Algozone

Backtests fail in predictable ways, and ours is no exception:

• Fills are assumed. The engine books the trigger price. In a fast market you get a worse price, or no fill at all on exactly the trades that would have run furthest.
• Costs bite harder than they look. Brokerage, STT, exchange and GST charges are small per trade and enormous across hundreds of them. On our own data, applying realistic costs removes a large share of the measured edge — which is why the “After costs” toggle exists and why you should use it.
• The same rules were tuned on the same history. When a strategy is designed and measured on one dataset, some of its apparent edge is a description of that dataset, not a law of markets. Live results are almost always worse than backtested ones.
• The instruments are correlated. Sixteen NSE names on one big index day are not sixteen independent bets. Trade counts overstate how much diversification is really present, which makes the results look statistically stronger than they are.
• Option prices are modelled, not quoted. Option figures use a pricing model with realised volatility. Real bid-ask spreads, especially on single stocks, can erase the entire edge.

The mildest of the three, and still capable of losing money steadily.

• You lose on a large minority of trades by design. The stop is hit on a meaningful share of setups. That is not the strategy failing — it is the strategy working. The edge, if it exists, only appears across hundreds of trades.
• Losing runs are longer than people expect. The calculator shows the worst consecutive losing streak in the backtest. Live, you will sit through one of those, and it will feel like the method has stopped working.
• Gaps ignore your stop. Intraday stops do not protect against a stock opening far below your level on news.
• Leverage multiplies both directions. Intraday margin makes a 2% adverse move a much larger loss on your own money. None of the figures on this page assume borrowed capital.
• Liquidity is not uniform. The results above are on large, liquid names. Thinner stocks behave far worse.

The NIFTY signals on this page are detected on the NIFTY futures chart — it is the cleanest continuous intraday series for the index, without the auction artefacts of the spot calculation. The trade itself is a bought at-the-money option, never a futures position. That is a deliberate risk decision:

• A futures loss is not capped. An adverse gap can leave you owing more than the margin you posted. A bought option cannot lose more than the premium paid — you always know the worst case before you enter.
• Margin calls arrive intraday. If the balance falls below requirement the broker squares you off, usually at the worst moment. Buying premium removes that mechanism entirely.
• One lot is large. A NIFTY futures lot controls a notional value many times the margin, so a fraction-of-a-percent index move is a large rupee swing. An option lot on the same signal costs a few thousand rupees.
• You cannot size below one lot in either instrument — but the option lot is affordable at account sizes where the futures lot is not.

If you choose to trade the signals as futures anyway, understand that every figure on this page is calculated on option economics and does not describe what futures would do. The leverage, the loss profile and the margin mechanics are all different.

The most attractive-looking numbers on this page belong to the most dangerous instrument on it. Option percentages are large because the base is small — a 40% loss on premium is a normal outcome, not a disaster scenario.

• Total loss is a routine result. A bought option can expire worthless. Losing 100% of the premium on a trade is ordinary, not exceptional.
• Time decay works against you every single day. If the underlying does nothing, you lose money. Being right about direction but late is the same as being wrong.
• Volatility can crush a correct trade. A fall in implied volatility can leave you losing money even when the index moves your way — particularly after events.
• Spreads and liquidity. Away from at-the-money NIFTY strikes, spreads widen sharply. On single-stock options they can be wide enough to make a profitable strategy unprofitable.
• Our option figures are modelled. They use a theoretical pricing model, not traded quotes. Treat them as the optimistic end of the range.
• Selling options is categorically worse. Nothing on this page covers option writing, which carries theoretically unlimited loss. Do not extrapolate these figures to it.

• Execution discipline. A backtest never skips a signal, never moves a stop, never doubles down after a loss and never takes a day off. If you do any of those things, you are not trading this strategy and these numbers do not describe you.
• Position sizing errors destroy more accounts than bad strategies do. Risking too much per trade turns a normal losing streak into a permanent loss.
• Taxes. Intraday and F&O profits are taxable as business income in India, with separate audit and reporting requirements. None of the figures above are after tax.
• Regime change. Market structure changes. A rule set that worked for a decade can stop working, and there is no reliable way to know in advance that it has.
• Opportunity and psychological cost. Active intraday trading demands attention every session and produces sustained stress. Most people who attempt it stop within a year, and most stop at a loss.

• Algozone is an educational research tool that shows how a rules-based structure method has behaved historically.
• It is not investment advice, not a recommendation to buy or sell any security, and not a portfolio management or advisory service.
• Capital Finplus Academy is an educational platform. Please consult a SEBI-registered investment adviser before acting on anything you see here.
• Past performance is not indicative of future results. This applies to every figure on this page without exception.
• Trade only capital you can afford to lose in full. If a total loss of the amount you entered in the calculator would change your life, the amount is too large.