Maximum Drawdown Explained: Formula, Types and Real Limits

By Lenka Rož Schánová · Operations & Risk, For Traders

Max DD is the standard abbreviation for maximum drawdown (also written MDD) — the largest peak-to-trough decline an account suffers before making a new equity high, calculated as MDD = (Trough − Peak) ÷ Peak. The denominator is the peak, not your starting balance, which is why a $100,000 account that runs to $112,400 and falls to $98,900 has a max DD of −12.01%, not −1.1%.

Key takeaways

• Max DD, MDD and maximum drawdown are the same metric: the largest peak-to-trough equity decline, measured from the running high-water mark.
• The maximum drawdown formula is MDD = (Trough − Peak) ÷ Peak × 100 — dividing by the starting balance instead of the peak understates the damage, often by a factor of 10.
• Most prop evaluations set a max drawdown ceiling of 8–10% and a separate daily loss limit of 4–5%; breaching either one ends the account independently.
• Prop firms measure drawdown four different ways — static, trailing, end-of-day and intraday floating equity — and the same trade sequence can pass one and breach another.
• Recovery is asymmetric: −10% needs +11.1% to get flat, −20% needs +25%, and −50% needs +100%.
• Risking 1% per trade gives you roughly 10 consecutive full losses inside a 10% ceiling; risking 2% cuts that to 5 before slippage and correlation haircuts.
• A good max drawdown for a discretionary retail trader is under 15% annually, and under 6% during a prop evaluation — roughly half your hard limit.

What Max DD Means in Trading

Max DD is shorthand for maximum drawdown — the largest drop your equity curve takes from a peak to the lowest point that follows, before it claws back to a new high. You'll see it written as "max DD" in prop firm rulebooks, "MDD" on performance tear sheets, and spelled out in full on your MT4/MT5 statement footer. All three labels point at the exact same number.

Max DD, MDD and maximum drawdown: same metric, three labels

If a prop dashboard flags "Max DD: 8.2%" and your MT5 report shows "Maximal Drawdown: 8.20%," you're not looking at two different risks — you're looking at one number wearing two outfits. The abbreviation changes by platform; the math behind maximum drawdown meaning stays fixed: it's always the worst peak-to-trough decline your account has logged, expressed as a percentage or dollar figure. When someone asks "what is max drawdown in trading," this is the full answer — nothing more exotic than that.

The high-water mark and why drawdown is peak-relative

Your high-water mark is the highest equity value your account has ever reached. Drawdown only measures distance from that peak — not from your starting balance, not from yesterday's close. This matters because the high-water mark doesn't reset on green days that fall short of a new record. Say your account grinds out small losses for three straight weeks after a strong run: as long as equity never prints a fresh high during that stretch, every one of those daily dips belongs to the same drawdown. It's one continuous peak-to-trough decline, not twenty separate mini-drawdowns stitched together. The equity curve only "resets" the moment it closes above the old high-water mark and starts building a new one.

Drawdown vs a losing trade vs a losing day

A losing trade is a single event. A losing day is a cluster of trades netting negative. Drawdown is neither — it's the cumulative distance between your last peak and your current equity, and it can span one trade or three months of choppy underperformance. This is also where max DD splits into two flavors people conflate constantly: current drawdown is where you sit right now relative to your most recent peak, and it shrinks the second you make a new high. Maximum drawdown is historical — the single worst peak-to-trough decline your account has ever recorded, and it never improves; it can only get replaced by something worse.

The distinction carries different weight depending on where you're reading it. On a prop dashboard, max DD is a hard rule with a dollar floor bolted to it — breach it and the account is done, full stop. In a standalone performance report or backtest, max DD is just a descriptive statistic, a data point for judging how rough the ride was to earn the return. Same formula, very different consequences depending on the context you're trading in.

The Maximum Drawdown Formula (MDD Formula) and How to Calculate It

The maximum drawdown formula is simple: MDD = (Trough − Peak) ÷ Peak × 100. That's it — no smoothing, no averaging, just the worst percentage decline your equity curve recorded from any high point to the lowest point that followed before a new high was made.

MDD = (Trough − Peak) ÷ Peak × 100

The part traders get wrong isn't the arithmetic — it's the denominator. Max drawdown is measured against the peak, not your starting balance. That's not a technicality. The peak is the actual capital you had on the table at the moment risk was live. If you grew a $100,000 account to $150,000 and then dropped to $135,000, you didn't lose 15% of your starting capital — you lost 10% of the $150,000 you actually had at risk. Peak-relative measurement reflects reality: risk scales with what's currently in the account, not with what you opened it with six months ago.

Worked example: a $100,000 account and the −12.01% vs −1.1% trap

This is the exact trap that catches traders comparing their own math to a prop firm's dashboard number. Same account, two very different-looking figures depending on which number you divide by.

At Week 6, the account sits at $98,900 — only $1,100 below where it started, a −1.1% figure that looks like nothing happened. But relative to the $112,400 peak reached at Week 3, that's a −12.01% max drawdown. If your firm's daily loss limit or max DD rule is measured off the peak (most are), the −1.1% number is not the one that matters — the −12.01% is. Confusing the two is how traders convince themselves they have more room than they actually do.

How to calculate max drawdown in a spreadsheet with a running peak

Building an equity curve spreadsheet to track this yourself takes four columns and one running peak formula:

Drag column C and D down as new rows are added; the running peak only updates when equity makes a new high, and column D automatically shows negative values everywhere else. Column E gives you the single worst number across the whole history — your max DD.

One catch that trips people up: use equity, not closed balance, if your prop firm measures floating drawdown. Balance only updates when a trade closes; equity moves tick by tick with every open position. If your evaluation rules track floating drawdown intraday, a spreadsheet built on end-of-day balance will understate your real max DD and give you false confidence right up until a breach you didn't see coming.

What Is a Good Max Drawdown? Real Benchmarks by Strategy and Stage

A good max DD is one that's small enough to survive, but never so small it means you barely traded. There's no single "good" number — it depends on your strategy, your holding period, and whether you're in an evaluation or trading a funded account. Here's what actually holds up across styles.

Benchmark ranges by strategy: scalping, intraday indices, swing gold, trend-following

These ranges come from how the strategy is built, not from someone's opinion of "acceptable risk":

A trend system sitting at 25% max DD isn't broken — that's the drawdown its edge requires to eventually catch the 300-pip leg. A scalper at 25% max DD is a red flag.

Return per unit of drawdown: the Calmar and MAR ratios

A drawdown number alone tells you nothing. 12% max DD is great if it bought you 40% annual return, and terrible if it bought you 6%. That's what the Calmar ratio (also called the MAR ratio) fixes: annual return ÷ max DD.

• Calmar around 1.0 — acceptable, the strategy pays you back roughly what it risks
• Calmar around 2.0 — strong, genuinely tradable edge
• Calmar above 3.0 — treat with suspicion; usually a short backtest window, a lucky sample, or curve-fitting rather than a repeatable edge

If someone shows you a track record with 2% max DD and 60% annual return and no Calmar caveat, ask for the sample size before you ask for the strategy.

What counts as good during an evaluation vs on a funded account

The bar moves depending on which side of the pass/fail line you're on. During a Two-Step Challenge with an 8–10% max DD ceiling, a good running max DD is under 6% — that last 2–4% isn't spending room, it's your buffer for the one bad NFP or FOMC print you didn't see coming. Burn through it early and you're trading the rest of the challenge with no margin for error.

On a funded account, tighten it further: target under half your stated limit. If your limit is 10%, you want to be operating comfortably inside 5%. That buffer is what keeps a single bad candle from ending the relationship — a funded account is worth more to you long-term than any one trade, and drawdown discipline is what protects it.

One more thing worth saying plainly: low drawdown with flat or no return isn't skill. It's inactivity. A 1% max DD strategy that also returns 1% a year hasn't solved risk — it's just avoided taking any.

The Four Drawdown Types Prop Firms Use

Prop firms don't all measure drawdown the same way — the four common drawdown types are static, trailing, end-of-day and intraday/floating, and which one governs your account decides whether the exact same losing trade gets you cut or gets you a Tuesday. Same sequence of price action, four different verdicts, because the floor moves (or doesn't) differently in each.

Static (absolute) drawdown

The floor is set once, from your starting balance, and it never moves again — not up, not down. A $100,000 account with a 10% static drawdown limit has a hard floor at $90,000 on day one and on day 300. This is the most forgiving of the drawdown limits once you're sitting on unrealized gains, because your cushion only grows as your equity does. It's common on standard Two-Step Challenge equity accounts where the firm wants a simple, predictable rule traders can plan around.

Trailing drawdown — and the point where it stops trailing

The floor follows your equity high upward, tick for tick, which means your own profits raise the bar you have to defend. Most trailing rules lock once you hit the profit target — the floor freezes at initial balance plus target and stops trailing from there, so a late pullback after you've already hit target doesn't retroactively disqualify you. Before that lock, though, trailing is unforgiving: a trader who runs the account up 6% to a new equity peak, then gives back 5% from that peak, breaches a trailing drawdown limit — because the floor had already climbed with the peak — while the exact same sequence never touches a static drawdown floor set 10% below the original balance. Same trade, opposite outcome. That's the core difference in trailing drawdown vs static drawdown, and it's the one traders get wrong most often when they switch account types mid-career.

End-of-day drawdown and why futures accounts use it

End-of-day drawdown recalculates the floor once, at the daily close, using your closing balance — not your lowest point intraday. This is the standard on CME futures programs trading instruments like ES E-mini S&P 500 and NQ/US100 futures, where volatility during the session can be sharp but the firm only cares where you closed. It lets real intraday heat — a scalp that goes 15 ticks against you before reversing — pass unpunished as long as the account recovers by settlement.

Intraday / floating equity drawdown

Floating equity drawdown counts unrealized P&L in real time, tick by tick, all session long. There's no forgiveness window. A −$4,800 floating loss on an XAUUSD position during FOMC can breach the limit and end the account even if the trade eventually closes green an hour later — the damage is done the moment floating equity crosses the floor, closed or not.

*Sequence: account opens at $100,000, runs up 6% to a new peak, then gives back 5% from that peak intraday before partially recovering by the close.

What Max Drawdown Means on a Funded Account: Closed Balance vs Floating Equity

On a funded account, max drawdown means your equity can never fall below a defined floor — but "equity" has two possible meanings, and which one your firm uses decides whether a rough trade is a bad day or a breach. Closed-balance basis only counts P&L after a trade closes. Floating-equity basis counts unrealized P&L the instant it moves against you, tick by tick, even if you close the trade in profit twenty minutes later.

The single rule detail that decides most breaches

This is the one line in the rules page that traders skip and then argue about in a support ticket. It's usually buried under a heading like "drawdown calculation method," and it says either "based on closed trades" or "based on floating/unrealized equity, calculated in real time." Same account size, same daily loss limit, wildly different survivability during a fast market. A firm running closed-balance drawdown on a $50,000 account effectively gives you room to ride out a spike as long as you don't panic-close at the worst tick. A firm running floating equity does not care what the trade eventually does — if your equity curve touches the floor mid-session, you're done, regardless of what the close price says an hour later.

XAUUSD through an FOMC print: the same trade, two outcomes

You're long XAUUSD going into an FOMC statement. Gold does what gold does on Fed day — it spikes 400 pips in ninety seconds, wipes stops, then reverses. Your open position swings to −$4,300 floating at the peak of the spike. Forty minutes later, after the dust settles and the market re-prices the statement, you close the trade at +$1,100.

On a closed-balance account: that's a winning trade. Your balance went up $1,100, full stop — the floating drawdown never touched the ledger.

On a floating-equity account with a tight daily loss limit: that same −$4,300 dip may have already breached your daily drawdown limit the moment it printed, before your eventual profit ever mattered. You could be locked out of the account with a green trade sitting on your screen. Same entry, same exit, same market move — two completely different outcomes depending on one line in the rulebook.

How to find your drawdown basis before you take a trade

1. Rules page first. Search "drawdown calculation" or "how is drawdown measured" — reputable firms, including For Traders, state the basis explicitly rather than leaving it implied.
2. Prop dashboard widget. Most platforms show a live drawdown meter. Watch whether it moves while a trade is still open — if it does, you're on floating equity.
3. MT4/MT5 terminal. Check the Equity field, not the Balance field, in your terminal's account summary. If Equity is what the firm enforces against, that number is your real-time floor — and it moves every tick your XAUUSD or futures position is open.

Knowing your basis before FOMC, NFP, or any high-ATR session isn't optional homework — it's the difference between sizing a trade you can survive and one that ends your evaluation on a spike you'd have recovered from ten minutes later.

Max Drawdown vs Daily Loss Limit: Two Rules, Two Ways to Fail

The daily loss limit and max drawdown are independent rules that measure different things over different time windows — you can breach one while barely touching the other. Daily loss limit caps how much you can lose in a single session (typically 4–5% of the day's starting balance or equity), while max drawdown caps your cumulative decline from any equity peak (typically 8–10%) and never resets. Most traders conflate the two until an evaluation ends on a rule they didn't think applied.

How the daily loss limit is calculated and when it resets

The daily loss limit is measured from your account's starting balance or equity at the moment of server rollover — not from your last trade, not from the week's high. At rollover (typically 00:00 platform time), the clock zeroes out and a fresh daily ceiling opens up, regardless of how deep in drawdown you were the day before. Miss that distinction and you'll either panic-close a position right before rollover for no reason, or assume yesterday's losses carry forward into today's limit — they don't.

The scenarios where you breach one but not the other

Two traders, same 10% max drawdown ceiling and 5% daily loss limit, two very different outcomes:

Trader A got greedy in one session and hit the wall fast. Trader B never had a bad day by daily-limit standards, but the drawdown never recovered — no new equity high between losses meant every loss stacked on the last. This is exactly why reading only max drawdown mechanics in isolation misses half the risk picture.

Sizing to satisfy both ceilings at once

Your real daily risk budget is the smaller of the two constraints, not the daily loss limit alone. If your daily loss limit is 5% but you're already 6% into a 10% max drawdown, your effective room for the day is 4% — not 5% — because breaching max DD ends the challenge regardless of what the daily counter says. A practical rule that keeps both ceilings honest: never let a single day consume more than half your daily loss limit. Cap it at 2-2.5% instead of running it to 4.8%, and you build a buffer against the low-probability session — the NFP spike, the gap after a surprise Fed print — that turns a normal loss into a rule breach. For the full mechanics on resets, session windows, and edge cases like weekend gaps, see our daily loss limit explainer.

Drawdown Recovery Is Asymmetric: The Maths Nobody Wants to See

A −10% drawdown needs +11.11% to break even, and a −50% drawdown needs +100%. That gap is drawdown recovery asymmetry, and it's the single most underrated number in risk management. Every dollar you lose is easier to lose than it is to win back, because you're compounding off a smaller base. The math isn't opinion — it's algebra: required gain = loss ÷ (1 − loss).

Required gain after −5%, −10%, −20%, −30% and −50%

Run the formula across the range traders actually experience and the curve stops looking linear fast:

Why the curve bends past −20%

Up to about −15%, the required gain to break even tracks close to the loss itself — annoying, but manageable. Past −20% the curve bends hard: you're no longer clawing back a percentage, you're clawing back a percentage of a shrunken account, so each additional point lost demands a disproportionately bigger swing to erase. That's why a −20% drawdown (+25% to recover) doesn't feel twice as bad as −10% (+11.11%) — it's structurally worse, not linearly worse. On a trailing drawdown account, this bites twice: getting back to your old equity high does not restore your old cushion, because the floor trailed up with your peak on the way there. You've spent your buffer and bought back only your balance, not your breathing room.

Why capping drawdown beats chasing returns

Cutting max DD from 20% to 10% has roughly the same effect on terminal equity as a large jump in win rate — but it's far easier to control. Win rate depends on the market cooperating; drawdown depends on your position sizing and your stop discipline, both of which you own outright. In R-multiple terms: if you're risking 1R per trade with a 2:1 risk to reward and averaging 2R winners, clawing back a 25% hole takes roughly 12-13 net winning trades in a row with zero drawdown resets along the way. That's the real cost of letting a drawdown run — not the loss itself, but the string of clean trades required to undo it. Traders who survive prop evaluations aren't the ones with the highest win rate; they're the ones who never let the denominator get small enough to need a miracle.

How Many Losses in a Row Can You Survive? Risk Per Trade vs Loss Streaks

At 1% risk per trade you can absorb 8 consecutive full losses before hitting an 8% ceiling, and 10 before hitting 10%. Drop to 0.5% risk and that number roughly doubles. Push to 2% and you're down to 4 or 5 losses before the account is closed — which is exactly how fast a bad week can end a challenge.

The loss-streak table: 0.5%, 1%, 1.5% and 2% risk inside 8% and 10% ceilings

This is why the 1% rule exists as the default in most trading education, not because it's magic, but because it buys you 8-10 consecutive losses of runway on a standard evaluation before you're staring at a breach. Run the same math at 2% and you're gambling that your losing streak stops at three — a bet the market doesn't owe you.

Slippage, gaps and correlation haircuts

The table above assumes every loss equals exactly your planned risk. It rarely does. A stop on XAUUSD during an NFP spike can fill 20-40% past your level, and a weekend gap on futures can blow through a stop entirely before the market reopens. Treat a "full loss" as 1.2-1.4× your intended risk once slippage and gaps are priced in, which means you should discount every row in the table by roughly 20-30% for real-world survivability — 8 planned losses at 1% risk behaves more like 6 in practice.

Correlation risk compounds this quietly. Long XAUUSD and short US100 are often the same macro bet wearing two tickets — both lean on the same dollar-weakness or risk-off narrative. Two "1%" positions that move together aren't 2% of independent risk; they're closer to one 2% position with extra commission. If you're sizing every trade in isolation without checking what else is open, you're underestimating your real exposure every single time correlated legs stack up.

Why a 60% win rate still delivers 6-loss streaks

A 60% win rate sounds safe until you run the binomial math: the probability of six consecutive losses in a row is (0.4)^6, or roughly 0.4%. Over a sample of 250 trades — a normal quarter for an active trader — that translates to a run of six-in-a-row showing up about once. That's not a broken edge or bad luck; it's losing-streak probability doing exactly what statistics predicts. Traders who don't understand this cut their risk in panic after loss four, right before the setup that would've recovered it.

The fix isn't guessing smaller position sizes when volatility spikes — it's ATR-based position sizing, where your stop distance and lot size adjust to the instrument's current range so 1% risk stays 1% risk whether gold is moving $8 a day or $28. Fixed lot sizes across changing volatility regimes are how a "1% rule" quietly becomes a 3% rule during a news week.

Why Your Live Max Drawdown Beats Your Backtest Every Time

A backtest showing 6% max DD routinely turns into 10–14% live because the backtest never modeled the four things that actually drain an account: fills, spread, correlation, and overfit parameters. If you're planning your risk around the number in your backtest report, you're planning around a number that was never real to begin with.

The four inflators: fills, spread, correlation, overfit parameters

Fills. Your backtest assumes you got filled at the mid-price, every time. Live, during the first sixty seconds of NFP or an FOMC print, gold and index fills slip 3-8 pips past where your model thinks you entered. That's not an edge case — that's the exact moment your strategy's edge is supposed to show up, because that's when the range is biggest.

Spread. Variable spreads widen precisely when your stop sits closest to price — during news, during rollover, during the last hour of the New York session. Your backtest used a fixed or average spread. Live, the spread does the opposite of average right when it matters.

Correlation. If you backtested each instrument separately — XAUUSD alone, NAS100 alone, EURUSD alone — you never saw the day all three lost together. Gold, indices, and dollar pairs share a risk-off trigger. A per-instrument backtest hides that correlation spike; a live account eats it in one session.

Overfit parameters. The exact stop distance, moving average length, or filter that minimized drawdown in-sample got selected because it minimized drawdown in-sample. That's overfitting, not edge. Run the same settings on out-of-sample data and the advantage that looked so clean usually evaporates, taking your low drawdown number with it.

Monte Carlo and trade-order shuffling as a stress test

Here's the fix: don't trust your historical max DD, stress-test it. Take your backtest's trade sequence and run a Monte Carlo simulation — shuffle the order of those same trades 1,000 times. Same win rate, same average R:R, different sequencing. Read the 95th percentile drawdown from that distribution, not the single historical path your backtest happened to produce. That 95th percentile number is a far more honest estimate of what a bad month can actually do to your equity curve, because trade order is random — your backtest just got lucky with the order it drew.

The planning rule: assume 1.5× your backtested max DD

Size for reality, not for the report. Take your backtested max DD and multiply it by 1.5. A strategy that shows 6% needs to survive a live 9% without blowing your daily loss limit or your evaluation ceiling. If 1.5× your backtested number doesn't fit inside your account's max DD rule with room left over, your position size is too big — full stop, before you even open the challenge.

Frequently Asked Questions