← AlgoRishi Blog · algorishi.com

Position sizing: the arithmetic of surviving a losing streak

8 September 2026 · AlgoRishi

Ask any trader who blew up an account what killed them, and almost nobody says "I was wrong too often." They say "I was wrong at the worst possible size."

A losing streak is not bad luck. It is arithmetic, and it is coming for every method that has ever existed. Position sizing is the only thing that decides whether you are still at the screen when it ends.

20SENSEX lot size
100strike step
1 in 100odds any 5 trades are all losers at a 60% win rate
−65%capital after 10 losses at 10% risk each

Streaks are normal, not cursed

Suppose you win 60 out of every 100 trades. That is a genuinely good hit rate. The chance that any particular run of five trades is all losers is 0.4 × 0.4 × 0.4 × 0.4 × 0.4, which is about 1 in 100.

That sounds rare until you notice how many five-trade windows there are in a year. Take two trades a day on SENSEX weeklies and you will see roughly 500 such windows in twelve months. A one-in-a-hundred event, offered five hundred times, is not an event. It is a Tuesday.

You are not sizing for the average trade. You are sizing for the worst run you will hit this year — and if your method has any losers at all, that run is longer than you think.

The vada pav stall

Picture a man running a vada pav stall outside a station. Every morning he decides how much batter to make. If he makes a normal batch and it rains, he loses a day's margin and opens again tomorrow.

If instead he spends his entire month's float on one enormous batch because yesterday was busy — and it rains — he does not have a bad day. He has no stall. Nothing about his cooking changed. Only the size did.

The arithmetic nobody wants to look at

Losses do not undo symmetrically. Lose 20% and you need 25% to get level. Lose 50% and you need 100%. The hole gets steeper faster than the ladder gets longer.

Capital lostGain needed just to get back to level
10%11.1%
25%33.3%
50%100%
65%186%
80%400%

Now run ten straight losers through two different sizing choices. Same trades, same method, same wrongness — only the risk per trade differs.

Capital remaining after 10 consecutive losses 100% 30% risking 2% per trade — 81.7% left risking 10% per trade — 34.9% left consecutive losing trades →
Both traders were wrong ten times in a row. One needs 22% to recover. The other needs 187%.

The green trader has a bad month. The red trader has a different life. Neither of them predicted the market any better than the other.

Turning that into SENSEX lots

SENSEX weekly options trade in lots of 20, with strikes every 100 points and expiry on Thursday. So the sizing question has a concrete answer, and it works backwards from your stop-loss, not forwards from your enthusiasm.

The formula is one line:

Lots = (capital × risk %) ÷ (stop-loss in premium points × 20)

Say the account is ₹3,00,000 and you are willing to risk 2% on a trade, which is ₹6,000. Say your rule is that you are wrong if the premium falls 22 points from entry.

Notice what did not enter that calculation: how confident you feel, whether the last trade won, or whether the premium is ₹120 or ₹450. Widen the stop to 40 points and the same ₹6,000 budget buys you 7 lots, not 13. The stop and the size are one decision, taken together, before you click.

Write the lot count down before entry. If you find yourself recalculating it after you are already in the trade, you are not sizing — you are negotiating.

Where the arithmetic lies to you

Two honest caveats, because option buyers face risks a clean formula hides.

First, a stop-loss is an instruction, not a guarantee. In a fast tape the fill can come several points worse than the trigger, so your real risk per trade is usually a little larger than the number in your spreadsheet. Size as though your stop will slip, because sometimes it will.

Second, Thursday is not a normal day. On SENSEX expiry, time decay accelerates hard and premiums can go from ₹60 to near zero in minutes. A position sized comfortably on Monday can behave very differently on expiry afternoon.

Option buyers can lose 100% of the premium paid. No sizing rule prevents a loss; it only limits how much of your account one loss can take. This article is educational and is not advice to buy or sell any contract.

The point

Everyone arrives at the market looking for a better entry. Almost nobody arrives looking for a smaller size. But entries only decide whether a single trade wins. Size decides whether you get to take the next hundred.

Disclosure: AlgoRishi publishes signal tracking on real market prices using virtual capital, for information and education only. Nothing here is investment advice or a recommendation to trade. Futures & options trading involves substantial risk of loss. Past performance is not indicative of future results. AlgoRishi is not a SEBI-registered investment adviser or research analyst.