Two things happened in your portfolio last year. You put ₹4 lakh into a midcap chemicals company after three evenings with the annual report, and it is down 38% because a plant in China restarted and the product price halved. And you put ₹2 lakh into a stock a colleague mentioned in the lift, on a Tuesday, without opening a single filing, and it has doubled because the parent announced a demerger nobody had modelled. At the family lunch you mention the second one. You do not mention the first. Both of those instincts are wrong, and between them they will decide how you invest for the next twenty years.
This lesson is not the comforting version of that idea — the one that ends with "so do not be too hard on yourself". It is the rigorous version, which is considerably less comfortable, because done properly it takes away the excuse as often as it grants one. It also takes as settled what an earlier lesson in this track established: the number of decisions needed to read an investing record statistically is larger than a career supplies. This one asks the question that follows from it. If the results cannot grade you, what can?
A surgeon with a 4% mortality rate is not obviously worse than one with 1%. If the first takes the emergencies at two in the morning and the second operates only on scheduled, healthy patients, the raw figures compare two different jobs. Hospitals know this, which is why serious audit compares each case against what was known when the decision to operate was taken, and against how comparable cases turn out — not against whether this particular patient walked out.
That is exactly the audit a portfolio needs, and almost nobody performs it. The result of any single investment is a mixture of what you decided and what the world then did. Judging the decision by the result alone is called resulting, and it is the standard method in every household in the country.
Why results are such poor evidence here
Investing is a system with a high noise-to-signal ratio and a small number of observations. That combination is unusual, and it is what makes the grading problem hard. In a system with fast feedback and thousands of repetitions — a bowler’s line, a factory’s reject rate — results are excellent evidence about process, because the noise averages out. In a system where a handful of consequential decisions are taken each year and each one takes years to resolve, the noise is still much larger than the signal by the time you run out of career.
| Kind of decision | How often you make it | What the result can tell you |
|---|---|---|
| An intraday or swing trade | Hundreds a year, resolved in days | More than any other row here, and still less than it looks. A few hundred is enough sample size for a large edge to show through the noise; a small one needs thousands |
| A stock purchase held for years | Perhaps five to fifteen a year, resolved over three to five | Very little individually. Across a decade a pattern in what kinds of idea go wrong may show; the selection record itself needs far more decisions than a career supplies |
| An asset allocation | Reviewed yearly, genuinely changed maybe four times in a life | Almost nothing. You will never have enough independent observations for the result to grade the choice |
| Buying a house, choosing an adviser, taking a career risk | Once, twice, three times ever | Nothing at all. There is no sample. Whatever happened, happened once, and the counterfactual was never observed |
Three categories, not two
The familiar version of this idea sorts outcomes into two bins: you followed your process and got unlucky, or you broke your process. That is right as far as it goes, and it lets far too much through, because it hides the most common and most correctable failure of all inside the word "unlucky". A useful audit has three bins, and the middle one is where almost all improvement lives.
- 1Did you follow your own written rule? — a process failure if not
Size, entry conditions, the checklist, the maximum weight, the requirement to read the accounts. This is a factual question with a yes or no answer, answerable from the record and not from memory. If the answer is no, the outcome is irrelevant to the grade. You broke the rule; whether it happened to pay is a separate and dangerous fact.
- 2Was the information you lacked obtainable at the time? — a knowledge failure if it was
This is the bin everybody misfiles. The receivables had been rising for six quarters and it is in the published accounts. The promoter’s pledge was disclosed on the exchange. The scheme document stated the share of the fund that could sit in derivatives. A promoter with a prior regulatory order against him is on the public record. None of that is bad luck. It was there, it was free, and you did not go and get it.
- 3Did the information simply not exist? — then it is an outcome failure, and only then
A regulator changed a rule. A war closed a shipping lane. A competitor’s plant that nobody knew was being built came online. A fire. A pandemic. Here the decision can be graded good and the result bad without contradiction, and the only correct response is to check whether the position was sized so that this was survivable.
- "Nobody could have known the auditor would resign" — the resignation letter and the reason are filed with the exchange
- "The debt just appeared" — the borrowings were in the balance sheet you did not open
- "The fund suddenly changed style" — the change in mandate was notified to unitholders
- "The stock got stuck in a circuit" — the price band applying to it, and any surveillance measure that narrowed it, are published by the exchange
- "The tax bill was a shock" — the treatment was in force before you sold
- A statutory rate or threshold changed after you committed
- A key customer of your company lost a contract in another country
- A fraud so constructed that the auditors and the lenders also missed it
- A commodity price that moved for reasons no forecaster had in their model
- A promoter’s sudden death
The fourth case: you followed the process and the process was wrong
There is a case the three bins do not cover, and it is the one people escape into once they learn the framework. Suppose you followed your rules exactly, the information genuinely did not exist, and the same thing has now happened four times. At some point "I followed my process" stops being a defence and becomes the finding. A process can be faithfully executed and still be a bad process — and the difference between the two sentences matters enormously.
| I failed the process | The process failed | |
|---|---|---|
| What the record shows | Rules exist, and this decision departed from them | Rules exist, and this decision complied with them |
| What the evidence looks like | One departure is enough to establish it | A pattern across many decisions, with a common feature |
| The right response | None to the rule. Fix the execution, or ask why the rule is one you keep breaking | Change one element of the rule, then wait for enough new decisions to judge it |
| The failure mode | Rewriting the rule so that what you did becomes compliant | Changing the rule after every bad result, so nothing is ever tested |
Set it to being right two times in five, with a gain four times the size of the loss — a decision worth taking every time it is offered. Then look at the path it draws, and at how long it spends below where it started. That stretch is what a bad quarter feels like from the inside, and it is also what a correct process looks like while it is working.
The review that produces findings
- Grade the decision before you look at the price. Open the note first, decide which bin it belongs in, and only then check what the holding has done. The order is not a formality — reversing it contaminates every grade you assign.
- Grade the wins with the same instrument as the losses. A profitable violation is the single most damaging event in an investing life, because the market has just paid you to break your own rules and the lesson sticks in a way no reasoning removes.
- Count the bins, not the rupees. Four knowledge failures in a year is a finding. A 9% loss is a number that could have arisen from anything.
- Write the finding as a behaviour, not a resolution. "Do not buy from tips" is a wish. "No purchase within 72 hours of first hearing the name, and never before the last two annual reports are open on the screen" is a rule that can be complied with or breached.
- Change one thing, then leave it alone long enough to be tested. For a long-horizon portfolio that means years, not quarters — which is uncomfortable, and is the actual cost of operating in a system with this few observations.
The annual review
Your portfolio is up 3% against an index up 14%. Going through the year: two holdings you researched properly are down heavily on sector-wide price falls that were not foreseeable; one holding you bought without reading the accounts is up 90%; and you exited a good business early because the position frightened you at a size you had chosen yourself.
You buy a smallcap after reading two annual reports and sizing it at 3%. Eight months later it falls 45% after the company discloses a related-party transaction that was disclosed, in the notes, in the year before you bought. How should this be graded?
Ek stock teen shaam padhkar liya — 38% gir gaya. Ek lift mein sunkar liya, bina kuch khole — double ho gaya. Ghar mein doosre wala hi sunaate ho. Par sahi hisaab teen dabbon mein hota hai: apna rule toda? ya jo baat likhi hui thi woh padhi hi nahi? ya woh baat kahin thi hi nahi? Beech wala dabba sabse bada hai, aur log usi ko "kismat kharab" bolkar chhod dete hain. Aur haan — rule todkar paisa ban jaaye, woh sabse khatarnaak din hota hai, kyunki market ne aapko galat aadat ka inaam de diya.
- Judging a decision by its result is resulting, and in a noisy system with few observations it teaches the wrong lesson.
- Sort into three bins: you broke your rule, the information was obtainable and unread, or it did not exist.
- The middle bin is where nearly all improvement lives, and it is habitually misfiled as bad luck.
- A rewarded rule violation is the most damaging outcome available, so grade the wins with the same instrument as the losses.
- Grades are only meaningful against a note written before the outcome was known — memory rewrites the reasoning to fit the result.
Mark it done to track your progress through the curriculum.
Common questions
Short, direct answers to what people ask about this topic.
- resulting meaning in investing
- Resulting is grading a decision by how it turned out rather than by what was known when it was taken. A loss can follow a well-researched, correctly sized purchase because the world then did something nobody could have foreseen, and a gain can follow a tip acted on in a day with no filing opened. Because investing supplies few decisions and a great deal of noise, the result is a poor witness to the quality of the reasoning behind it.
- judging a decision by how it turned out rather than by how it was made is known as
- Resulting. The term comes from poker and describes treating the outcome as a verdict on the process, which is unreliable in any system where luck contributes heavily to each individual result. A more honest audit asks three questions instead: did you follow your own written rule, was the information you lacked obtainable at the time, and did the fact simply not exist yet.
- how do I tell a bad investment decision from bad luck
- Ask whether the fact that hurt you was obtainable, free, before you paid. Rising receivables, a promoter pledge, an auditor resignation, a change in a scheme’s mandate and a surveillance measure narrowing a price band are all published, so a loss caused by one of them is a knowledge failure rather than bad luck. Genuine bad luck is confined to information that did not exist at the time — a rule change, a fire, a competitor’s plant nobody knew was being built.
- how many trades do you need before your returns prove you have skill
- More than most investing lives supply, and the number rises sharply as the edge gets smaller. A few hundred short-horizon trades can let a large edge show through the noise, while a small one needs thousands. A long-term selection record of five to fifteen decisions a year, each resolving over three to five years, and an asset allocation genuinely changed four times in a lifetime, never accumulate enough independent observations for the returns themselves to grade the process.
- is a profitable trade still a mistake if I broke my own rules
- Yes, and it is the most damaging outcome available, because the market has just paid you for breaking a rule and that lesson sticks harder than any argument against it. Grade it as a process failure and record it as one, whatever the profit was. The real cost is rarely the money made on that position — it is the size of the next position taken the same way.