Grading the exit

You have learned when a buy is sane, the only three honest reasons to sell, and why timing the market is a losing game. This closing module asks the question that decides whether you actually improve: after a sale is done and the price has moved on, how do you judge whether you sold well?

The obvious answer is the wrong one. Most people grade the sale by what the price did afterwards — sold and it fell further, good call; sold and it rebounded, terrible mistake. But that grades the outcome, and the outcome is a mix of your decision and a great deal of luck you did not control. To learn anything, you must grade the process: the quality of the reasoning you had at the moment you decided, using only what was knowable then. This is the discipline that turns a lifetime of investing into a lifetime of learning to invest — and it is the right note to end the whole book on.

Every investing decision is followed by an outcome, but a thick layer of chance sits between the two. You can reason perfectly and lose money; you can reason terribly and make it. Over a single decision, the result tells you almost nothing about the quality of the thinking, because luck had the final say. Over hundreds of decisions, good reasoning does show through in the results — but any one result is mostly noise.

Judging a decision by its result has a name — outcome bias: rating the choice by how it turned out rather than by whether it was sound when made. The poker champion Annie Duke calls the everyday version resulting — working backwards from "it lost money" to "it was a bad bet." Resulting feels like accountability. It is actually the enemy of learning, because it teaches you to copy whatever happened to work and abandon whatever happened not to, regardless of whether the reasoning was any good.

This is the principle the whole shelf has been building toward. Grade the process, not the outcome — the idea, argued by Michael Mauboussin, that in any field where luck plays a large role, the only reliable path to improvement is to judge the quality of decisions independently of their results. A good decision that loses is still a good decision. A bad decision that wins is still a bad decision. Master that one uncomfortable sentence and you have the core of investing maturity.

Separate the two axes and every sale falls into one of four boxes. The decision was either good (you sold for an honest reason, or held because none applied) or bad (you sold on a feeling, or froze through a real change). The outcome was either good (the price move flattered you) or bad (it embarrassed you). Crossing them gives the grid the whole module turns on.

The two diagonal boxes are easy — good all round, or bad all round. The learning lives in the off-diagonal boxes, the ones people almost always misread. Good decision, bad outcome is the sound sell that later looks foolish because the stock rebounded; the temptation is to swear off selling for reasons, which would be a disaster. Bad decision, good outcome is the panic-sell that happened to dodge a crash; the temptation is to bank it as skill and trust your gut next time, which is how disasters are trained. An honest grader defends the good process in the first box and refuses to be seduced by the lucky result in the second.

Here is the grid drawn out. Read it by row, not by column — the row is what you controlled.

Notice the two amber corners share a label — people misread this. They are the whole reason grading the process is a skill and not just common sense. A win off a bad process and a loss off a good one both feel like verdicts on your judgement, and both are lying to you.

Grade three real sales the honest way. illustrative You are reviewing your year, and three exits are on the table.

Sale one — the sound sell that looks foolish. You held a mid-cap at ₹500. Its results showed the growth story had genuinely reversed — falling volumes, a real debt problem — so the thesis broke and you sold at ₹470, taking a small loss. Three months later, a surprise order sent it to ₹700. Your gut screams terrible sell. Grade the row, not the column. At the moment you decided, the thesis had broken and you had no way to know an order was coming; you sold for an honest reason on the facts you had. This is good decision, bad outcome — still a good sell. If you rewrite it as a mistake, you will hesitate next time a thesis truly breaks, and that hesitation will one day be very expensive.

Sale two — the lucky escape. You sold your whole book in a panic during a scary fortnight, with no thesis broken and no reason beyond fear. The market then fell 20%, and you feel like a genius. Grade the row. The decision was to sell on a feeling — precisely the non-reason the last two modules warned against — and it happened to pay. This is bad decision, good outcome — dangerous luck. The honest entry in your journal reads: process was poor, result was lucky, do not repeat. Bank this as skill and you will do it again in a fortnight that instead rockets upward.

Sale three — the clean one. You trimmed a winner that had grown to 30% of the book back to 15%, purely to rebalance; the thesis was intact and you kept most of the position. It kept climbing, and your trimmed slice would have made a little more. No regret is warranted. The decision answered an honest reason — size — and the small "cost" of the continued rise is exactly the insurance premium survival requires. Good decision, and the outcome is beside the point.

Process-grading is powerful, and it has limits worth stating plainly so it is not misused.

It does not mean outcomes never matter. Over a long run of decisions, a good process should produce good results, and a strategy that keeps losing money for years is telling you something — perhaps that the "process" was never as sound as it felt. The rule is that a single outcome cannot grade a single decision, not that results are irrelevant forever.

It cannot be done from memory. After the fact, the mind quietly rewrites what you believed before you knew the result, so "I always knew that would happen" is usually the outcome editing your recollection. Honest grading needs a record made at the time — which is why a decision journal, a written note of why you made each call and what would prove it wrong, is the one tool that makes this whole discipline real. Without it, you are grading a memory the outcome has already tampered with.

And it cannot turn a good process into a guarantee. Reasoning well tilts the odds in your favour over many decisions; it never removes luck from any single one. The promise is a better long-run distribution of results, not certainty on the next trade.

Even people who nod along to "process over outcome" slip on these.

1. Banking luck as skill. A reckless call that paid off feels like proof of a gift. Unless the reasoning was sound at the time, it was dangerous luck, and treating it as skill trains the very habit that will eventually ruin you.

2. Punishing a good decision for a bad result. The sound sell that later rebounded gets rewritten as a blunder, so next time you freeze when a thesis truly breaks. Protecting a good process from an unlucky outcome is the hardest and most valuable move here.

3. Grading from memory instead of a written record. Hindsight silently edits what you believed beforehand, so without a decision journal you will grade a flattering fiction, not the decision you actually made.

4. Confusing a noisy year with a verdict. One good or bad year is a small sample soaked in luck. Judge the process across many decisions, not by the last line of one year's statement.

5. Using "process over outcome" as an excuse. It is a tool for honest review, not a shield for laziness. If the results are poor over a long stretch, the honest response is to interrogate the process, not to keep repeating it and calling every loss unlucky.