What Is Drawdown in Trading? How It's Measured and Why Getting Back Costs More
Drawdown is the fall in a trading account or strategy from its highest value to a later low, stated as a percentage of that high, and it lasts until the account makes a new high. This covers the formula and the running-peak method with a worked example, maximum and average drawdown, time under water, the CFTC's definition, why recovery takes a bigger gain than the loss, how losing streaks at different risk-per-trade levels stack into a drawdown, why there's no universal good max drawdown, and what to do when you're in one.
What is drawdown in trading? It's the drop in your account, or in a strategy's equity, from its highest point to a later low, stated as a percentage of that high. It starts the moment you come off a peak, and it only ends when the account makes a new one.
I trade momentum, and a choppy week can hand you five or six stopped-out breakouts in a row. None of them is a big loss on its own. Together they're a drawdown, and the drawdown is the number that decides if you're still trading next month. This post covers the definition, how to calculate it, maximum drawdown and the recovery math. Then the part most explainers skip: how the size of each loss turns into the size of the hole.
Quick Answer
Drawdown is how far your account has fallen from its highest value, measured as a percentage of that high. A $12,000 peak that drops to $9,600 is a 20% drawdown. It lasts until the account makes a new high. Maximum drawdown is the deepest one on record, and a 20% drop needs a 25% gain to get back.
Every account and dollar figure on this page is hypothetical. The regulatory definition comes from the CFTC, the recovery numbers are plain arithmetic, and the one real-world statistic is from a Morgan Stanley study, linked where it's used.
What Is Drawdown in Trading?
The drawdown meaning in trading, plainly: how far the account sits below its own high right now, or how far it fell at its worst. Traders quote it as a percentage of that high. If the account peaked at $12,000 and sits at $10,800, you're in a 10% drawdown.
It's measured from the running peak
The peak is the highest value the account has reached so far. A running high. Your deposit has nothing to do with it, which is why a trader who's up on the year can still be deep in a drawdown. In the worked example below, the account starts at $10,000, runs to $12,000, and on day 5 closes at $10,200. That's up 2% from the start and 15% under the peak. Both are accurate. The drawdown figure is the one that shows how much of the run-up you handed back.
Every new high resets the count. The old peak stops mattering, and any loss from the new high starts a fresh drawdown measured from there.
Dollars or percent
You can state it in dollars too. A drop from $12,000 to $9,600 is a $2,400 drawdown. Percent is what makes accounts comparable, since $2,400 is noise on a big account and most of a small one.
Current drawdown vs maximum drawdown
Current drawdown is where you are right now against the last peak. Maximum drawdown is the deepest one in the whole record. The current number moves every day. The max only moves when a drawdown goes deeper than the previous worst. It gets its own section below.
The regulator's definition
US futures regulation actually defines this. CFTC Regulation 4.10 says draw-down "means losses experienced by a pool or account over a specified period," and defines worst peak-to-valley draw-down as "the greatest cumulative percentage decline in month-end net asset value due to losses sustained by a pool, account or trading program during any period in which the initial month-end net asset value is not equaled or exceeded by a subsequent month-end net asset value."
That's the running-peak rule in legal English. The drawdown lasts until the account gets back to where it was. One catch: it uses month-end values, so a nasty dip in the middle of a month that recovers by the 30th never shows up. Your own daily numbers will often show a deeper drawdown than a month-end series of the same account.
Why it matters
Returns tell you what a strategy made. Drawdown tells you what you had to sit through to get it. Win rate tells you neither, because it says nothing about how losses cluster or how big they were. Every strategy has drawdowns, good ones included. What separates them is how deep they get, how long they last, and what the trader does while in one.
(Outside trading, drawdown also means pulling money from a credit line or a retirement pot. Same word, different thing.)
How Do You Calculate Drawdown?
To calculate drawdown in trading, take the peak, subtract the low that came after it, and divide by the peak.
Drawdown % = (peak - trough) / peak × 100. A $12,000 peak that falls to $9,600: ($12,000 - $9,600) / $12,000 = $2,400 / $12,000 = 20%.
Some spreadsheets and platforms show it as a negative number, minus 20%. Same drawdown. The two inputs have rules, though. The peak is the running high, and the trough is the lowest point after that peak. Grab the highest and lowest values from a whole year without checking which came first and you can get a number that never happened.
The running-peak method
For a real record you need a column that tracks the running peak. Each day, the running peak is the larger of yesterday's running peak and today's balance. Drawdown that day is (running peak - balance) / running peak. In a spreadsheet with balances in column B, that's =MAX(C1,B2) for the peak in C2, with C1 set to =B1, and =(C2-B2)/C2 for the drawdown in D2, filled down. The table runs twelve days of a made-up account through it.
| Day | Closing balance | Running peak | Drawdown from peak |
|---|---|---|---|
| Day 1 | $10,000 | $10,000 | 0% (at a high) |
| Day 2 | $11,000 | $11,000 | 0% (at a high) |
| Day 3 | $12,000 | $12,000 | 0% (at a high) |
| Day 4 | $11,100 | $12,000 | 7.5% |
| Day 5 | $10,200 | $12,000 | 15% |
| Day 6 | $9,600 | $12,000 | 20% |
| Day 7 | $10,300 | $12,000 | 14.2% |
| Day 8 | $11,000 | $12,000 | 8.3% |
| Day 9 | $11,600 | $12,000 | 3.3% |
| Day 10 | $12,300 | $12,300 | 0% (at a high) |
| Day 11 | $11,800 | $12,300 | 4.1% |
| Day 12 | $12,700 | $12,700 | 0% (at a high) |
Same twelve days drive the diagram below and the examples after it. The running peak column only goes up or stays flat. It sits at $12,000 until day 10, when the account finally closes above it.
Maximum drawdown
Max drawdown meaning: the largest number in that drawdown column. Here it's 20%, from the day 3 peak to the day 6 low. It's the worst stretch the record contains, the worst case that has already happened. Maximum drawdown in trading gets quoted more than any other risk number, for strategies, funds and backtests alike, because it's one number and it's the one that hurts to look at.
Average drawdown
Average drawdown is the typical depth across all the separate drawdown episodes. This record has two episodes: 20% and 4.1%. Average, about 12%. Two is too few for that to mean much. On a real record you'd want dozens before trusting it.
Duration: time under water
Depth is one dimension. Time is the other. Duration runs from the peak until the account first gets back to it. Here the peak was day 3, the low was day 6, and the account closed above $12,000 again on day 10. That's 7 trading days under water: 3 going down, 4 climbing back. Two drawdowns of the same depth feel nothing alike if one lasts a week and the other lasts six months.
One drawdown, start to finish (hypothetical account)
What max drawdown can't tell you
Max drawdown is backward-looking. It reports one path of trades, the one that happened, and says nothing about how often drawdowns show up. A longer record can only match it or show a deeper one, never a shallower one, so the number tends to grow the longer you trade.
Order matters too. Take six trades on a $10,000 account: three winners of $600 and three losers of $400. Alternate them, win then loss, and the max drawdown is 3.8%. Take the three losses first and it's 12%. Same trades, same $600 profit at the end, more than three times the drawdown. That's why a backtest's max drawdown reads better as a floor than a ceiling. There's more on what a backtest can and can't tell you in the comparison with pre-trade grading.
Platforms also add their own labels, like absolute and relative drawdown, and those don't always mean the same thing from one platform to the next. Check your platform's definition before comparing its number with anyone else's.
Drawdown vs a Losing Trade: What's the Difference?
A loss is one trade. A drawdown is the account. The losing trade is a single closed result, say minus $300, while the drawdown is the cumulative fall from the account's peak, which can contain several losers and a few winners along the way. In the twelve-day example, days 7, 8 and 9 were all up days, and the account was still in a drawdown the whole time.
One thing a few explainers get wrong: there's no such thing as a loss that sits outside a drawdown. If you're at a new high and take a loss, that new high is the peak and the loss starts a drawdown from it. Day 11 in the example is exactly that, a fresh 4.1% drawdown one day after a new high. Most drawdowns are small and over quickly. They still count.
The distinction matters because you control the two differently. The size of one loss you control directly, with your stop and your position size. The drawdown is what a sequence of those losses adds up to, and the sequence is partly luck. You can't pick the order your trades come in. You can pick how much each loss costs, and how many marginal trades you let into the sequence.
Why Is Climbing Out of a Drawdown Harder Than Falling In?
Because you're climbing back on a smaller base. Lose 20% of $10,000 and you have $8,000. Getting back to $10,000 means making $2,000 on $8,000, which is 25%. The general version: if d is the drawdown as a decimal, the gain you need is g = d / (1 - d). It falls straight out of (1 - d) × (1 + g) = 1. What's left, times the growth, has to equal where you were.
| Drawdown | Left from a $10,000 peak | Gain needed to get back to the peak |
|---|---|---|
| 5% | $9,500 | 5.3% |
| 10% | $9,000 | 11.1% |
| 20% | $8,000 | 25% |
| 25% | $7,500 | 33.3% |
| 30% | $7,000 | 42.9% |
| 40% | $6,000 | 66.7% |
| 50% | $5,000 | 100% |
| 60% | $4,000 | 150% |
| 75% | $2,500 | 300% |
| 90% | $1,000 | 900% |
| 100% | $0 | Can't be recovered (nothing left to grow) |
Up to about 10% the gap barely registers. Past 30% it gets ugly fast, and at 50% you need to double what's left. CME Group's course on controlling risk puts it as "the amount you have to gain will be larger than the amount you lost," with a similar table running down to a 100% loss, labelled "Broke!"
The formula gives you the size of the gain. It says nothing about how long that takes, and nothing in the table is a time estimate. A 25% gain might take one trader a month and another a year, depending on their edge and how often they trade.
Recovery isn't guaranteed
You'll read that a drawdown is "temporary" or "not a permanent loss." Closed losses are real money, and a drawdown ends only when the account makes a new high. Nothing guarantees that. The best base rate I found is for stocks, not trading accounts. In Morgan Stanley Investment Management's 2025 study of drawdowns and recoveries, covering more than 6,500 US stocks from 1985 to 2024, "about 54 percent of stocks never return to par after hitting bottom."
That's buy-and-hold price data. The study also left out companies that were delisted, which it notes typically means a 100% drawdown. It isn't a trading statistic and it says nothing about your account. It's a reminder that "it'll come back" is a hope. So the cheapest drawdown to climb out of is the one that never got deep, which brings it back to the size of each loss.
How Do Per-Trade Losses Stack Into a Drawdown?
A drawdown is per-trade losses stacking up from a peak. Obvious, maybe, but it means the inputs are things you set before each trade: how much each loss costs (stop distance times position size), and how many losing trades get into the streak, including the ones you didn't need to take.
The table below shows what a losing streak does at different risk levels. Each loss is exactly the planned risk, the stop fills where it was set, and there are no winners in between. Nothing here says how likely a streak is. It just shows what one costs.
| Risk per trade | 3 losses in a row | 5 losses in a row | 10 losses in a row | Gain to recover from 10 losses |
|---|---|---|---|---|
| 0.5% | 1.5% | 2.5% | 4.9% | 5.1% |
| 1% | 3.0% | 4.9% | 9.6% | 10.6% |
| 2% | 5.9% | 9.6% | 18.3% | 22.4% |
| 3% | 8.7% | 14.1% | 26.3% | 35.6% |
| 5% | 14.3% | 22.6% | 40.1% | 67.0% |
These numbers are compounded. Each trade risks the stated percent of the balance at that moment, so the dollar risk shrinks as the account shrinks. On a $20,000 account at 1%, the first loss is $200 and the tenth is about $183, leaving $18,088, a 9.6% drawdown. Risk a fixed dollar amount instead and it's plain multiplication: ten losses at 1% of the starting balance is a 10% drawdown. CME's course makes the same point about fixed-percentage sizing, that it "reduces the size of each succeeding trade, which slows your account's decay when you experience a losing streak."
Read across the rows. At 1% risk, ten straight losses is a 9.6% drawdown that needs a 10.6% gain to fix. At 5%, the same streak is a 40.1% hole that needs 67%. The streak is identical. The difference is entirely the size you picked. A common rule of thumb is to risk 1% to 2% of the account per trade. It's a convention, nothing official, and the guide on how much to risk per trade walks through where it comes from and how to size to it.
A stop is what makes the loss a number
All of that assumes each loss is the size you planned. Without a defined stop it isn't a number at all, and the table stops applying. The stop sets where you're wrong, and position size turns that distance into a dollar amount you chose in advance. The stop loss guide covers how a stop sets your position size, and why a stop isn't a promise of the exact fill when price gaps through it. On fast momentum names this matters even more, since a wide stop means fewer shares. The risk section of the momentum trading playbook goes through sizing off the stop on those.
Leverage and margin scale every row. Double the position on the same stop and you've doubled the risk per trade, so a 1% trader is suddenly reading the 2% row, where ten losses is 18.3% instead of 9.6%.
Investors cut drawdowns by spreading money across things that don't fall together. The day-trading version is simpler. Five long breakouts in five chip stocks on the same morning are closer to one trade than five. If the sector rolls over, all five stops can go in the same few minutes, and your 1% risk was really 5%.
What you do inside a drawdown decides how deep it gets
The math above assumes you keep doing the same thing. People don't. A few losses in, the urge is to size up and win it all back in one trade, or to start taking setups you'd normally pass on because sitting out feels like accepting the loss. That's revenge trading, and it's how an ordinary losing streak becomes a drawdown you remember. The breakdown of revenge trading and overtrading covers why willpower alone rarely stops it.
Deep, long drawdowns also get into your head. They're where traders abandon a plan that was fine, or jump to a new strategy every two weeks and never find out whether any of them worked. The guide to trading psychology goes into the biases behind that.
Where a chart grade fits, and what it can't see
SnapPChart can't see your account, so it can't tell you your drawdown. It doesn't connect to a broker or a prop firm, doesn't know your balance, doesn't read live prices and doesn't enforce any limit. It reads one chart screenshot that you upload and grades the setup from A+ down to F (on futures, forex, indices, metals and crypto charts the grade tops out at B).
What it can do sits at the two inputs above. When a setup is worth planning, it proposes an entry, a stop with its reasoning, and targets, plus a size helper where you type your account size and a risk percent (it starts at 1%) and it works out the share count from that plan's entry and stop. It doesn't save what you type. If the first target doesn't pay at least 2:1 by default, about twice the risk, the verdict is skip. So the loss on the next trade becomes a number you picked before entry, and a C-grade setup you'd only take to win money back gets called what it is. I'm not going to claim it shortens a drawdown. It doesn't know you're in one.
The size helper only shows when there's a live entry and stop, so not on skips, not on wait-for-a-pullback plans, and not on the second free analysis, where entry and stop are hidden. For more on what AI reads off a chart and where it's blind, there's an overview of AI chart analysis and a section on what AI can't see on a chart.
Know what the next loss costs before you take it.
Upload a chart screenshot. When the setup is worth planning, a full analysis gives you an entry, a stop and targets, and sizes the position to that stop from the account size and risk percent you type in. An F grade, or a first target under about twice the risk, gets a skip. It can't see your account or your drawdown.
Grade this chartWhat Is a Good Max Drawdown?
There's no universal good number. You'll see 5%, 10 to 20% and 20 to 30% quoted as normal or acceptable, usually without a source. Any of them could be right for one trader and wrong for another, because a max drawdown only means something next to a few other things.
- The return it bought. A 15% max drawdown on a strategy that makes 60% a year is a different deal from 15% on one that makes 8%. The Calmar ratio puts a number on this: annualized return divided by max drawdown.
- How you size. From the streak table, a trader risking 2% a trade is ten straight losses away from an 18.3% drawdown. If that's unacceptable, the thing to change is the 2%.
- How long the record is. A longer record can only match or deepen the max, so a three-month max drawdown is a weak guide to the three-year one.
- Could you actually sit through it? The real limit is the drawdown where you stop following the plan. Be honest about that number, and write it down before you get there.
Two hypothetical strategies both make 30% in a year. One gets there with a worst drawdown of 8%. The other drops 35% before it comes back. Same return, a Calmar of 3.75 against about 0.86, and two very different years to live through. A high win rate can hide the same thing, if the few losses are big. That's the case for asking about drawdown before you commit money to any strategy, signal service or fund: what was the max drawdown, how long did it take to recover, and could you have held through it?
US regulation treats it as a headline number too. CFTC Regulation 4.25 requires commodity pools to show "the worst peak-to-valley draw-down during the most recent five calendar years and year-to-date" in their performance disclosures.
Funded and evaluation accounts are the exception where a good max drawdown is defined for you. It's whatever keeps you inside the firm's limit, and breaching it usually ends the account. Those limits come in trailing and static versions that behave very differently, and the prop firm drawdown calculator shows how much room you have left under one.
How to Recover From a Drawdown in Trading
Start by working out what kind of drawdown it is. Compare it with your own record. If it's within the depth and length of drawdowns you've had before while following the same rules, it may just be a normal losing streak. If it's deeper or longer than anything in the record, that's the signal to review, keeping in mind that the record's max was only ever a floor. Judging whether a strategy has an edge at all takes a real sample size, more trades than most people assume.
- Cut size first.Halving your risk per trade halves the damage from the next streak. It also slows the climb back, and that's fine. Sizing up to get back faster is how a 15% drawdown turns into 30%.
- Check whether the losses followed the plan. Go through every trade in the drawdown and tag it: planned setup, planned size, stop honored, or not. Losses inside the plan are the cost of running the strategy. Losses outside it are a different problem with a different fix. A journal template with those fields makes this a ten-minute job, and tracking how often you take a setup versus skip it shows whether marginal trades crept in.
- Trade less, and only what's in the plan.The setups that cost the most in a drawdown are the ones you'd skip on a good day. If a setup grades C, skipping it counts as a decision too, and usually the better one.
- Cap the day. A daily loss limit stops one bad session from doing a week's damage. The risk rules post covers when to stop trading for the day.
- Take a break if you can't stop clicking. A day off costs nothing. Another revenge trade costs real money.
Pick your cut-off points before you need them
The time to choose drawdown thresholds is when you're at a new high and calm. Three red days in is too late. Pick a level where you cut size and a level where you stop live trading and review. For illustration only, a trader might halve size at 8% below the peak and stop for a week at 15%. Those aren't standards. Several sites publish their own numbers and none of them agree, because the right levels depend on your sizing and your strategy's history. What matters is that the rule exists before the drawdown does.
Frequently Asked Questions
Does an open losing trade count toward drawdown?
It depends which drawdown you're looking at. Balance drawdown only counts closed trades. Equity drawdown includes open positions marked at the current price, so a trade that's down but still open already counts. Equity drawdown is often the deeper of the two, because it catches losses before they're closed and swings that never show up in the closed-trade record, and it's often the one brokers and prop firms watch. If you hold losers hoping they come back, your balance drawdown can look fine while your equity drawdown is already ugly.
Is drawdown the same as a daily loss limit?
No, though they're related. A daily loss limit caps how much you can lose in one session, usually measured from that day's starting balance, and it resets the next morning. Drawdown is measured from the account's highest point and doesn't reset until the account makes a new high. You can respect your daily limit every single day and still end up in a deep drawdown, one capped bad day at a time. Prop firms usually enforce both, and plenty of retail traders set both for themselves.
How long does a drawdown usually last?
There's no typical length. It depends on how often the strategy trades, how big its edge is and how deep the hole got. The recovery formula tells you the size of the gain you need, never how long it takes. The most honest estimate you have is your own record: find the longest time under water so far, then assume the next one can be longer, since a longer record can only show longer stretches. If you don't have a record yet, you don't know, and that's worth admitting before you size up.
Should I stop trading during a drawdown?
Not automatically. If the drawdown is within the depth and length you've seen before and you've been following your rules, stopping can mean missing the trades that would have pulled you out. If it's deeper than anything in your record, or you've caught yourself breaking rules, stepping away from live trading to review is reasonable. There's a middle option too: keep the routine but cut size hard, or move to a simulator for a week, so you keep reading charts without adding to the hole.
Can SnapPChart tell me my drawdown or warn me when I'm in one?
No. It doesn't connect to a broker or a prop firm, doesn't know your balance and doesn't track your trades, so it has no way of measuring a drawdown or alerting you to one. It grades one chart screenshot at a time. What it can do is help with the next trade: when a setup is worth planning, a full analysis gives an entry, a stop and targets, and works out a position size from the account size and risk percent you type in, so the loss if the stop hits is a number you chose. An F grade, or a first target under about twice the risk, gets a skip. Your actual drawdown lives in your broker statement or your journal.
Everything here is general education. It isn't investment, financial or trading advice, and nothing on this page recommends any security, strategy, broker or prop firm. The draw-down definitions are from CFTC Regulations 4.10 and 4.25 (17 CFR 4.10 and 4.25). The recovery and fixed-percentage quotes are from CME Group's Controlling Risk course. The stock recovery figure is from Mauboussin and Callahan, "Drawdowns and Recoveries: Base Rates for Bottoms and Bounces," Morgan Stanley Investment Management, May 2025, which covers buy-and-hold US stock prices, not trading accounts. All were read in October 2026. Every account, balance, streak and strategy on this page is hypothetical, and the losing-streak table says nothing about how likely a streak is. Day trading carries a real risk of losing the money you put in. SnapPChart grades one static chart screenshot. It does not connect to brokers or prop firms, know your balance, measure your drawdown, read live prices or place trades.
Writes about AI-assisted day trading, technical analysis, and the systems traders actually use to stay disciplined.
In a drawdown? Make the next loss a number you picked.
Upload a chart screenshot. SnapPChart grades the setup A+ to F and, when it's worth planning, on a full analysis it gives you an entry, a stop with its reasoning, targets, and a position size worked out from that stop and the risk percent you choose. If the first target doesn't pay about twice the risk, it says skip. It doesn't connect to your broker or know your balance. One skipped C-grade trade can pay for the subscription.