The Professional Trading Mindset

Every trade has two properties, and they move independently. The first is decision quality: whether the position was one your rules permitted, sized so a bad outcome is survivable, with an exit written before entry. The second is the outcome: the rupee number on the contract note. Markets are one of the few arenas where these two come apart routinely. A doctor with a sound diagnosis usually gets a recovering patient. A trader with a sound decision gets a loss often enough that the connection is easy to lose faith in.

Take an illustrative case. Suppose NIFTY is at 24,500 and India VIX is elevated after a run of wide sessions. You sell a 24,500 straddle, sized so that a 2 per cent adverse gap costs about 1 per cent of your capital, with a written rule to exit if the combined premium doubles. Overnight, an unexpected policy headline gaps the index 320 points. You exit at the level you wrote down, and you lose. Nothing about that decision was wrong. The loss was the price of holding a position that was always going to lose sometimes.

Now reverse it. You buy a far out-of-the-money weekly call twenty minutes before the close because the index is running, with no exit level and a quantity that is a fifth of your account. The index adds another 200 points and the premium triples. Your account is up. Your process was indefensible. This is the more expensive of the two cases, because the account never sends a bill for it — the bill arrives three months later, when the same behaviour meets a session that does not cooperate.

The four combinations of decision and outcome each teach something different, and only two of them teach it clearly. The table below is worth keeping somewhere you can see it, because the two diagonal cells are where almost all lasting damage and almost all unnecessary self-doubt come from.

A single trade is a sample of one, and a sample of one supports no conclusion. A short strangle that expired worthless does not prove the strikes were far enough out; it proves the index did not travel that far this once. A long call that went to zero does not prove the view was wrong; it may mean the view was right and the expiry was too near. Once you accept this, the question you ask before entering changes shape entirely. You stop asking whether this trade will work and start asking what happens if you take this exact position a hundred times.

That reframing does real work at the moment of sizing. If you would be comfortable repeating a position a hundred times at two lots but not at eight, then eight lots was never a view about the market — it was a bet on this particular instance being the good one. The size is where probabilistic thinking either exists or does not. It is also the reason position sizing and risk per trade are treated as separate modules in this book rather than as a footnote to strategy selection.

The practical version of this fits in one line in a notebook. Before the market opens, for any structure you might take that day, write the loss number first. If NIFTY gaps 2 per cent against this at the opening tick, what is the account worth, and can I take the next five signals afterwards? If the honest answer is no, the idea is not the problem. The size is. Traders almost never abandon a sound approach because it stopped working; they abandon it because they were holding too much of it during the ordinary bad patch.

Expectancy is the average rupee result you can expect per trade if you keep repeating the same decision. It is the only number that answers whether an approach is worth executing, and it is deliberately indifferent to how the last trade went. It has four inputs: how often the approach wins, how much it makes when it wins, how often it loses, and how much it loses when it loses. Costs are the fifth input and the one retail traders leave out most often, which is why so many approaches look positive on paper and are negative on a contract note.

Work an illustrative example. Suppose a NIFTY debit spread setup produces 100 round trips. Thirty-eight of them make an average of ₹6,200 and sixty-two lose an average of ₹3,100. Gross expectancy is (0.38 × 6,200) − (0.62 × 3,100) = ₹2,356 − ₹1,922 = ₹434 per trade. Now subtract round-trip costs. If brokerage, exchange transaction charges, STT, stamp duty and GST together came to roughly ₹250 per round trip, net expectancy falls to about ₹184. The edge survives, but nearly six rupees in every ten of it went to costs. Verify current charges against the NSE circulars and your own contract notes rather than assuming a figure.

Notice what the example does not require: a high hit rate. Thirty-eight winners in a hundred is a losing majority, and it is still a positive expectancy because the winners are twice the losers. The reverse trap is more seductive. An approach that wins 70 times out of 100 for ₹1,500 but loses 30 times for ₹4,000 has an expectancy of (0.70 × 1,500) − (0.30 × 4,000) = ₹1,050 − ₹1,200 = negative ₹150 per trade. It feels wonderful to trade and it drains the account. Selling far out-of-the-money options without a defined stop produces exactly this shape.

Expectancy is also what makes a drawdown interpretable. If you know the number and you know it was measured over a decent sample, then a bad month is a distribution doing what distributions do. If you never calculated it, every drawdown becomes an open question about whether the whole approach is broken, and open questions during drawdowns get answered by fear.

Consider a sequence that plays out constantly in weekly options. You buy a NIFTY call at ₹120. Two sessions later it is ₹55 and your written exit was ₹70, which you did not take. Instead you double the quantity at ₹55, reasoning that your average cost is now ₹87 and the index only needs a small bounce. The index obliges. You exit the combined position at ₹95 and book a small profit. Your account is unharmed. Your behaviour has just been rewarded for ignoring a stop and increasing size into a loss, and the brain does not label the reward as an accident.

The damage is not in that trade. It is in the next one, and the one after, where the same move is made with more conviction and a larger quantity because it worked. Averaging into a losing long option is a strategy with a specific and violent failure mode: theta keeps working against the larger position and, if the move never comes, the whole thing goes to zero rather than to a stop. The one time it fails is arithmetically capable of removing several months of the times it worked.

A profitable win also distorts the next decision through the house-money effect. Traders who are up on the session routinely take a position they would have declined at the open, on the reasoning that they are only risking the market’s money. There is no such category. Once the trade is closed the money is yours, and risking it loosely is the same as risking your opening capital loosely. The clean fix is that the day’s risk cap is written before the open in rupees, and a profitable morning does not raise it.

The only defence against a well-rewarded mistake is a record that grades the decision separately from the result. That is precisely what the behaviour journal in the dual-journaling module exists to capture, and it is why a profitable trade can and should be logged as a deviation.

Losing runs are not evidence of anything until they exceed what ordinary variance produces, and ordinary variance produces far longer runs than most traders expect. The arithmetic is simple enough to do on a phone. If an approach wins 40 times out of 100 and each trade is treated as independent, the chance of five losses in a row is 0.6 raised to the power five, which is about 7.8 per cent — roughly one in every thirteen sequences of five trades. A trader taking four positions a week will meet that several times a year without anything having changed.

The table below runs the same arithmetic for two hit rates. It is a mathematical illustration under an independence assumption, not a claim about how any real approach behaves — real trades cluster, because market regimes cluster, so genuine runs tend to be longer than the arithmetic suggests. That makes the conclusion stronger, not weaker. If eight consecutive losses is already an ordinary event on paper, it is certainly not proof that the approach has stopped working.

What follows from this is a discipline about when you are allowed to change anything. A rule change made in the middle of a drawdown is almost never a response to evidence; there is not enough evidence yet. It is a response to discomfort. The honest structure is to fix a review point in advance — a number of trades or a calendar date — and to hold the rules unchanged until you reach it, while reducing size if the drawdown crosses a threshold you also set in advance.

Size is what converts a survivable run into a terminal one. Risking a fixed small percentage of equity per trade means each loss shrinks the next position slightly, which is exactly the behaviour you want during a bad patch. Risking a fixed rupee amount, or worse a fixed number of lots, means a run of losses hits an account that is getting smaller with positions that are not.

Everything in this module is a frame, and a frame that is not converted into a routine evaporates at 09:15. The routine below takes fifteen to twenty minutes and is deliberately ordered so that every decision requiring judgment is made before any price is on screen. The sequence matters more than the content: once a live chart is in front of you, the decisions you make are contaminated by what the chart is doing at that instant.

Two of the steps are the ones traders skip and then regret. The first is the calendar check, because a large share of avoidable F&O losses come from holding a short-premium position through an event that was on a published calendar. The second is the honest note about your own state. There is no way to trade well through a night of no sleep or a personal crisis, and the only available response is a smaller size or no size at all.

Notice that this routine produces written artefacts — a rupee risk cap, a set of levels, a pre-written exit. Those artefacts are what the mechanical systems module turns into permanent rules, and what the dual-journaling module compares your actual behaviour against. Without them, a review at the end of the week has nothing to compare anything to.

At the close, write down two numbers rather than one. The first is the P&L, which you record and then set aside. The second is an execution grade: what proportion of the day’s trades were on the plan, at the planned size, exited at the planned level. Only the second number is allowed to influence tomorrow’s position size. This feels backwards for about a month and then it stops feeling like anything, because it is simply the correct way to score a process whose outputs are noisy.

Grading is easier when the questions are fixed. Did I take only the trades my rules permitted? Was every size the size I wrote before the open? Did I exit where I said I would, or did the level move while the position was live? Did I skip the trades my rules told me to skip, including the ones that then went on to work? That last question matters, because a skipped trade that would have paid is the single most reliable trigger for abandoning a rule the following session.

A session where you followed every rule and lost ₹6,000 is a better session than one where you improvised and made ₹15,000. Written down, everyone agrees with that sentence. Lived through, almost nobody does, which is exactly why the grade has to be written rather than felt. The record is what survives the mood.

From here the sequence continues. What happens to this frame once a position is live and moving against you is the emotional control module. How to turn the artefacts of your routine into rules that need no judgment is the mechanical systems module. How to record and review all of it so that the leaks become visible is dual-journaling. This module gives you the scorecard; the next three keep you using it.