How to stop overtrading on Angel One

Most advice about fixing your trading starts with "from tomorrow". Start a journal from tomorrow. Track your discipline from tomorrow. Set your limits from tomorrow.

Which means you spend the next three months collecting data before you learn anything, and in the meantime you keep doing the thing that is costing you.

You do not have to wait. Your Angel One tradebook already contains months of evidence about exactly how you trade badly. It is sitting there right now. This post is about reading it.

What your existing tradebook can already tell you

You might think a record of past trades only shows P&L. It shows much more than that, because most discipline failures leave a timestamp.

From an export of trades alone, with no other data, you can reconstruct:

• Trades per day, and how the bad days differ from the ordinary ones
• Time between trades, which collapses when you are on tilt
• Consecutive losses, and what you did immediately after each streak
• Position size relative to your own average, which is where size creep shows
• Entry times, and the P&L of everything opened after 2 PM
• Whether you kept trading past the point you had said you would stop

That is most of a discipline audit, from data you already have. The only things you genuinely cannot recover are the live ones: what your unrealised P&L was doing minute by minute, and what you were feeling. Everything else is in the file.

Do the audit

Set aside an hour. This is worth more than anything else in this post.

1. Export the last three months from Angel One. Their reports section gives you tradebook and P&L statements. Three months, not one, because you want enough bad days to see a pattern.

2. Build a trades-per-day column. Note the median and the worst day. Almost everyone guesses the median right and the worst day badly wrong.

3. Add a "minutes since previous trade" column. Sort ascending. Look at everything under five minutes. Then check what the trade immediately before each of those was. If the answer is usually "a loss", you have found your problem, and it has a name.

4. Add a "size versus my average" column. Flag anything more than 1.5 times your normal. Then total the P&L of just those trades. For most people this is a large negative number produced by a small handful of trades.

5. Sort into two buckets and total each.

• A: I would take this again. You can describe the setup. You had a reason before entering.
• B: everything else. Cannot explain it now. Minutes after a loss. Wrong size. Instrument you do not normally touch.

Bucket A is usually somewhere between flat and decently profitable. Bucket B is where the account bleeds. Your strategy was never the problem.

6. Score each day out of 100. Start at 100 and deduct 10 for each trade taken within ten minutes of a loss, 15 for each oversized position, 10 for each trade beyond your usual count, and 25 for any day you kept trading after a loss you had called your limit.

Now plot the daily score against daily P&L. Two findings, reliably:

Your worst discipline days are not always your red days. Green days scoring 45 are the dangerous ones, because nothing about them felt wrong at the time.

The score falls before the money does. Discipline decays over several sessions before the account notices. A run of 60s is a warning. A big red day is the invoice.

This is what LossGuardian does automatically on Angel One: roughly 80% of the discipline stats, including limit breaches, execution-window violations and consecutive-loss counts, are re-derivable from your tradebook alone, for sessions the app was never running. The stats that genuinely need live data are marked "not available" rather than guessed at, which matters, because a fabricated number is worse than a missing one.

What the audit will tell you to do

You will not need a general rule. You will have your own.

If the "minutes since previous trade" column is full of two-minute re-entries after losses, your rule is a cooldown: ten minutes, no new position, after any losing trade. Leave the screen, because a cooldown spent staring at the chart is a countdown.

If the oversized-position column holds your worst losses, your rule is about size, and specifically about never increasing it inside a losing session. Work your quantity backwards from your per-trade risk, as in how to manage risk in trading.

If the damage is concentrated after 2 PM, your rule is a cutoff time, and it is the easiest rule on this list to keep because it is not a judgement call.

If your trade count on bad days is triple your median, your rule is a count. Median plus one. You cannot argue with a count the way you can argue with a rupee figure.

Most people find two of these apply. Pick those two. Do not adopt all four.

Fill the gaps as you go

One practical thing about running a tool alongside Angel One. You will not have it open every session. You will be travelling, or on your phone, or you closed it and forgot.

Trades executed while the app was closed still get detected and folded into your state and discipline data afterwards, so your history does not develop holes. What it deliberately does not do is retroactively fire alerts for something you can no longer act on. An alarm about a trade from four hours ago is noise, and noise is how people learn to ignore alarms.

There is also a local history database, scoped strictly per broker account and rebuildable from the underlying archive, with day-by-day trends and an intraday drill-down for any past day. The point is that the audit above becomes something you can run any evening in a minute, instead of an hour in a spreadsheet.

The three numbers, before 9:15

Whatever your audit told you, these go on paper before the market opens:

• Maximum trades today: your median plus one
• Maximum loss today: roughly three times your per-trade risk
• After a loss, wait: ten minutes

Method for the second one in how to set a daily loss limit you cannot talk yourself out of, with the arithmetic in the loss limit calculator.

What is still missing

The audit tells you what you did. It does not stop you at 1:47 PM.

LossGuardian reads your positions and orders from your Angel One account as they happen, tracks them against the limits you set that morning, and puts a full-screen alert in front of you when you cross one. Not a corner notification you swat away. Something that interrupts and quotes your own rule back at you.

Angel One is a bring-your-own-key integration. You create the API key in your own Angel One developer account, so it is issued to you and revocable by you at any time, without asking us. It lives in your computer's secure keychain, on your machine, never on our servers.

It reads. It does not act. It cannot place, modify or cancel an order, and it will not square off a position for you. Here is why.

Details on LossGuardian for Angel One, and how it compares with other risk tools.

Start with the export

Do not wait for tomorrow. Download three months tonight and build the trades-per-day column. That single column will tell you whether you have a problem and roughly how big it is, in about fifteen minutes.

---

Related: the general method is in how to stop overtrading, and the anatomy of the afternoon collapse is in why discipline fails at 1:30 PM.