Drawdown is the decline from an earlier high-water mark to a later value in an account or modeled equity curve. Recovery arithmetic answers one narrow question: how large a gain on the smaller remaining balance would bring it back to that earlier level? It does not say that the gain will happen, how long it might take, or whether the trading process should continue.

drawdown and recovery arithmeticTwo fictional balance paths start at $10,000. The first falls 20 percent to $8,000 and needs a 25 percent gain to return; the second falls 50 percent to $5,000 and needs a 100 percent gain. Values illustrate arithmetic, not trading results or forecasts.SAME $10,000 PEAK · DIFFERENT RECOVERY BASES20% drawdown−20%+25% on $8,000PEAK$10,000TROUGH$8,000RECOVERED$10,000$2,000 loss · $2,000 gain on the smaller balance50% drawdown−50%+100% on $5,000PEAK$10,000TROUGH$5,000RECOVERED$10,000$5,000 loss · another $5,000 to regain the peakIllustrative balance arithmetic · the rebound is not a prediction
Two fictional balance paths start at $10,000. The first falls 20 percent to $8,000 and needs a 25 percent gain to return; the second falls 50 percent to $5,000 and needs a 100 percent gain. Values illustrate arithmetic, not trading results or forecasts. Illustrative teaching graphic · not live market data

Calculate the recovery gain

Write the loss as a fraction d of the peak. The remaining fraction is 1 − d. To return to the peak, the required gain on what remains is:

recovery gain = d ÷ (1 − d)

The denominator matters: after a loss, the base used to calculate the recovery is smaller. The table rounds to one decimal place; the formula gives the exact result.

Decline from peak Remaining equity Gain needed to regain peak
10% 90% 11.1%
20% 80% 25.0%
30% 70% 42.9%
40% 60% 66.7%
50% 50% 100.0%
75% 25% 300.0%
90% 10% 900.0%

Consider a fictional $10,000 equity peak. A 20% decline leaves $8,000, so the shortfall is $2,000. That shortfall is 25% of $8,000. A 50% decline leaves $5,000 and requires another $5,000, or a 100% gain on the remainder. These are balance-sheet calculations, not trade returns or forecasts. At a complete 100% loss, no capital remains in the model, so there is no finite percentage gain that restores the starting balance without adding money.

Losses compound against the changing balance. If equity falls 20% and then another 20%, the path is $10,000 × 0.8 × 0.8 = $6,400, a 36% decline from the original peak—not a 40% decline. Returning from $6,400 to $10,000 would require a 56.25% gain. Keep the peak, each successive balance and the measurement convention visible when reviewing a sequence.

Measure the path consistently

For each observation, compare equity with the highest earlier value in the chosen sample. The largest peak-to-later-trough decline is maximum drawdown. In a backtest or journal, record the peak date and value, trough date and value, dollar loss, percentage loss, whether and when the curve later made a new high, and the equity sampling frequency. State whether equity includes open-position marks and which fees, funding, slippage and liquidation assumptions were applied. A start-to-finish return cannot reveal the deepest interim decline or time spent below a prior peak.

Here is a separate fictional month-end series. The drawdown column compares each balance with its running peak; values are rounded to two decimal places.

Observation Equity Running peak Drawdown from peak
M0 $10,000 $10,000 0.00%
M1 $10,500 $10,500 0.00%
M2 $10,200 $10,500 −2.86%
M3 $9,200 $10,500 −12.38%
M4 $9,800 $10,500 −6.67%
M5 $10,300 $10,500 −1.90%
M6 $10,500 $10,500 0.00%
M7 $10,600 $10,600 0.00%

At M3, the $1,300 decline divided by the $10,500 prior peak gives 12.38%. The series first returns to the old peak at M6, six intervals after M0: five intervals from peak to peak, with two from peak to trough and three from trough back to the old peak. It makes a new high only at M7. These invented values explain measurement; they are not market or account history.

fictional equity path with peak, trough and recoveryEight invented month-end balances show a 12.38 percent peak-to-trough drawdown from $10,500 to $9,200. The old peak is first regained at interval six; a new high follows at $10,600. The line is a teaching example, not account or market data.FICTIONAL MONTH-END EQUITY · PEAK, TROUGH AND RETURN$11,000$10,500$10,000$9,500$9,000prior peak · $10,500trough · $9,200old peak regainedM0M1M2M3M4M5M6M7($9,200 ÷ $10,500) − 1 = −12.38% · invented values only
Eight invented month-end balances show a 12.38 percent peak-to-trough drawdown from $10,500 to $9,200. The old peak is first regained at interval six; a new high follows at $10,600. The line is a teaching example, not account or market data. Illustrative teaching graphic · not live market data

Sampling changes what the measure can see. A monthly closing series may miss a deeper intramonth fall; a daily close series may miss an intraday liquidation or a breach that recovered before the close. Name the observation convention and do not describe close-only drawdown as the worst price path. If cash was added or removed, disclose whether the series is a raw account balance or a cash-flow-adjusted performance measure. Deposits should not appear as earned return.

Depth and duration are separate dimensions. A 2026 International Journal of Forecasting paper by Tommaso Proietti develops a framework that tracks drawdown size alongside elapsed time from running highs and lows. Its empirical series is the S&P 500 daily close from January 2000 through 30 August 2023, plus component stocks; that is a measurement reference, not evidence about crypto drawdowns or a forecast for a trading strategy. In the fictional series above, five intervals from peak back to the old peak and three from trough to the old peak answer different duration questions. Define which clock your report uses. The CFA Institute summary likewise notes that drawdown is historical and has no direct predictive value. A larger recovery percentage is arithmetic on the residual balance, not a probability or a timetable.

External deposits can make an account balance rise without the trading process recovering. Withdrawals can make it fall even when the underlying positions gained. Keep cash flows visible and choose a documented return or equity method before comparing periods; do not count a deposit as trading profit. If you compare a portfolio with a benchmark, also define a benchmark-relative method. The CFA Institute’s drawdown research summary explains that absolute portfolio drawdown and benchmark-relative drawdown can answer different questions; subtracting two maximum-drawdown percentages does not automatically create a valid active-drawdown measure.

Use drawdown as a review prompt

  1. Freeze the period, equity series and valuation frequency before looking for the worst decline.
  2. Identify the previous high-water mark and the lowest later value; calculate both the dollar and percentage decline.
  3. Measure time to a new high separately. If no new high occurs before the sample ends, label recovery as unobserved rather than complete.
  4. Check the ledger assumptions and any deposits or withdrawals before attributing the move to strategy performance.
  5. Compare the result with preselected limits and benchmarks; explain what the metric omits.
workflow for measuring and reviewing an equity drawdownA consistent review fixes the sample and valuation convention, identifies the running peak and later trough, reports percentage depth and elapsed time separately, then checks cash flows, costs, open marks and prewritten limits.MEASURE THE PATH BEFORE EXPLAINING IT01 · Freezedates · frequencyequity convention02 · Find marksrunning peaklater trough03 · Reportdepth · dollarsduration · dates04 · Auditcash flows · costsmarks · gaps05 · Compareprewritten limitsthen review processA close-based curve cannot reveal every intraperiod low.The report describes the chosen series and model; it does not predict a recovery.
A consistent review fixes the sample and valuation convention, identifies the running peak and later trough, reports percentage depth and elapsed time separately, then checks cash flows, costs, open marks and prewritten limits. Illustrative teaching graphic · not live market data

Set any review threshold before seeing the result. If a measured limit is exceeded, the next step is to check data, cash flows, instrument rules, costs and the written risk plan. It is not automatically a signal to hold, add capital, widen a stop or increase leverage. The appropriate response depends on the account, contract, jurisdiction and the trader’s precommitted process.

The CFTC cautions that losses in margined virtual-currency futures can be amplified and that a participant can be required to add funds or close positions. That is a specific futures and margin risk; it does not define every spot account or venue rule. A larger recovery percentage is not evidence that a strategy has a positive edge, and increasing exposure to recover faster can deepen the next loss.

The drawdown recovery calculator works out the arithmetic for an entered decline. SuperQuant’s backtest metric is based on its own modeled equity and assumptions, including synthetic examples; neither tool forecasts recovery or demonstrates live-account experience. For related checks, see the backtest metrics guide and the position-size calculator. This is educational material, not individualized investment advice.