Options, Losses and the Hedge I Didn't Need

I came across a project called TradingAgents. Over 1.2 lakh stars on GitHub.

It builds a trading firm out of AI agents. Analysts read the numbers, the news and the charts. A bull and a bear argue. A trader decides. A risk team says yes or no.

Clever. Genuinely clever.

Then I read the fine print. It's a research tool, not advice. Run it twice on the same stock and you may get two different calls.

I stopped there. If a team of machines can't give the same answer twice, what are we buying when we trade on a chart and a tip?

I studied options for my MBA exams. Ninety-nine percent of the chart work, I've forgotten. What stayed is this: research is interpretation. Facts just sit there. Judgement decides what they mean.

So before judging anyone else, I tested mine.

Six things I believed

I wrote down six beliefs and checked each one. Three held. Two held only partly. One fell over.

One belief held better than I expected. Prices run ahead of results. People buy on what they expect, so by the time the quarter's numbers come out, the move has usually happened. Results just confirm it. The exception is a real surprise. Then the price keeps drifting for a few weeks.

Two of my beliefs were quietly fighting each other. If I can forecast earnings, the market can't be a coin toss for me. Both can't be fully true. I'm still sitting with that one.

If retail F&O were a company

Nine out of ten individual F&O traders in India lose money. Not once. Year after year. That's not my guess. It's SEBI's own data.

I read it the only way I know. Like a P&L.

Revenue is whatever traders win. Cost of sales is brokerage, transaction tax and exchange fees, about ₹26,000 per trader in FY24 alone. The margin goes somewhere else. Prop desks and foreign investors booked gross profits of ₹33,000 crore and ₹28,000 crore that year, before costs. Brokers, exchanges and the government take their cut through the fees.

Unit economics? Nine of ten customers lose. Retention? More than 75% of the loss-makers came back after two straight losing years.

No board would fund this business. I wouldn't sign off on it. Yet lakhs of people put their own savings into it.

Why it happened

My first instinct was to blame the traders. People with no clue what an option is.

SEBI's data doesn't let me off that easily. Experience didn't really improve results. So it isn't only ignorance. The system is built this way.

It looks like investing. Charts, data, an app on your phone. It behaves like gambling.

The product looks cheap. You pay only the premium, a few hundred or a few thousand rupees. Feels like a small bet. SEBI itself called expiry-day trading in cheap options largely speculative. Even in FY26, 59% of index options turnover happened on expiry day.

The other side is a machine. In FY24, 97% of foreign investors' F&O profits and 96% of prop traders' profits came from algorithms. A person on a phone app is trading against them.

The wrong people walked in. About 35% of individual derivatives traders held no shares at all, and 78% held less than ₹1 lakh in equity. Over 75% declared an income below ₹5 lakh. They never started with investing. They went straight to betting.

And the losses stay hidden. Wins get posted. Losses don't. So everyone sees winners.

Follow the money and it's simple. Brokers and exchanges earn on volume. Algo desks earn the spread. Influencers earn attention. Win or lose, everyone gets paid while the trader keeps trading. Everyone except the trader.

So why does SEBI allow it?

I asked myself this. Isn't it just a game?

The tool isn't. How most of us use it is.

Derivatives do real work. A mutual fund buys index puts to protect its portfolio, the way a factory insures its plant. Futures tell you where the market thinks prices are headed. Arbitrage traders keep futures and cash prices in line. India brought in index futures around 2000 and options around 2001.

Same contract, two lives. For a fund, a put is insurance. For someone earning ₹5 lakh a year, an expiry-day option is a lottery ticket.

Ban it, and the hedgers suffer while the gamblers move to dabba trading, off the exchange, with no protection at all. So in October 2024 SEBI raised the bar instead. The minimum index contract went from ₹5–10 lakh to ₹15–20 lakh. Each exchange now gets a weekly expiry on only one benchmark index. Buyers pay the premium upfront, and sellers post extra margin on expiry day.

Losses did come down, from ₹1,05,603 crore in FY25 to ₹91,685 crore in FY26. Nine in ten still lose.

Calls and puts, in plain numbers

My exam notes made this complicated. It isn't.

An option buyer has to be right three times. Direction. Size. Timing. All at once.

Illustrative numbers only. Nifty at 24,000. One lot is 75 units. Everything settles in cash.

A call is a bet that Nifty goes up. I buy the 24,200 call at ₹100. I pay ₹7,500.

A put is a bet that Nifty falls. Or insurance against the fall. I buy the 23,800 put at ₹90. I pay ₹6,750.

Now say Nifty climbs to 24,150. I was right. It went up. And I still lose every rupee I paid for the call, because it didn't go up enough.

The premium also melts a little every day. That's why an expiry-day option is so cheap. There's almost no time left for anything to happen.

The seller sits on the other side. Collects the premium on day one. Smiles. Until one sharp move takes back far more.

The hedge I almost bet on

Then I turned it on myself.

Say I hold ₹1 lakh in a Nifty 50 index fund. I want it safe if I sell at the end of December. A put sounds like the obvious answer. I even caught myself saying I need to "bet on" a put.

At this size, it doesn't work.

One Nifty contract is now worth ₹15–20 lakh. One put would cover 15 to 20 times what I own. If Nifty fell, I'd make far more than I lost. That isn't protection. That's a bearish bet dressed up as a hedge.

I'd also pay premium on ₹15 lakh to protect ₹1 lakh. And monthly options expire on a fixed weekday, which may not be the day I want to sell.

The fund has its own rules anyway. NAV once a day. No stop-loss. Redeem before 3 pm and you get that day's NAV. The money lands in two or three working days.

Tax matters more than any of this. Sell within 12 months and it's 20% on the gain. After 12 months it's 12.5%, and the first ₹1.25 lakh of long-term gains each year is tax-free.

So the real hedge is boring. If I need the money by a date, I move it to a liquid fund before that date. If I don't, a month's swing on ₹1 lakh is just what equity does. I hold.

What stays with me

A team of machines can't give the same call twice. ₹91,685 crore in one year is what it costs when people trade as if they can.

A hedge isn't a bet. I knew that for my exams. It took a ₹1 lakh question to feel it.

Maybe that's all a hedge really is. A thehrav before the trade.