Every Friday after the cash close we do a journal pass on the week's trades. The good ones are interesting; the bad ones are useful. This week's Bank Nifty directional gave up 1.6R on a setup that, on paper, looked textbook. The post-mortem below is written exactly the way we run the internal review — so you can see what a loss looks like when it is broken apart honestly.
A reminder before we dive in: everything in this post is an educational reconstruction of a trade we studied on our own books. It is not a recommendation, a tip, or a call. We are a market-education community. We are not SEBI registered. Nothing here should be copied into your own account without your own independent judgement.
The setup
We entered the week with a bearish lean on Bank Nifty. Three pieces of evidence were stacking up:
- Weekly structure: Bank Nifty had printed a lower high on the daily chart after failing to reclaim the 48,900 supply zone for the third time in April.
- Sectoral internals: The top-five private banks were losing relative strength vs the Nifty 50 — a trend we track on a 10-day rolling basis.
- Flow: Index futures OI had built up +7.3% week-on-week while price was flat to down. That is classic short build-up, not long unwinding.
On Monday's pre-market bias sheet, Bank Nifty was tagged as a sell-on-strength candidate, with 48,750 as the invalidation for the week and 48,100 / 47,850 as the two downside objectives. The thesis was clean.
The trade
On Tuesday morning, price opened near 48,680 and within the first 30 minutes rejected the 48,720 zone for the third time on the 15-minute chart. The rejection was exactly what we were waiting for. We entered a directional short via an at-the-money put debit spread on the weekly:
INSTRUMENT : Bank Nifty weekly options STRUCTURE : Long 48,700 PE / Short 48,300 PE (debit spread) ENTRY DEBIT : ~145 per lot MAX RISK : 145 × lot-size = defined TARGET : 300 (roughly 1.9R net) STOP : Spot close above 48,780 on 15m
The position was sized at 1R of our notional weekly budget. Entry log, screenshot, bias sheet — all filed. So far, so clean.
What went wrong (explained)
By Wednesday noon the trade was at +0.8R. Price had drifted to 48,420 on schedule. Our plan said: book half at the 0.8–1.0R zone, let the rest run to the 47,850 objective with a trail.
We didn't do that. Instead, we stared at the chart, told ourselves "it's working, why clip it," and held the full size. By Wednesday 14:45, a short-covering move took Bank Nifty back to 48,610. Our paper gain evaporated. On Thursday morning the index gapped up on a US-session tailwind, our 15-minute stop triggered at 48,810, and we closed for a 1.6R loss instead of a 0.8R partial lock plus a remaining-half break-even.
The setup was correct. The thesis was correct. The entry was correct. The management was where a winning trade was converted into a losing one.
The rule that broke
Our process rule on directional option spreads is mechanical. It is written down. It reads:
When unrealised P&L ≥ 0.8R, → book 50% of the position at market. → move stop on remaining 50% to scratch.
This rule exists for one reason: options are wasting assets. A debit spread that is +0.8R on Wednesday can be flat on Thursday even if price does not move, because theta and vega keep grinding. Locking in half de-risks the week without capping the upside on the other half.
We didn't execute it. We overrode a written, tested, boring rule because the screen felt good. That is the definition of a discretionary override, and it is the single most common mistake in our own internal review log — we see it roughly once every six to eight weeks.
Why the override happens
We've been tracking this behaviour for three years. It almost always comes from one of three sources:
- Recency bias — the last two discretionary holds worked, so we assume this one will too.
- Narrative anchoring — the underlying thesis is so clean that we forget we are managing a position, not proving a view.
- Screen stickiness — staring at a winning position makes it psychologically harder to touch it.
The override is rarely driven by new information. It is almost always driven by emotion dressed up as conviction.
Going into next week
The thesis on Bank Nifty hasn't changed — the structural bias is still to the downside, and the 48,780 supply zone is still intact. But the way we will express it changes:
- We are downsizing the weekly directional allocation from 1.0R to 0.6R until two consecutive weeks of clean execution.
- The 0.8R partial-book rule is being enforced with a pre-set limit order at the computed spread value, entered at the same time as the initial fill. No staring at the screen and deciding.
- A post-trade timer: any override of a written rule must be typed into the journal in the moment, with the reason. If we can't type it, we don't do it.
These are not new ideas. They are old ideas that we have to keep re-learning. That is the honest truth of discretionary trading — the edges are small, the rules are well-understood, and the hard part is being a machine that executes them.
Key takeaways
- A correct thesis and a correct entry do not guarantee a winning trade — management is a separate skill.
- Debit spreads decay. Partial-booking at 0.8R is an anti-theta rule, not a profit-taking rule.
- Discretionary overrides of written rules are the most consistent source of large losses in our own logs.
- Pre-placed limit orders for partial exits remove the moment-of-decision that emotions exploit.
- Write down the override in real time. If you can't articulate it, you shouldn't be doing it.
We will keep publishing these honestly — the wins are never as interesting as the losses, and the education has always been in the latter. If you are studying alongside us, the exercise for this week is simple: pull up your last five trades and flag which ones were executed strictly by the rules you had written before entry. For most traders the number is lower than they think.