Drawdown and Risk of Ruin: The Numbers Signal Providers Rarely Show

Drawdown is how far an account falls from a previous high before it recovers, and maximum drawdown is the largest such fall in a track record. Risk of ruin is the probability that losses push an account down to a level from which it cannot, or realistically will not, recover. Both depend far more on how much is risked per trade than on how good the signals are.

In short: a 50% drawdown needs a 100% gain just to get back to where you started. Losing streaks that feel impossible are routine over a few hundred trades. And the same signals, followed at 1% risk per trade or at 10% risk per trade, produce completely different chances of a crippling loss. If a provider won’t tell you their maximum drawdown, measured on equity including open positions, you can’t judge the risk you’d be taking.

What maximum drawdown measures

Drawdown at any moment is the percentage fall from the highest balance reached so far. Maximum drawdown is the worst of those falls over the whole record.

An example account, month by month:

Month	Equity	Previous peak	Drawdown
0	$10,000	$10,000	0%
1	$10,800	$10,800	0%
2	$11,500	$11,500	0%
3	$10,900	$11,500	5.2%
4	$9,800	$11,500	14.8%
5	$9,200	$11,500	20.0%
6	$10,100	$11,500	12.2%
7	$11,900	$11,900	0%
8	$12,600	$12,600	0%
9	$11,300	$12,600	10.3%

The maximum drawdown is (11,500 - 9,200) / 11,500 = 20%, even though the account ended the period up 13%. Someone who subscribed in month 2 spent the next five months watching a fifth of their money disappear.

Balance drawdown vs equity drawdown

This distinction matters more than almost anything else in a signal record.

Balance drawdown counts only closed trades.
Equity drawdown includes open positions at their current, floating value.

A strategy that never closes a losing trade (it holds and waits, or adds more positions) can show a smooth, rising balance while the equity sits deep in the red. Grid and martingale systems are the classic case, explained in why win rate is misleading. Always ask for equity drawdown.

The recovery maths

Losses and gains are not symmetric. After a loss of d percent, the gain needed to recover is d / (100 - d):

Drawdown	Gain needed to recover
5%	5.3%
10%	11.1%
20%	25.0%
25%	33.3%
30%	42.9%
40%	66.7%
50%	100.0%
60%	150.0%
75%	300.0%
90%	900.0%

The curve steepens fast. Small drawdowns are an inconvenience; large ones are close to permanent. A 75% loss needs the account to quadruple.

Losing streaks are normal

Even a decent system has long runs of losses. Here is the probability of seeing at least one losing streak of a given length, computed exactly for independent trades:

Losing streak	Over 200 trades, 45% loss rate	Over 1,000 trades, 45% loss rate	Over 200 trades, 50% loss rate
5 in a row	88%	effectively 100%	97%
8 in a row	16.5%	60.4%	32%
10 in a row	3.5%	17%	9%

So a system that wins 55% of the time has better-than-even odds of producing eight straight losses somewhere in 1,000 trades. What that streak does to you depends on position size. With a fixed fraction of the account risked per trade, the drawdown after n straight losses is 1 minus (1 - risk) to the power n:

Risk per trade	After 5 losses	After 10 losses	After 20 losses
1%	4.9%	9.6%	18.2%
2%	9.6%	18.3%	33.2%
5%	22.6%	40.1%	64.2%
10%	41.0%	65.1%	87.8%

Ten losses at 10% risk per trade takes away nearly two-thirds of the account, which then needs a gain of about 187% to recover.

Risk of ruin

“Ruin” doesn’t have to mean zero. For most people it means a loss big enough that they stop, or that the account can no longer trade at a sensible size. Here are two ways to estimate it.

The classic formula (fixed-size bets)

If you risk a fixed dollar amount on each trade (one “unit”), with a 1 : 1 payoff, win probability p and loss probability q, and you keep trading indefinitely, the probability of eventually losing N units is (q / p) raised to the power N (this is the gambler’s ruin result, valid when p is greater than q).

With p = 0.55:

Account size in risk units	Probability of eventually losing it all
5 units (20% risked per trade)	36.7%
10 units (10% per trade)	13.4%
20 units (5% per trade)	1.8%
50 units (2% per trade)	0.004%

With a thinner edge, p = 0.52, the same 20 units carries a 20.2% ruin probability and 10 units carries 44.9%. A small edge needs a lot of room.

A simulation (fixed-fractional risk)

Most traders risk a percentage of the current balance instead, which shrinks bets after losses and makes total loss mathematically impossible. But a 50% drawdown is very possible. Here are the results of simulating 20,000 accounts, each taking 500 trades at a 55% win rate and 1 : 1 risk-reward, before costs:

Risk per trade	Share of accounts that hit a 50% drawdown	Median ending balance (start = 1.00)
1%	0.0%	1.61
2%	0.8%	2.46
5%	60.1%	6.53
10%	99.9%	12.23
20%	100%	0.93

Two things stand out. First, at 10% risk the median account ends far ahead, yet almost every account suffered a 50% drawdown on the way, which most people would not sit through. Second, at 20% risk the median account ends below where it started, even though every trade has a positive expectancy. Oversizing turns a winning system into a losing one. Why that happens is the subject of position sizing and the Kelly criterion.

Now remove the edge entirely: a 50% win rate, which is roughly what a signal with no real skill delivers before costs. The share of accounts hitting a 50% drawdown within 500 trades becomes 0.5% at 1% risk, 28.7% at 2%, 96.9% at 5%, and 100% at 10%. Resampling an actual trade history in the same way is explained in Monte Carlo simulation for trading systems.

Why the real risk is usually higher

These numbers assume independent trades and that a stop loss always fills at its price. Real trading is less kind:

Leverage. Retail forex is traded on margin, so a position can be much larger than the account. A run of losses can trigger a margin close-out before your planned stops. See leverage and margin in forex.
Gaps. Over weekends and around major news, price can jump past a stop, making the loss larger than planned.
Correlated signals. Several signals on dollar pairs at once behave like one large trade.
Changing conditions. A system’s win rate is not a fixed property; it can fall when markets change, exactly when you’re already in a drawdown.

Worked check: scaling a provider’s drawdown to your size

Suppose a provider advertises a maximum drawdown of about 8%, produced by four straight losses at 2% risk per trade: 1 minus 0.98 to the 4th power is 7.8%. A subscriber who is told the signals are “low risk” and sizes up to 6% per trade would see the same four losses produce a drawdown of 1 minus 0.94 to the 4th power, or 21.9%, nearly three times as deep, and needing a 28% gain to recover. The record didn’t change; the position size did. A drawdown figure is only meaningful alongside the risk per trade it was produced at, which is why the risk per trade is on the list below.

What to look for in a provider’s record

Maximum equity drawdown, with dates, and how long recovery took.
The risk per trade the results were produced at, so you can scale to your own.
The longest losing streak in the record, compared with what the win rate implies.
Whether position size ever grows after losses. That raises the risk of ruin sharply.

The short version

Maximum drawdown tells you the worst fall a record has suffered; the recovery maths tells you why big drawdowns are so hard to come back from. Losing streaks of eight or more are routine over enough trades. Risk of ruin rises steeply with position size, and an oversized bet can make even a positive-expectancy system lose. The safest assumption is that the real drawdown will be worse than the best-looking record suggests.