it is always larger than the fall itself because the gain is measured against a smaller base. A 30% fall requires a 42.86% gain, and a 50% fall requires 100%.
Getting back to the peak
Defaults take a 100,000 portfolio down 30%, recovering at an assumed 8% a year.
The years figure assumes a steady return from the bottom, which is not how recoveries happen — they arrive in bursts separated by nothing. Read it as the time a constant rate would take, not as a schedule.
Runs entirely in your browser. Nothing you type is sent anywhere or stored.
Covered on this page: TradingView.
The gain needed to recover from a drawdown is 1 ÷ (1 − fall) − 1. A 30% fall needs a 42.86% gain and a 50% fall needs 100%, because the recovery is measured from the smaller amount left at the bottom.
The time is the less certain half. In all 11 falls of 20% or more in the S&P 500 since 1950, the index got back to its old record sooner than a steady 8% a year would have: a median of 1.3 years from the bottom, where the calculator’s constant-rate estimate for the same falls is 5.3 years.
How the number is built
Two percentages that sound like they should match and do not, because they are measured against different amounts.
Gain needed = 1 ÷ (1 − fall) − 1
A 30% fall is measured against the peak; the recovery is measured against the bottom. The bottom is smaller, so the same absolute amount of money is a larger percentage of it.
A worked example
Take the defaults: $100,000 falling 30%, recovering at an assumed 8% a year.
The bottom is $70,000.
Getting back to $100,000 from $70,000 requires $30,000 — which is 42.86% of $70,000.
At 8% a year that takes 4.63 years.
The clean case to memorize is 50%. Halving requires a double, and a double is not what a diversified portfolio does quickly. That single pairing communicates the asymmetry better than any table.
The curve steepens
Walk down the scale and the requirement accelerates. A 10% fall needs 11.11%. A 20% fall needs 25%. A 30% fall needs 42.86%. A 50% fall needs 100%. A 70% fall needs 233.33%. A 90% fall needs 900%.
The first twenty points cost 25 points of recovery and the last twenty cost 667. That is the argument for every constraint a risk framework imposes, stated in one sequence.
Time behaves the same way. At 8% a year, recovering 10% takes 1.37 years, 30% takes 4.63, 50% takes 9.01 and 70% takes 15.64. The years are not proportional to the fall either.
What being under water is actually like
In the S&P 500 since 1950 the deepest bear market was a 56.8% fall, from October 2007 to March 2009, and the index did not close above that peak again until 28 March 2013. By the arithmetic above, a 56.8% fall needs about 131% to get back. Deep as it is, depth is not the most useful figure.
93% of trading days in that history closed below the previous record, and the longest wait for a new record was 7.5 years, from January 1973 to July 1980 — and the index still went on to new records. Being below a high is the ordinary condition of holding anything, not a signal that something has gone wrong.
Combining depth with duration is a fairer summary of an experience than a maximum drawdown alone. Two falls of the same depth, one recovered in months and one that took seven years, show the same maximum drawdown and are nothing alike to live through.
What deepens the hole
Fees and trading costs are charged on the way down as well. On this site’s arithmetic, 75 basis points a year removes 20.2% of a thirty-year pot — and during a drawdown that is money coming out of the base the recovery has to be measured from.
Selling into a fall converts a drawdown into a permanent loss. The recovery arithmetic only applies to a position still held; a position closed at the bottom has no denominator left to recover.
Why small losses are a different category
The flat part of the curve is where a loss stays a nuisance. Under about 20%, recovery needs roughly the same order of magnitude as the fall — 25% against 20% — and an ordinary year of returns covers it.
Past about 40% the two numbers separate permanently. A 40% fall needs 66.67% and a 60% fall needs 150%, so what changes is not the size of the problem but its type: one is a bad stretch and the other is a decade.
Which is what a maximum-loss rule is actually buying. Capping any single position’s contribution to portfolio decline through position sizing is not caution for its own sake; it is a decision to stay in the region where ordinary returns can undo an error.
The original data
Every S&P 500 fall of 20% or more since 1950 beat the calculator’s clock. Measured on daily closes from Yahoo Finance, from a record close to the lowest close before the next record, there were 11 such falls from 1956 to 2022.
The table sets the time each one actually took from its bottom back to a new record against the years this calculator returns for the same fall at its 8% default.
| Bottom | Fall | Gain needed | Bottom to new record | Calculator at 8% |
|---|---|---|---|---|
| Oct 1957 | 21.5% | 27.3% | 0.92 years | 3.14 years |
| Jun 1962 | 28.0% | 38.8% | 1.19 years | 4.26 years |
| Oct 1966 | 22.2% | 28.5% | 0.57 years | 3.26 years |
| May 1970 | 36.1% | 56.4% | 1.78 years | 5.81 years |
| Oct 1974 | 48.2% | 93.1% | 5.79 years | 8.55 years |
| Aug 1982 | 27.1% | 37.2% | 0.23 years | 4.11 years |
| Dec 1987 | 33.5% | 50.4% | 1.64 years | 5.30 years |
| Oct 2002 | 49.1% | 96.6% | 4.64 years | 8.79 years |
| Mar 2009 | 56.8% | 131.3% | 4.05 years | 10.90 years |
| Mar 2020 | 33.9% | 51.3% | 0.41 years | 5.38 years |
| Oct 2022 | 25.4% | 34.1% | 1.27 years | 3.81 years |
The median recovery took 1.3 years from the bottom; the steady-rate estimate for the same falls is 5.3.
Even the slowest, after October 1974, climbed at about 12.0% a year, and the median pace implied by the table is 28.6% a year. These are price-only closes with no dividends, so the true recoveries were quicker still. Every row, with the peak dates and closes, is in the bear-market recovery table.
The bottom is only known afterwards, so the fairer start is the first close 20% below the old record. Counted from there, the median wait for a new record was 1.6 years, and 8 of the 11 beat the 2.90 years the calculator gives for the 25% gain needed at that point.
Read that as a fact about an index, not a promise about a holding. Recoveries from a bottom have tended to start steeply, which is why a constant rate overstated the wait in every row.
The whole episode, peak to new record, still took a median of 2.0 years, and a single stock has no mechanism that makes its row look like these.
Of the 24,971 unique videos collected for this site’s search study, 7 have drawdown in the title, at a median of 3,813 views across 7 channels. Breakeven appears in 2 instruction-shaped titles at 40,893.
Seven videos at a 3,813 median against two at 40,893 for the same arithmetic under a different name. The subject people search for is the per-trade version, which the breakeven after a loss calculator covers; the portfolio version has almost no audience despite mattering more.
The question: Down thirty percent. Average down to recover faster?
The answer is that averaging down lowers the price you need, and raises the amount at stake. Adding capital at 70 reduces the average cost and increases the position — so the recovery percentage falls while the money that can be lost rises. It is a different trade with a different size, not a repair to the existing one, and treating it as a repair is how a manageable drawdown becomes an unrecoverable one.
When it fails
The formula assumes a recovery happens, and for an individual holding it may not. An index has a mechanism for recovering — its worst members are replaced — and a single company has none. The calculator returns 42.86% just as confidently for something that will never trade there again, and the honest reading of a large drawdown in one position is that the arithmetic of recovery is the least of the questions.
The second failure is treating the years figure as a schedule. Recoveries arrive in bursts.
A third is ignoring contributions. New money changes both the base and the timeline.
A fourth is measuring from an intraday high. Peak-to-trough on closes is the honest version.
A fifth is forgetting inflation. Getting back to the same number is not getting back to the same purchasing power.
And a sixth is averaging down to shorten the recovery. That is a new position, not a repair.
Related
Drawdown covers how the fall is measured and why the peak matters. Volatility is what produces the falls in the first place. And index funds is the structure with a built-in recovery mechanism.
Risk of ruin is the same curve seen from the other end: the chance that a run of losses pushes an account so far down the steep part that no ordinary return brings it back.
The number that changes behavior is not 42.86. It is the shape of the curve past about 40%, where each further point of decline starts adding several points to what recovery requires. Everything a risk framework does — position limits, stops, diversification — is really about staying on the flat part of that curve.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money. Some links on this page earn a commission if you buy through them. It costs you nothing and it does not decide what appears here or in what order — how these pages are made is set out in our methodology.