Position sizing as the only damage cap

Almost nothing about a stock is under your control. You cannot make it rise. You cannot stop it falling. You cannot force the management to be honest, the results to be good, or the market to agree with you this quarter. You picked the company as carefully as you could, and then the world does what it does.

There is exactly one decision that is fully, entirely yours, made before any of that uncertainty arrives: how much money you put in. That single number — the position size, the share of your total portfolio you commit to one holding — is the only real lever you have over risk. Everything else is a hope. This is a control.

And it is a powerful one, because it decides the size of your worst day. A stock you cannot control can still only hurt you in proportion to how much you gave it. Put 5% of your money in a name and even a total wipeout costs you 5%. Put 50% in and a mere halving costs you 25% of everything you own. Same company, same fall — the damage is set entirely by the number you chose in advance.

This module is about treating that number as the most important decision in the whole process. Not which stock. How much. Because which stock decides whether you are right; how much decides whether being wrong can end you.

Beginners spend almost all their energy on selection — which stock, which sector, which story — and almost none on sizing. This is backwards, and dangerously so. Selection decides your returns in a good year. Sizing decides whether you are still in the game to have a good year at all.

The reason is a hard piece of arithmetic that this whole shelf keeps returning to: losses and gains are not symmetric. If a holding falls 50%, it must then rise 100% just to get you back to where you started. Fall 80%, and you need a 400% gain to recover. A big enough loss is not a setback you can grind back from; it is a hole with steep walls. And once your capital is gone, no future good idea can help you, because you have nothing left to put in it. A loss large enough to wipe you out is permanent in a way an ordinary loss is not — there is no coming back from zero, so the first job of a portfolio is not to grow fast but to make sure no single mistake can be fatal.

Position sizing is the tool that guarantees this. By capping how much rides on any one name, you cap how much any one mistake can cost — and you know the cap in advance, in rupees, before the fear arrives. It is a margin of safety built not into the price you pay but into the size you take: leave enough room that even a bad outcome on this position leaves the rest of the book, and you, intact.

There is a second, gentler reason, which the next module develops fully. If you have some genuine edge — some reason to expect this bet to pay more often than not — then sizing is also how you press that edge sensibly, putting a little more into your best ideas and a little less into your weakest, without ever letting any single one grow large enough to sink you. Bet more where the edge is real and less where it is thin — but never so much that one wrong call ends the game. This module fixes the survival half of that rule. The edge half comes next.

The honest way to size a position is to work backwards from the loss you are willing to take, not forwards from how much you like the stock. It runs in three plain steps.

Step one — decide the most you will lose to any single mistake. Pick a number as a share of your whole portfolio: say, no single holding should ever cost you more than 3% of the book. On a ₹10,00,000 portfolio, that is ₹30,000. This is your damage cap — the wound you have decided you can absorb and walk away from, calmly, without it changing your life.

Step two — estimate how far this stock could fall before you would act. Maybe you would exit if the thesis broke and the stock had fallen 40%. Maybe, for a wobblier small-cap, you assume it could simply halve. This is your honest guess at the downside on the money committed.

Step three — size so that fall equals your cap. If ₹30,000 is the most you will lose, and the stock could fall 40%, then the position must be small enough that 40% of it is ₹30,000 — about ₹75,000, or 7.5% of the book. If instead you assume it could fall the full 50%, the position shrinks to ₹60,000. The riskier the holding, the smaller the slice. The tolerable loss sets the size; the size does not set the tolerable loss.

That is the entire method. What makes it powerful is what it does to a single mistake. Below, watch the same disaster — one holding halving — land on a ₹10,00,000 portfolio at four different position sizes.

Notice that nothing about the stock differs across those four bars. It is the same company, having the same terrible day. The only thing that changed is a number you set, calmly, before any of it happened — and that number is the difference between a 2.5% scratch and a 20% wound that needs a 25% recovery just to undo.

Watch two investors meet the same disaster. illustrative

Both hold a ₹10,00,000 portfolio. Both own the same promising mid-cap — call it a specialty manufacturer that both researched and genuinely liked. Investor A, brimming with conviction, made it 40% of her book: ₹4,00,000. Investor B liked it just as much but obeyed a rule he set long ago — no single holding above 10% — so he put in ₹1,00,000.

Then the unexpected happens, as it periodically does: a sudden regulatory blow, a large customer lost, an accounting question. The stock halves. Not a permanent verdict, perhaps — but a brutal, fast 50% fall.

Investor A is down ₹2,00,000 — 20% of everything. Her whole portfolio now needs to climb 25% just to get back to where it started, and worse, she is now frightened and second-guessing every other holding. The size of the loss has taken over her judgement. Investor B is down ₹50,000 — 5% of the book, an annoyance he can absorb without it touching his sleep or his other decisions. Same stock. Same fall. Same research. The only difference was a number each chose before the trouble arrived.

The uncomfortable truth in that table is that Investor A did nothing wrong in her analysis. She may even have been right about the company in the long run. But she let conviction set her size, and conviction is a statement about being right, not a defence against being wrong. Investor B may have felt exactly the same excitement — and then quietly refused to let it choose the number. That refusal is the whole skill.

Sizing is the most reliable risk tool you have, but it is not magic, and believing it does more than it can is its own trap.

It cannot make a bad stock good. A carefully sized position in a failing company still loses money — just a controlled amount of it. Sizing decides how much a mistake costs, never whether you made one. Selection and sizing are two separate jobs; doing the second well does not excuse doing the first badly.

It cannot protect you from correlation. If you cap every holding at 8% but ten of your holdings are the same bet in disguise — the last module's five-banks problem — then a shock hits all ten at once and the "8% caps" add up to a 60% blow. Sizing controls single-name damage; it assumes your names are genuinely different. Size and correlation must be managed together.

It cannot survive a gap. Position sizing assumes you can exit somewhere near your estimated downside. In a real crash a stock can fall past your level before you can act, or become impossible to sell at all. Your assumed 40% downside can turn into 70% overnight. Size with some room for the fall to be worse than you planned — this is why real caps are usually stricter than the arithmetic strictly requires.

The same sizing mistakes recur, and each one feels perfectly reasonable in the moment.

1. Letting conviction set the size. "I'm really sure about this one, so I'll go big" is the single most expensive sentence in investing. Conviction belongs in selection; it must not be allowed to touch the size, because the size is your defence against conviction being wrong.

2. Sizing by how much you can afford to buy, not how much you can afford to lose. Having ₹4,00,000 spare is not a reason to put ₹4,00,000 in one name. The right question is always the loss, never the availability.

3. Averaging a mistake into a giant. A position falls, you add to bring the average down, it falls again, you add again — and a small holding quietly becomes your largest, sized entirely by the falling price rather than any decision. Adding down is a separate, deliberate choice, covered later; it must never happen by drift.

4. Ignoring the recovery maths. People underestimate how a large loss compounds against them. A 33% loss needs a 50% gain to undo; a 50% loss needs 100%. Small, capped losses can be earned back; large ones quietly become permanent.

5. Setting the cap and then ignoring it. A rule written calmly is worthless if it is abandoned the moment excitement or fear arrives. The entire value of sizing is that the number was chosen before the emotion — which only helps if you then obey it.