What a Drawdown Does to Your Position Size

Almost everyone who has been given a risk rule has been given the same one: risk a small fixed percentage of your account on any single trade. It is the right rule. It is close to the only rule that reliably keeps a bad run from becoming a final one, and this blog has argued for it before.

What the rule is rarely explained alongside is what it does to you during a losing run. A percentage needs something to be a percentage of, and that something is your account, and your account is exactly the thing a losing run is making smaller. So every loss does two things at once: it takes money, and it shrinks the size of every trade you take afterwards. The first effect is the one you notice. The second is the one that decides how the next two months go.

Nobody sits down and decides this. It arrives as a side effect of a rule adopted for a different reason, it is not written down anywhere, and it is the mechanical fact underneath most of the bad decisions people make while they are down.

The denominator is doing something you did not ask it to

Work through the arithmetic once, because it is short and most people have never actually done it. Take an account, take a risk budget expressed as a percentage of it, and lose a few trades. The percentage has not changed — you are as disciplined as you were on day one. But the number of currency units that percentage now represents is smaller than it was, because the balance it is calculated from is smaller than it was.

Which means your next winning trade, sized correctly under exactly the same rule, brings back less than the equivalent losing trade took out. Not because you did anything wrong, and not because the market changed. Because you are now betting a percentage of a smaller number.

Run it far enough and the shape is unmistakable: the way down is quicker than the way back, and the gap widens the deeper the hole gets. This is not a market observation or a claim about win rates. It is arithmetic, and it is true for every trader who sizes off a live balance, including the ones with an edge.

A drawdown does not only cost you money. It resizes every trade you take afterwards, and nobody decided that it should.

It is worth being clear that this is the rule working, not failing. Automatic de-leveraging into weakness is exactly the property that keeps an account alive through a run of losses that would otherwise end it. The problem is not the behaviour. The problem is that it is invisible, so it is never planned for — and an effect you have not planned for is one you will improvise around at the worst possible moment.

Two policies, and almost nobody picks one on purpose

There are two coherent answers to what the percentage should be a percentage of, and they behave differently enough that the choice matters more than the percentage itself.

• Size off the current balance. Your risk in currency terms falls as you lose and rises as you win. Drawdowns get shallower than they otherwise would, recoveries get slower, and the account is very hard to destroy. This is the conservative answer and it is the right default for most people.
• Size off a fixed reference — a starting balance, a high-water mark, or a figure you re-set on a schedule. Your risk in currency terms stays constant through a losing run, so a recovery moves at the same speed the drawdown did. It also means the drawdown goes deeper, because you keep betting the old size on a smaller account.

Both are defensible. What is not defensible is the third thing, which is what most people actually do: drift between them without noticing. And the drift has a direction. After a run of wins, the balance is up and the calculator gets used, because the bigger number is pleasant to recalculate from. After a run of losses, the calculator gets skipped and yesterday's lot size gets reused, because recalculating means typing a smaller number and admitting where you are.

The net effect is the worst available combination: sizing off the balance when it flatters you and off the high-water mark when it does not. Risk rises after wins and does not fall after losses. That is not a policy, and it is not a discipline problem either — it is an unwritten rule losing to a written one.

The recovery trade

Now put a person inside the arithmetic. They are down, they can see that trades at the current size will take a long time to repair it, and the market is offering something that looks good. The thought that arrives is not reckless on its face. It is: this one is a strong setup, I will take it a little bigger, and that gets me back to level.

Look at what has actually happened to the sizing decision there. Every legitimate input to position size is a property of the trade — the distance to the invalidation level, the risk budget, the value of a unit of movement on the instrument. The input that just determined this one is your profit and loss. The trade got bigger because you were behind, and it would have been smaller if you had been ahead, which means the market did not participate in the decision at all.

If your P&L is an input to your position size, the size is not sized. It is a reaction.

There is a second problem stacked on top, and it is the one that turns a drawdown into an ending. The moment you increase size specifically because you are down, you have raised the cost of being wrong at the exact point in the sequence where your capacity to absorb it is lowest. The rule that would have protected you is the rule you suspended, and you suspended it precisely when it was doing the most work.

This is also why it is a bad idea to sort out your sizing policy while you are in a drawdown. Every decision made from inside one is contaminated by wanting out of it. The policy has to exist before it is needed, which means writing it down on an ordinary day when nothing is at stake.

The deposit that hides the hole

There is a version of this that is quieter and, in some ways, worse. An account is down, and money is added to it. The reasons are usually reasonable — this was always going to be a staged deposit, the balance is now inconveniently small, the minimum lot has become awkward.

But a deposit changes the denominator without any trading having happened. The percentage rule now sizes off a bigger number, so position size goes back up. And the account's own record of the drawdown gets blurred, because the balance curve now contains a rise that was not a profit. Look at that chart six months later and you will read a recovery that never took place.

The fix is bookkeeping rather than willpower: record deposits and withdrawals separately from trading results, and evaluate performance on the trading line alone. If you want to know whether the method is working, the money you put in has to be kept out of the answer.

Writing the policy down

This is short, and it is not complicated. What makes it work is that it is decided in advance and specific enough to be checked afterwards.

• Name the denominator explicitly. Current balance, current equity including open positions, or a fixed reference figure — pick one, in writing, and note the date you picked it.
• Say when it is allowed to change. On a schedule is the cleanest: recompute the sizing figure at a fixed interval, use that number until the next one, and do not recalculate per trade. A number that only updates on a set day cannot be quietly updated in your favour.
• Handle deposits and withdrawals in the same sentence. Either they change the denominator at the next scheduled recompute, or they do not change it at all. The one thing they must not do is change it silently on the day they land.
• Set a drawdown level at which you reduce deliberately, and write what reducing means — smaller risk, fewer positions, or a pause. A planned reduction is a decision; an unplanned one arrives as capitulation after a worse week.
• State plainly that P&L is not an input to position size. It is the one line that stops the recovery trade, and it only works if it was written before you needed it.

Note what is not on that list: a target for getting back to level, and a date. Both are ways of letting the drawdown set your behaviour, which is the thing the policy exists to prevent. The account does not know where its previous high was, and neither does the market.

What this looks like if you follow a signal channel

A published record of calls has no drawdown in it. It has outcomes — this one hit target, this one hit stop, this one was closed manually — and outcomes are the same for everyone reading. The drawdown lives in your account, and it is built out of the order you took them in, the ones you skipped, the ones you were asleep for, and the size you used on each.

Which is why two people can follow the same channel, take the same calls, and end the month in genuinely different places. Start in the middle of a losing sequence and you are sizing down through the part where it turns; start a fortnight earlier and you are sizing up into it. Neither person did anything wrong and neither result says much about the calls.

The practical consequence is the same one this blog reached from the other direction when it looked at how many trades you actually have on: the channel supplies the trade, and everything about how much it costs you is decided at your end. A call cannot know your balance, your drawdown, or which rule you are sizing under. If a source ever does tell you what size to take, that is a much bigger claim than it sounds like, and it is being made without the information required to make it.

Common questions

What drawdown should make me stop trading?

We are not going to name a percentage, because a number that fits one account's volatility, timeframe and purpose is wrong for the next one, and a borrowed threshold is one you will talk yourself out of the first time it triggers. The useful test is about the form of the rule, not its value: it has to be a level you set while you were flat and calm, it has to be checkable without interpretation, and it has to say what happens next — reduce, pause for a fixed period, or review before resuming. A threshold with no stated consequence is not a rule, it is a worry.

Balance or equity — which should I size off?

Equity, meaning the balance including the current value of anything open, is the more honest figure, because it is the money you would actually have if everything closed right now. Sizing off balance while carrying open losers means you are calculating from a number that has already been spent. The cost of the honest version is that your sizing figure moves during the day, which is exactly why the scheduled recompute is worth having: take the equity figure at the fixed moment, and use that one until the next one.

Doesn't sizing down make it impossible to recover?

It makes recovery slower, and that is the trade being made on purpose. Sizing down buys you a much longer runway in exchange for a longer climb. The alternative — holding size constant through a drawdown — recovers at full speed if the run turns and digs a materially deeper hole if it does not. Both are legitimate. What is not legitimate is choosing the second one halfway through the drawdown, because at that point you are not choosing a policy, you are chasing a number.

My minimum lot size is bigger than my risk budget now. What then?

Then that trade is not available to you at that stop distance, and the answer is to skip it rather than to round up. This is the same wall covered in the piece on the one field a signal cannot contain, and a drawdown brings you to it sooner because the budget in currency terms is smaller. The honest options are a wider structure where the stop distance suits the minimum size, a different instrument where the unit is smaller, or not trading it. Rounding the size up is not a fourth option — it is the recovery trade with a technical excuse.

Should I change my method after a bad run?

Not on the basis of the drawdown alone, because a drawdown does not distinguish between a method that has stopped working and one that is having the losing run it was always going to have. What tells them apart is your record: whether the trades were taken as specified, whether the conditions the method needs were present, and whether the same rule was applied when it was inconvenient. That is a job for the journal, and it is the reason the journal has to survive the bad month rather than stop at the first one.

Why we published this

We publish free calls on Telegram — entries, stops and targets, with the outcome of each one posted afterwards, losing trades included — and we run a paid room and say so openly. So it is worth being direct about what an article like this does for us, which is nothing obvious: it tells readers that during a bad stretch the correct response to our calls is to take them smaller.

It is here because the failure it describes is the one that ends accounts, and it ends them quietly. Nobody blows up on the trade they planned. They blow up on the trade they sized while they were behind, in a week when the rule they had been keeping for months was the one thing standing between a rough month and a final one. A reader who has written down their denominator, their recompute schedule and their reduction level before they need any of them is a reader who is still trading next quarter, and that is the only kind of reader a channel like ours is any use to.

It also sets a limit on what our published record can honestly claim. Our outcomes are public and they are posted either way, so you can check what the calls did. What you cannot read off them is what they did to anybody's account, because that depends on size, sequence and starting point — three things that belong to the reader and to nobody else. Any channel that implies otherwise is describing your results using information it does not have.