How to set a daily loss limit you cannot talk yourself out of

Almost every retail trader has set a daily loss limit. Almost none of them are still honouring the one they set three months ago.

The failure is rarely willpower in the way people assume. It is that the limit was never built to survive contact with a losing afternoon. It was a round number picked in a good mood, stored in the trader's head, with no definition of what counts and no consequence for passing it. Of course it broke.

A limit that holds has four properties. It is calculated, it is written, it is unambiguous, and it is placed where you cannot avoid it. Here is how to build one.

1. Get the number from your account, not from your mood

The number people usually pick is whatever loss feels tolerable. That is the wrong input, because your tolerance moves: it is generous on a Monday and nonexistent after three red days.

Start from arithmetic instead.

Step one: decide what a bad month is allowed to cost. Take your trading capital and pick the drawdown at which you would genuinely stop and reassess. Not the figure that wipes you out; the figure at which you have a problem. For most people that is somewhere between 6% and 10% of capital. On ₹5,00,000, call it 8%, or ₹40,000.

Step two: divide across the month's losing days. In a month of roughly 20 sessions, plan for six to eight losers. That is not pessimism, it is a normal distribution of outcomes. ₹40,000 across seven bad days is a daily limit of about ₹5,700. Call it ₹5,000, because round numbers are easier to hold in your head under stress.

Step three: sanity-check it against your average win. If your daily limit is four times your typical good day, it is not a limit, it is decoration: one bad day would erase four good ones and you would still be inside your rule. A daily loss limit ought to sit at or below one to one and a half times your average winning day. If it does not, the limit is too wide, or the strategy has a risk-reward problem the limit is hiding.

That is the number. Notice what it did not come from: how confident you feel, what you lost yesterday, or what someone on YouTube trades.

If you would rather not do the arithmetic by hand, the daily loss limit calculator runs these three steps on your own figures and writes the rule out for you. It works entirely in your browser, and nothing you type is sent anywhere.

2. Write it down as a sentence, not a figure

"₹5,000" is not a rule. It is a quantity, and a quantity can be reinterpreted at 1:30 PM by a motivated person. A rule is a sentence with no room in it.

Write this, in your own words, somewhere permanent:

If my realised plus unrealised loss for the day reaches ₹5,000, I close everything and stop trading for the day. Not a smaller size. Not one more setup. Stop.

Then close the loopholes explicitly, because every one of these is a real argument a real trader has made with themselves:

• Realised or unrealised? Both. Counting only booked losses means holding a losing position is how you stay under your limit, which is precisely the wrong incentive.
• Per account or across accounts? Across. If you trade two brokers, the limit is the sum. A limit you can evade by opening a second app is not a limit. This is the single most common way a rule quietly dies, and it is exactly why a risk tool has to see every broker you use rather than one at a time.
• What about a hedge? It counts. Every open leg counts.
• Does a green afternoon reset it? No. Once you have stopped for the day, you have stopped, even if the market hands you an obvious setup at 2 PM. Especially then.
• Can it be raised mid-session? Never. It can be reviewed on a weekend, with a month of data in front of you, and changed for next week. Not today.

That last one is the load-bearing clause. Most limits do not get broken; they get amended, in the moment, by the person least qualified to amend them.

3. Add the two limits that fire earlier

A rupee limit is a backstop, and it is a late one: by the time you reach it, the sequence that got you there has been running for an hour. Pair it with two rules that trip before the money does:

A trade count. "Six trades, then I am done." A count is beautifully unarguable. You cannot be halfway past it, cannot round it down, cannot decide that this one does not really count.

A cooldown after a loss. "No new position for ten minutes after a losing trade." This one is aimed at the exact hinge of a bad day: the re-entry taken ninety seconds after a loss, for the wrong reason. Ten minutes is usually enough for the urgency to fade to the point where you can look at a chart rather than a grievance. There is more on why the afternoon breaks people in why discipline fails at 1:30 PM.

Together these three cover different failure modes: the count catches overtrading, the cooldown catches tilt, and the rupee limit catches the day that goes wrong quickly.

4. Put it somewhere you cannot avoid meeting it

Here is where most of this advice usually stops, and where it usually fails.

A limit written in a journal is checked when you feel like checking it, which is never the moment it matters. A limit in your head is renegotiated silently. A mental note plus a broker P&L screen means you have to be the one who notices, and at 1:30 PM on a losing day, you are the least reliable observer available.

The limit has to be enforced by something that is not you, at the moment you cross it, in a form you cannot dismiss with a reflex click. That is the entire job LossGuardian does: it watches your positions across your brokers, and when you cross a line you wrote, it puts a hard-stop alert in front of you that states your own rule back at you and does not go away.

What it will not do is close the position. It is read-only: it cannot place, modify or cancel an order, by design, and that is a decision we are not going to reverse. The closing is yours. That is what makes the rep count.

It works the same way on every broker we support, and the price does not change between them: Zerodha, Fyers, Dhan, Upstox, Angel One and Groww. If you trade two of them, both are counted against the same limit, which is the point.

5. Review it monthly, with data

At the end of each month, sit down with the record and ask three questions:

1. How many days did I hit the limit? More than three or four out of twenty and the number is too tight for your strategy. You are being stopped out of normal variance, which is its own kind of damage.
2. How many days did I pass it? Any answer other than zero is the thing to fix, and it is a process problem, not a number problem. Do not widen the limit to make the breaches disappear.
3. What did the breach days look like? Not the total; the sequence. Where was the first loss, how long until the next entry, when did size change. The pattern is almost always the same one, and once you can name yours you can watch for it.

Then adjust once, deliberately, for next month. On a weekend, with the data in front of you, by the version of you that is not down ₹5,000.

That is the whole method. The number is arithmetic, the rule is a sentence with the loopholes closed, the early warnings are time and count rather than money, and the enforcement lives outside your head. None of it is complicated. It is just built to be difficult to argue with, which is the only property that matters at 1:30 PM.

---

See also: how LossGuardian compares with other Indian risk management tools, and what it stores and where on the security page.