The written thesis

You have spent the earlier parts of this shelf learning to catch your own wiring in the act — the anchor that will not let go of a buy price, the confirmation reflex that hires evidence instead of weighing it, the way memory edits itself after the result is known. This part is where the catching becomes a habit you can keep. It begins with the plainest defence of all, and the one everything else is built on: writing the thesis down before you act.

Start with a sentence a beginner writes all the time. "This is a high-quality company with great management, and it will compound for years." It may even be true. But notice what it cannot do. It names no evidence you could be shown later. It draws no boundary — no price at which the case stops making sense, no fact that would end it. And crucially, there is no version of the future in which this sentence is wrong, because nothing in it can be checked. Six months on, whatever the stock does, you can nod along with it. A claim that survives every outcome has told you nothing.

That gap — between an opinion you hold and a claim reality can settle — is the whole subject of this module. A written thesis is the tool that closes it: a short, dated statement, set down before the feeling hardens into a trade, that commits you to a claim specific enough to be proven wrong. It is the single defence the earlier biases cannot quietly rewrite, because it exists outside your head, on paper, timestamped, refusing to change its story to match the price.

The reason the earlier modules kept returning to writing is not a stylistic tic. It is the one move that works. You cannot out-feel your wiring: the fast, emotional system is quicker than you, always present, and floods hardest in exactly the moment a decision matters. Resolve — "this time I'll stay disciplined" — is itself a feeling, and the market is very good at drowning feelings with other feelings. A thesis you merely hold in your mind is stored in the same place the biases live, and it will be edited by them without your noticing. Anchoring will slide your sense of "fair value" toward whatever you paid. Confirmation bias will let the good news in cheaply and make the bad news pass a committee. Hindsight will convince you that what happened is what you always expected.

A thesis written down in a calm hour does none of that. It does not get tired, greedy, or frightened. It does not rewrite itself when the price moves, because ink on a page cannot. When you wrote it, you did not yet know how the story would end, so it preserves your real uncertainty — the thing the outcome most wants to erase. The decision you write before you feel is the only one the feeling cannot later revise. That is the entire mechanism: not more willpower, but a slower, honester witness set down ahead of time and then obeyed.

There is a discipline hidden in the writing, too. To fill the fields honestly you have to admit what you actually understand about the business and what you are guessing at — and the act of writing exposes the difference between the two far more ruthlessly than thinking ever does. A vague thesis embarrasses itself on the page. That embarrassment is the tool working: it is cheaper to feel it while writing than after buying.

A usable thesis fits on a single page, and it has seven fields. Short enough that you will actually fill it in; specific enough that vague thinking has nowhere to hide. Here is the whole template, with what each field is really for.

1. The business, in one line. What does it sell, and to whom? If you cannot say it in a sentence a friend would follow, you do not yet understand it well enough to own it. This is your circle of competence, tested out loud.

2. Why it is worth owning — the evidence. The facts, not the feeling. Cash that tracks profit, a widening distribution, a clean balance sheet — things you could point to in a filing. If every line here is an adjective ("great", "strong", "quality"), you have written a mood, not a case.

3. Why now, and at what price. What makes today an entry rather than "someday"? A number belongs here — the price you are paying and what it implies — because a good business at a punishing price is still a bad purchase.

4. The key risks. The two or three that could actually hurt the case, written by you before you are attached to the outcome. This is the reject pile, built first: the honest search for what would break the thesis, done while your mind is still fair.

5. What would change my mind — the break condition. This is the field that makes it a thesis, and it has a name: the invalidation, the single checkable fact at which the case is dead and you act. Not "if I lose faith" — faith follows the price. A number you could look up and a stranger could verify.

6. Position size and max loss. How much of the portfolio, and the rupee loss you can carry and still read bad news calmly. Size is not just maths; it decides how honestly you can think after the price falls.

7. The review date. The day you will next open this on purpose — not on a whim, and not because the price twitched.

Fields 5 and 6 are load-bearing. You can be a little vague in the others and the thesis still limps along; drop either the break condition or the size and it stops being a thesis at all. The first becomes a hope with no exit; the second, a bet with no floor.

The two load-bearing fields are worth one more word each, because they are where beginners go wrong.

The break condition has to be falsifiable. Falsifiability is a plain idea borrowed from science: a claim only counts as real if there is some result that would show it to be false. "The management is good" is not falsifiable — no fact can settle it. "Receivable days will stay under 90" is: you check a number, and it either did or it did not. A break condition that your mood can reinterpret is not a break condition; it is a feeling in disguise. This is the same instinct as disconfirmation from the confirmation-bias module — you write, in advance, the evidence you would least like to find.

And the size has to protect your conduct, not just cap your exposure. Position sizing is the deliberate choice of how much of the portfolio one idea gets, made by how much damage it can do if wrong — not by how excited you are. A position that looks rational at 6% can become unbearable at 25%: the thesis did not change, but your ability to think about it did. Writing the max loss in rupees, plus a rule like "no averaging down without a fresh thesis", is what turns a percentage into armour.

Templates stay abstract until you see one completed. Here is a worked thesis for a made-up company — every figure invented to teach, none a claim about any real business.

Read field 5 again, because it is the hinge. "If receivable days exceed 90 for two consecutive quarters, exit." That single line does three things at once. It is checkable — receivable days sit in every set of accounts, and you can read them off without asking anyone's permission. It is patient — "two consecutive quarters" refuses to panic at one noisy print. And it is pre-committed — you decided the exit while calm, so the frightened version of you months later does not get a vote. Contrast it with what a beginner usually writes: "I'll sell if it starts going wrong." What is "wrong"? By the time you feel it, the price has already fallen and your judgement is already compromised. The difference between those two sentences is the difference between a thesis and a wish.

Field 6 does the same job for behaviour. "6%" alone is arithmetic. "6%, max loss ₹30,000, no averaging down" is a rule your future self can be held to — a floor under the mistake and a lock on the one destructive habit a falling price invites.

The best way to feel why fields 5 and 6 are load-bearing is to try to finish a thesis without them. The builder below has all seven fields. Fill them — or prefill the worked example — and then delete the break condition and the size and watch what the checker does. illustrative

It refuses to call it a thesis. Not because the other five fields are weak, but because a case with no way to be proven wrong and no floor under the loss is, by definition, a hope. The tool will not let a glowing write-up and a fair price stand in for the two lines that actually make the claim usable.

Now put the finished thesis to work on an ordinary bad day. Suppose the packaged-foods holding is down 20%, and your feed is full of worry about a weak festive season. Without a written thesis, that 20% is a wound, and the only instrument you have to read it with is how much it hurts. With the thesis, the fall becomes a question with an answer: is the break condition tripped? You open the accounts, find receivable days at 62 — nowhere near the 90 you named — and the thesis holds. The price fell; the case did not. You do nothing, calmly, because the plan already accounted for a price that wanders while the business stays intact.

A written thesis is only a defence if you use it the way it was designed, which is sparingly. The commonest way to ruin a good thesis is to reopen it constantly. A red price tempts you to reread the whole case every evening, and each rereading is another chance for the fast mind to renegotiate the terms — to soften the break condition, to explain away the risk, to talk yourself into "just a little more" size. A thesis you revisit daily is not being consulted; it is being eroded.

There are exactly two honest reasons to reopen it. The first is the review date — the pre-set day, usually tied to results, when you deliberately check whether the evidence in fields 2 and 4 has improved, weakened, or stayed unresolved. The second is the day the break condition actually triggers. That is it. A price that is merely red, a headline a friend forwards, a bad feeling at midnight — none of these are triggers unless they move a fact you named. This is what it means to obey the paper: the paper decides when you think about the position, so that the feeling does not.

When the review date comes, the discipline is to update the thesis honestly rather than defend it. If the evidence has genuinely weakened but the break condition has not tripped, you note it and watch more closely. If the break condition has tripped, you act — even, especially, when you do not want to. And when you close a position, the thesis becomes something more valuable still: a record of what you actually believed, which the next module turns into a decision journal you can learn from without your memory quietly rewriting the past.

Writing a thesis protects you from a great deal, and it is honest to be clear about what it does not do.

It does not make the thesis correct. A beautifully structured, fully falsifiable thesis can still be wrong about the business — you can name a clean break condition on a company you have simply misjudged. The template disciplines your thinking; it does not supply insight. What it guarantees is not that you were right, but that when you were wrong, you will find out cleanly and early instead of slowly and expensively.

It does not work if the break condition is theatre. A falsifiable-looking line you never actually check, or one set so loose it could never trip, is worse than none — it gives the comfort of a defence without the defence. The test has to be one you will really look up, at a level that could really be crossed.

And it cannot survive your dishonesty at the review date. If, when the break condition triggers, you quietly move the goalpost — "90 days was arbitrary, let's say 110" — the thesis has not failed; you have overruled it, which is the exact behaviour it was built to prevent. The discipline is not in the writing. It is in obeying what you wrote.

The same few substitutions turn a real thesis back into an opinion. Named once, they are easy to catch on the page.

1. Writing reasons to buy and calling it a thesis. The pleasant half — why it is wonderful — is not the thesis. The thesis is the half that could exit you. A note that is all case and no break condition is a sales pitch to yourself.

2. An unfalsifiable break condition. "Exit if the story changes" or "if I lose conviction" cannot be checked, and conviction obligingly follows the price. If a stranger could not verify whether your break condition had triggered, neither can you.

3. A size with no rule attached. "6%" tells your future self nothing about how to behave when the position is down. Without a max loss and a no-add rule, the size is a number, not armour.

4. Rereading it whenever the price moves. Each reopening is a chance to renegotiate. A thesis consulted only on the date and the trigger stays a plan; one reread nightly becomes a mood.

5. Moving the goalpost when the break condition triggers. The moment you rewrite the exit to avoid taking it, the thesis is dead and confirmation bias is driving. The line was written calmly for exactly this moment; honour it.