Free Options Backtester — Replay Strategies on SPY and SPX History

What this backtester does, and what it does not do

It takes a rule set — strategy, delta or strike, days to expiration, profit target, stop loss — and walks it forward one day at a time through the history you picked. Every entry the rules would have triggered becomes a simulated trade, aggregated into win rate, annualized return, drawdown and a trade log.

Use it to ask whether a 30-delta covered call held to expiration beat a 50% profit target, or what a cash-secured put program did through 2022. It does not model share ownership — the covered call backtest books the short call only — has no intraday simulation, and cannot place an order.

The price history ends on March 9, 2026. SPY and SPX bars are stored up to that date, so a backtest that asks for a window running to today is actually run through that session. Pick an end date at or before it to see the full range you asked for, and check the date on the first and last trade in the log before you quote a result.

How the data is built

Underlying prices are real: daily OHLC and volume for SPY and the SPX index, plus VIX closes and Cboe put/call readings. SPY first traded on the American Stock Exchange on January 29, 1993, the natural floor for an ETF test; the SPX series reaches back to 1950, but VIX closes start January 2, 1990.

Option prices are not real. Each day the engine builds a chain with a Black-Scholes model that takes that day's VIX close as its volatility input, applies a fixed 4.5% risk-free rate, bends implied volatility upward below spot for the skew, and adds a modelled bid/ask spread. VIX is itself derived from SPX option prices, so it is a reasonable input — but it is one number for one tenor, not the surface a market maker would quote.

Three consequences. Direction is grounded in real prices while option marks are theoretical: a SPY close is a SPY close, but a $3.90 credit on a 30-delta call will differ from the NBBO you could have hit. Daily bars cannot see a volatility spike, a gap through your stop, or the difference between selling at the open and at the close. And early assignment is not modelled — there is no dividend calendar or ex-dividend exercise logic, exactly the window when an in-the-money short call is most often assigned. SPX compounds this: SPX options are cash-settled and European-style, while SPY options are American-style and settle in shares.

Covered calls vs cash-secured puts vs the wheel, on the same underlying

All three sell premium on a stock you are willing to own and are marketed the same way: steady income, high win rate, limited risk. Run them on the same underlying and the differences show up in the tails. The table is one illustrative run: SPY, January 3, 2022 to December 31, 2024, 21–45 days to expiration at entry, 30 delta, 50% profit target, otherwise held to expiration, entries skipped on high-impact event days, $100,000 capital, one position, 100-share multiplier, modelled prices.

Illustrative only: the short option leg's profit and loss against a $100,000 capital base with one contract open at a time, on modelled prices. Not a portfolio return; excludes the share leg, dividends, commissions and assignment; not a forecast.

The cash-secured put won more often (86.6% vs 84.4%) and still made less money, because its average loss was 37% larger. Both are the same trade in different clothes — a short put and a covered call have identical expiration payoffs, which is why the Options Industry Council calls the cash-secured put's risk profile identical to the covered call's. The covered call also carries a cost the model cannot see (owning the stock gives away upside above your strike) and a benefit it cannot see (the dividend). The wheel chains the two: sell a put, take assignment, then sell covered calls until the shares are called away — which this tool backtests as two separate passes, neither of which books the share leg.

Reading the results: annualized return, max drawdown, win rate

Annualized return (CAGR) compounds total profit and loss over the window, so it is only meaningful next to the capital base and concurrency you set: one contract on $100,000 leaves most of the account idle and looks tiny, while three positions on $30,000 each look far better and quietly triple correlation risk.

Max drawdown is measured on realized profit and loss against the running equity peak — not your account statement, because open positions are not marked to market. A short put $4 out of the money with three days left shows zero drawdown here and feels different live.

Win rate hides the most. Premium selling wins by construction, so 75–90% is normal and says little. What matters is average win against average loss: above, the covered call's average loss was 2.6 times its average win, so eight wins funded about three losses. That is why the results report profit factor and expectancy — compare those between runs, not win rates. The Sharpe ratio shown uses per-trade P&L, not daily account returns, so rank runs of this tool against each other with it, never against a fund fact sheet.

Where backtests mislead

Overfitting. Delta, DTE window, profit target, stop level, VIX filter, event filter — six dimensions, and the best cell looks excellent whether or not an edge exists. 64 trades cannot separate a five-point win-rate gap from luck.

Survivorship. SPY and SPX did not go to zero, were not acquired, and have the most liquid options anywhere, so a strategy that works here was tested on the friendliest possible case. Single-name options can gap through a strike on earnings.

Dividends and assignment. Because the share leg is not booked, SPY's quarterly dividends never enter these results, so a covered call's true return is understated by roughly the yield. The mirror risk is missing too: an in-the-money short call is most likely to be exercised the day before an ex-dividend date. FINRA calls this dividend risk, and it is invisible in any backtest that treats assignment as an expiration event.

Transaction costs. Fills happen at the modelled mid. Real premium selling crosses the spread twice, and a 50% profit target means paying it on a position that only had half its credit left. Add commissions and fees and 64 trades at $352 average wins lose real edge.

Path dependence. Real accounts stack correlated positions, and a test that enters a fresh put every week through a decline shows a smooth curve nobody could have held. 2022–2024 also holds one bear market and two strong up years; a window ending earlier differs.

Explorer coverage. The explorer loads up to 1,000 daily bars per request, so the ten-year and maximum presets show a bounded window rather than every session since 1993. Use the strategy builder for long runs.

How this differs from a screener, a paper-trading account and a broker simulator

A screener filters the universe as it is right now and stops there, with no history or outcome. Paper trading places simulated orders in the live market and makes you wait: good for mechanics, but you learn one realization of the future. A broker's simulator mimics its platform's tickets and fills — right for learning the buttons, wrong for measuring a rule set.

This backtester replays a rule set across years of daily history and hands you the whole distribution at once: every trade, the equity curve, the monthly returns. Best for stress-testing rules across regimes, worst for learning a platform. For the strategy tour rather than the statistics, see the options simulator overview; to price one trade before you place it, use the covered call calculator or cash-secured put calculator.

Frequently asked questions

Related tools and guides

Backtesting reading

• How to use the free OptionsPilot backtester — a walkthrough of the builder, the filters and the results tabs.
• Does options backtesting actually work? — the limitations, in more depth than the section above.
• Covered call vs cash-secured put, backtested — the same comparison with different windows and strikes.
• Best free options backtesting tools (2026) — how this tool compares with the alternatives.
• What counts as a good win rate — why 85% is normal for premium sellers and still not enough.
• Best delta for covered calls — the delta choice this backtester defaults to, and the trade-off behind it.
• Short puts: payoff, margin and naked vs cash-secured — what the CSP leg costs you in margin.

Run your first backtest

Pick a strategy, set the rules, and read the whole distribution — free, no account, no risk.

Sources

• U.S. Securities and Exchange Commission, "SPY: The Idea That Spawned an Industry" — SPY first traded on the AMEX on January 29, 1993.
• FINRA, Options — American vs European exercise, cash vs share settlement, standard 100-share contract, dividend risk and assignment.
• OCC / Options Industry Council, Covered Call (Buy/Write) — maximum gain and maximum loss for a covered call.
• OCC / Options Industry Council, Cash-Secured Put — payoff, breakeven, and the note that its risk profile is identical to the covered call's.
• Cboe, Cboe Volatility Index — VIX is derived from SPX option prices and was introduced in 1993.
• Cboe, S&P 500 Index Options Product Specifications — SPX multiplier of 100, settlement and trading hours.
• FINRA, Rule 4210 (Margin Requirements) — short option margin percentages and the treatment of covered calls.
• Internal Revenue Service, Publication 550 — tax treatment of listed options, including the section 1256 60/40 rule for broad-based index options.
• OptionsPilot's own backtest engine source (src/lib/backtester/) — the pricing model, risk-free rate, skew, synthetic spread and the exclusions described on this page.

Options involve risk and are not suitable for every investor. The figures on this page are illustrative, generated by a model on historical data, and are not a promise or projection of future results. Nothing here is investment, tax or legal advice. Read the Characteristics and Risks of Standardized Options before trading.