The Options Profit That Stale Prices Invented

An options backtest produces a profit. The strategy sells premium, waits for the position to lose value, and buys it back for less. The explanation sounds sensible: time passes, and time value decays.

But what if the apparent profit came from comparing prices that were never available together?

This installment of StonkHub Research Diaries follows the September 22, 2026 iron-condor research. Its most useful finding was a pricing defect that made an options strategy look better than the evidence justified.

The lead: collect premium within a range

An iron condor combines four option legs: a short put and call, with a protective option farther out on each side. It receives a credit when opened and seeks to benefit if the position can later be closed for less, or expires with a favorable payoff.

Time decay is part of that story, but it is not a promise of profit. Price moves, changes in implied volatility, and execution costs can outweigh it. The protective legs define the spread's payoff limits; they also need real prices.

The research examined same-day-expiry condors and longer-dated swing condors. The available historical window ran from February 3 through September 19, 2025, with shorter windows for some same-day experiments. The data consisted of trade-based aggregate bars, not synchronized bid and ask quotes.

The first attractive result had a hidden shortcut

The report records an early XSP swing run with nine simulated trades and a profit factor of 2.05. Profit factor compares total winning-trade profits with total losing-trade losses. Above one can look encouraging.

The problem was that the simulator could carry a protective wing's old trade price forward for up to 72 hours. Far-out-of-the-money options sometimes trade infrequently. An old print could therefore stand in for a current price while the other legs used more recent observations.

A four-leg position marked that way may describe prices that never coexisted. The report says the stale marks manufactured unusually large entry credits and apparently quick profit-target exits. The appealing result was not evidence of executable profitability.

Stricter marks changed the question

The revised approach required same-session observations with at least three trades for the legs. With stricter marks, the XSP swing run produced only three trades in approximately seven and a half months.

That did not prove the strategy rarely had opportunities. It showed that the available bars rarely supported pricing every required leg. A missing observation is a data limitation, not a market verdict.

Other defects mattered too. Last trades from different times could imply a negative vertical-spread value. A leg that stopped printing could prevent a losing position's stop from firing. The recorded fixes used session volume-weighted marks, payoff bounds, and an intrinsic-value floor for conservative stop handling.

These changes made the simulation harder to fool. They did not turn aggregate trade prices into executable quotes.

The profitable-looking alternative was still unresolved

The report also recorded a longer-dated SPX configuration with 16 simulated trades and approximately $3,823 net profit under its cost assumptions. That is the sort of result a blog could easily turn into a success story.

It did not meet the study's minimum of 30 trades. Its headline drawdown was 51.6% against the $5,000 starting-capital basis, above the study's 25% ceiling. Other tables use different capital bases, so those drawdown percentages cannot be mixed.

More fundamentally, session volume-weighted prices are not fills. The report notes that some entry and exit decisions used full-session marks extending beyond the nominal decision time. It also questions profit targets reached faster than time decay alone would explain.

We therefore describe this as an unresolved simulation result. The report's stronger language about an established trading edge goes beyond what these marks can support.

What we would do differently

The next test would start with timestamped quotes for every leg, a common decision time, and realistic execution assumptions. It would record missing or stale quotes explicitly and price exits as carefully as entries.

It would also separate the strategy's payoff from account suitability. A spread that looks attractive on paper can still require more buying power or tolerate more loss than the account allows. Historical research should document that constraint without turning it into a trade recommendation.

A longer evaluation window and prospective paper observations would follow. Those are proposed improvements, not completed evidence that repairs the historical result.

The conclusion

The first XSP profit was produced by unreliable marks. Better data is the next experiment; tuning the strategy around those prices would only make the fiction more elaborate.

This article discusses historical simulations, not investment advice or live trading results. Options can lose money. Read the financial disclaimer.

Explore the StonkHub blog or start this series with our Reddit attention experiment.