The "monthly income" iron condor
If you've spent any time in trading YouTube, you know this one. Sell an out-of-the-money call spread, sell an out-of-the-money put spread, collect the premium, let time decay do the work. "Consistent monthly income." "Works in sideways markets." "85% win rate."
We built it the way the videos describe it, and ran it over six years of index option history — tens of thousands of option chain snapshots, reconstructed at 1-minute resolution. The first result was spectacular.
Round 1: the naive backtest says you're rich
Rules exactly as popularly described: enter weekly condors at the market open, wings placed by delta, hold to expiry or a stop. The simulation — with fills taken at the bar's printed open price — produced a win rate near 85% and an annual profit in the lakhs on a single-lot-scaled book. The equity curve was a staircase to heaven. This is the chart the videos show you.
Round 2: one bug, and the staircase wobbles
Auditing the code before believing it, we found a subtle but classic backtesting bug: the entry logic was quietly using information from the same bar it entered on — a form of look-ahead that inflates every premium-selling result. Textbook stuff, embarrassingly easy to write, present in more public backtests than anyone admits. Fixing it removed roughly a third of the profit. Still profitable, though. Still a nice story.
Round 3: the fill-realism gate — and the story dies
Here is the correction almost no public backtest survives. The naive test assumed you sold four option legs at the printed opening price of the day's first bar. In reality: index options in the first minutes have wide, unstable spreads; a four-leg combo pays the spread four times; and by the time a real order is realistically working — call it half an hour after the open — the prices the backtest "sold" no longer exist.
We re-ran the identical strategy with one change: fills at prices achievable ~30 minutes after the open, paying realistic spreads per leg.
| Simulation version | Assumption | Outcome (same rules, same data) |
|---|---|---|
| Naive | Open-price fills, no spread | Strongly positive — the YouTube result |
| Bug-fixed | No look-ahead, open fills | Positive, ~⅓ smaller |
| Honest | Achievable fills + real spreads | Negative expectancy |
Not "smaller profit". Negative. The entire multi-lakh annual "income" was an artifact of two assumptions — one coding bug and one fantasy about execution. Every rupee of edge lived in prices no retail trader ever receives.
The receipts: three condor variants, every era negative
Beyond the flagship re-test, we ran the popular daily-entry variants through the full engine — out-of-sample windows, costs and achievable fills included. One-lot index options book:
| Variant (as taught online) | Out-of-sample result | Recent-era re-test |
|---|---|---|
| Daily short iron condor, 500-point wings | −₹66k to −₹83k | negative |
| Daily short iron condor, 300-point wings | −₹100k | negative |
| Daily short iron butterfly, 200-point wings | −₹119k | negative |
Three constructions, two market eras, one conclusion. The strategy family the internet calls "monthly income" has been, in honest simulation, a monthly expense.
Why this matters to you
Nobody selling a course shows you Round 3. It is not that iron condors are "bad" — a structure is just a structure — it's that the specific popular pitch (open-fill, mechanical, income-machine) has never been available to a real account paying real spreads. If someone shows you a backtest, your first three questions should be: where exactly are the fills? what do the spreads cost? did the code peek?
Send it in — selected ideas get the full treatment, free, results published here.