← AlgoRishi Blog · algorishi.com

Averaging down on a losing option: the arithmetic of why it rarely works

18 November 2026 · AlgoRishi

You buy a SENSEX weekly call at ₹240. By 11:30 IST it is trading at ₹120. Half your money is gone and the index has not even moved much. A voice in your head says the obvious thing: buy another lot here, bring the average down to ₹180, and you only need a small bounce to get out clean.

That voice is doing real arithmetic. It is just leaving out half the sum. Let us do the whole thing with actual SENSEX numbers.

20SENSEX lot size
₹4,800one lot at ₹240 premium
+50%bounce needed after averaging
₹7,200money now at risk

The setup, in rupees

SENSEX options trade in lots of 20, with strikes every 100 points, and both weekly and monthly contracts expire on Thursday. So a premium of ₹240 is ₹4,800 for one lot. A premium of ₹120 is ₹2,400.

Say you bought one lot of a weekly call at ₹240 — outlay ₹4,800. Premium drops to ₹120, so you are down ₹2,400 on paper. You add a second lot at ₹120 for ₹2,400.

Now you hold 40 quantity at an average premium of ₹180, and you have ₹7,200 in the trade instead of ₹4,800.

What the second lot actually bought you

It bought you a lower break-even. Before averaging, premium had to climb from ₹120 all the way back to ₹240 for you to be whole — a 100 per cent move. After averaging, you only need ₹180, a 50 per cent move. That is a genuine improvement and it is why the idea feels so clever.

It also bought you 50 per cent more exposure to being wrong. Here is every outcome side by side.

Premium at exitOne lot onlyAfter averaging down
₹240 (back to entry)₹0+₹2,400
₹180 (averaged break-even)−₹1,200₹0
₹120 (where you added)−₹2,400−₹2,400
₹40−₹4,000−₹5,600
₹0 (expires worthless)−₹4,800−₹7,200

Look at the third row. At ₹120 — exactly the price where you averaged — both columns are identical. That is the hinge of the whole decision.

Averaging down helps you only if the premium finishes above the price where you added. Below that price, it is strictly worse, and it gets worse faster the further price falls.

So the real question is never "will this bounce a little?" It is: "am I confident enough that this finishes above ₹120 to put another ₹2,400 behind an option that has already fallen 50 per cent?"

Premium path, and the line that decides everything avg ₹180 — new break-even ₹120 — the hinge buy 1 lot at ₹240 add 2nd lot at ₹120 ₹240 ₹120 ₹0 Thursday expiry Above amber you profit. Between amber and red you beat one lot. Below red, averaging costs you more.
Averaging down lowers the break-even but moves the whole position deeper into the zone where a further fall costs more.

Why options are different from shares

Averaging down has a respectable history in equity investing. If you believe a company is worth more than its share price, a lower price is a better price, and you can wait years to be proved right.

An option has no such patience. A SENSEX weekly call expires on Thursday whether you are right or not. Think of a run chase: you need 100 runs and there are four overs left. Buying a second bat does not get you more overs.

Worse, the decay accelerates. On the final day, time value drains hardest, which is precisely when a half-dead premium looks cheapest and most tempting. The option that fell from ₹240 to ₹120 did not become good value. It became a shorter-dated bet on the same view that has already gone against you once.

The most expensive version of this is averaging down on expiry Thursday. If the index does not reach your strike by close, the entire ₹7,200 goes to zero — not most of it, all of it.

The part nobody writes down

There is a second cost, and it is not in the table. When you averaged, you changed your position size after the trade had already gone wrong. Your original plan risked ₹4,800. You are now risking ₹7,200 — a number you never approved when you were calm.

Do that twice in a week and a ₹50,000 account is running positions it was never sized for. Most blown accounts we have seen were not killed by one bad view. They were killed by positions that quietly grew every time a view went bad.

Decide the maximum rupees you will lose on a trade before you enter, and treat it as the total for that idea — including anything you add later. If adding a lot would breach it, you do not have room to average.

What our own machine does

Our SENSEX engine cannot average down, by construction. Each entry gets a fixed stop-loss set before the order goes in, and there is no code path that adds quantity to a position showing a loss. If the stop hits, the trade is finished and the engine goes back to waiting for the next confirmed level break.

That rule costs us. There are sessions where price dips through our stop, reverses, and runs exactly where we thought it would — and we watch it from the sidelines with a realised loss. We keep the rule anyway, because the alternative is letting the losing trades choose their own size.

This is description, not advice. The arithmetic above is simply arithmetic: run it with your own premiums and lot counts before you decide anything.

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.