Skip to content

Maximum Drawdown

Paste your results and find the worst stretch — how deep it went, how long it lasted, and how long it took to come back.

One result per line, oldest first. A bare column of numbers works; so does a CSV with a profit_loss column and an optional date.

$

Optional, but percentage drawdown has no meaningful denominator without it.

Maximum drawdown

−$300.00

The worst peak-to-trough fall, from 2026-01-09 to 2026-01-18 — 5.7% of the bankroll at its peak. Measured against the running peak, not the start, which is why a record that is up overall can still show a large drawdown.

Net P/L
+$320.00
Results
12
Drawdown as a percentage
-5.70%

Against the equity peak it fell from.

Length of the fall
3 results

2026-01-09 → 2026-01-18

Longest time underwater
6 results

The longest stretch spent below a previous high before regaining it.

Currently down from peak
−$50.00
Equity curve · the shaded stretch is the worst peak-to-trough fall

Everything is parsed and calculated in your browser — nothing is uploaded, and no file leaves your machine. Rows without a valid number are skipped and reported above rather than dropped silently, because a series quietly missing three results gives a drawdown figure that is simply wrong. Where no dates are supplied, row order is used.

Runs entirely in your browser · nothing is sent or saved

What is maximum drawdown?

Maximum drawdown is the largest fall from a peak to a later trough in a betting record — the worst stretch, measured from the most you had rather than from where you started.

It is the number that describes what a record felt like to live through, and it is invisible in the figures people usually publish. Two bettors can finish a season at identical profit having had completely different years.

A bankroll that doubles and then falls back a quarter has had a 25% drawdown even though it is still well ahead. That quarter is what you actually lost from the high-water mark, and it is what decides whether a staking plan was survivable.

How it works

Maximum drawdown is the largest fall from a peak to a subsequent trough in your equity curve. It is the single best measure of what a betting record felt like to live through, and it is invisible in the figures people usually report. Two bettors can finish a season at exactly the same profit having had completely different years.

The measurement is against the running peak, not against where you started. A bankroll that doubles and then falls back a quarter has had a 25% drawdown even though it is still well ahead — and that is the right way to count it, because the quarter is what you actually lost from the high-water mark you had already reached.

Two other numbers matter as much as the depth. Duration is how many results the fall took. And time underwater is the longest stretch spent below a previous high before regaining it — frequently much longer than the fall itself, and psychologically the harder part. A three-week decline that takes four months to recover is four months of betting well and seeing no progress, which is when most people abandon a working process.

A percentage is only reported when you supply a starting bankroll. A percentage of a bare cumulative P/L series has no sensible denominator — the first losing bet would be an infinite fall from a peak of zero — so rather than print a number that looks meaningful and is not, the calculator asks for the bankroll or withholds the figure.

Everything happens in your browser. The file is read locally, parsed locally, and never sent anywhere.

The formula

equity_t       = starting bankroll + cumulative P/L up to t
running peak_t = max(equity_0 … equity_t)

drawdown_t     = equity_t − running peak_t     (≤ 0)
drawdown %_t   = drawdown_t / running peak_t

maximum drawdown = the most negative drawdown_t

duration        = results from that peak to that trough
time underwater = longest run below a previous high
                  before it is regained

Every point is compared with the highest equity seen before it, which is why a record that is up overall can still show a large drawdown.

A worked example

The sample series loaded above — twelve results on a $5,000 bankroll — finishes at $5,320, up $320. Read the bottom line alone and it was a good stretch.

It was not, in the middle. The equity peaks at $5,260 on the third bet, then loses three in a row down to $4,960 — below where it started. That is a maximum drawdown of −$300, or 5.70% of the peak it fell from. Measured against the $5,000 start it would read as −0.8%, which is why the peak is the right denominator.

The number that hurts most is neither of those. It is six results spent below a high-water mark before regaining it — half the series, betting perfectly well, watching the account go nowhere. That is the stretch people quit during, and it does not appear in the final figure at all.

Common questions

What is a normal maximum drawdown?
Larger than most people expect. A profitable flat bettor at 1–2% units routinely sees 15–25% drawdowns over a season, and deeper ones are not evidence that anything is broken. The bankroll simulator shows the distribution for your own edge and staking.
Why is it measured from the peak instead of the start?
Because the peak is the money you actually had. Falling from $6,200 to $5,400 is an $800 loss regardless of where the account started, and measuring from the start would report zero drawdown for a bankroll that has never yet dipped below its opening balance.
Why do I need to enter a bankroll?
For the percentage only. A percentage drawdown on a bare cumulative P/L series has no meaningful denominator — early in the series the running peak is near zero and the percentage explodes. Without a bankroll, the dollar figures are still exact.
What CSV format does it accept?
A bare column of numbers, or a CSV with a profit_loss column and an optional date. Columns are found by name rather than position, negatives can be signed or in parentheses, and currency symbols and commas are stripped.
Is my data uploaded anywhere?
No. The file is read and parsed in your browser and nothing is transmitted. There is no server involved in this page at all.
Why is time underwater longer than the drawdown itself?
Because recovering takes longer than falling. Losing 20% requires a 25% gain to get back, and at a normal edge that takes considerably more bets than the decline did. It is usually the harder part of a drawdown to sit through.

The guide behind this calculator

Related calculators

For informational and analytical purposes only. These tools do not predict outcomes and do not recommend wagers.