← All case studies
REALITY CHECK #3 · INTRADAY BREAKOUT · EDUCATIONAL POST-MORTEM

The opening range breakout — the internet's most-taught trade

Historical simulation on Indian index intraday data + a 14-year commodity sweep + live paper confirmation · windows through April 2026 (≥3-month lag) · published July 2026

If intraday trading content had a national anthem, it would be the ORB. Mark the high and low of the first 5, 15 or 30 minutes. Buy the break of the high, sell the break of the low. Ride the momentum. Every platform, every language, every year — the same diagram with the same two arrows.

It is also the strategy family we have tested most thoroughly. Nine variants across two indices, a fourteen-year commodity sweep, and — because we take falsification seriously — a live paper-trading deployment. Here is everything we found.

Round 1: the backtest looks fine (of course it does)

The 5-minute ORB posted a training-period Sharpe of 1.73. The 30-minute variant showed profitable years whenever the market trended. If we had stopped here — where most course-sellers stop — this would be a very different article.

Round 2: out-of-sample, everything dies

VariantInstrument scopeOut-of-sample result
5-min ORB (entry from 09:45)NIFTY−₹2.8L (after train Sharpe 1.73)
5-min ORBBank index−₹2.8L to −₹3.4L
30-min ORB (entry from 09:45)NIFTY−₹2.4L
30-min ORBBank index−₹2.5L to −₹2.7L
30-min ORB (entry from 10:00)NIFTY−₹2.2L to −₹3.6L
30-min ORB (entry from 10:00)Bank indexdeep negative

Full-sample forensics on the 30-minute NIFTY variant: over 1,200 trades, ~45% win rate, cumulative loss with losers consistently larger than winners. Walk-forward testing — retraining and re-testing era by era — confirmed it: −₹1.0L where the in-sample version had claimed a profit. "Trend-year dependent, negative full-sample" is the engine's verdict line.

Round 3: we let it trade paper money anyway

Because forward-testing is the only judge that can't be fooled, ORB variants ran in our live paper-trading fleet — real-time prices, realistic fills, no hindsight. They lost there too (five figures across both indices) and were retired. Backtest, walk-forward, live paper: three independent courts, same verdict.

Round 4: maybe it's an index thing? No.

We swept fourteen years of 1-minute data across gold, silver, copper, crude oil and natural gas — every opening-session breakout construction we could define. None survived. (For the curious: the only family that showed robustness in that entire sweep was the opposite trade — mean-reversion late in the session. Fading, not chasing.)

Why ORB fails — the mechanics. The opening minutes have the day's widest spreads and its most violent two-way noise. The "range" you're breaking out of is largely an auction finding its footing, and the breakout order pays the worst prices of the day at exactly the moment false breaks are most common. The strategy's premise puts you at maximum cost during maximum randomness — the arithmetic never recovers.

The honest summary

In-sample, ORB looks profitable. This is why it spread.

Out-of-sample, every variant we tested lost lakhs. Both indices, all entry timings.

Live paper trading confirmed the losses in real time. No hindsight involved.

Five commodities and fourteen years agree. It's not the market — it's the idea.

Trading an ORB variant right now?
Maybe yours is different — specific filter, specific market. Send the exact rules; we'll run it through all four rounds, free.
Disclaimer. This is an educational post-mortem of publicly known strategy concepts, based on historical simulation with data shown at a minimum three-month lag. It is not investment advice, not a research report, not a recommendation to trade or avoid any security, derivative or strategy, and not a solicitation. YCAI is not registered with SEBI in any capacity, charges nothing and sells nothing. Historical simulations — including negative ones — do not predict future results. Markets involve substantial risk of loss. Consult a SEBI-registered investment adviser before making investment decisions.