SPY-T Trading SPY-T Trading

Edge Lab: the Overnight Simulator

What would overnight trading actually do to your account? Every simulated night the market moves by a gap it really printed β€” fat tails included, no bell curves. A win means you called the direction, long or short. Drag the dial from a coin flip upward and watch what calling it right is worth.

Data
Position & vehicle
The dial

How often do you call the gap’s direction right? 55%

Loading gap history…
Left end (50%) = flipping a coin on direction. Everything to the right is a hypothetical: what if something helped you call it? Switch vehicles and watch the breakeven move β€” short-dated options pay more per correct call but charge decay every night. The dial runs to 85% so you can stress the upside β€” sustained real-world accuracy that high is extraordinary.

How to read it β€” what’s real here. 400 paths of Monte Carlo, but not the lazy kind: nights are drawn from this ticker’s real gap history in contiguous runs (stationary block bootstrap, ~8-night average blocks), so volatility clusters the way it really does β€” calm months and violent weeks, not evenly sprinkled randomness. On shock nights beyond the 90th-percentile move, your dialed edge fades toward a coin flip, because genuine surprises are the least predictable nights for any signal. Option premiums breathe: each night’s contract is priced at the IV implied by its surrounding stretch of history, so options cost the most exactly when moves are biggest, and spreads widen with volatility. You win each night’s move β€” long or short β€” at your effective win rate, and lose it otherwise. Option trades are the real overnight round trip: bought near the 4PM close and sold at the next open at exact Black-Scholes mark-to-market β€” never held to expiration, so remaining time value is kept. The ~17.5-hour hold is charged 0.75 of a day of decay (slightly conservative: market makers pre-decay part of the overnight before the close). IV and strikes are solved from each ticker’s own volatility at load, so QQQ options price hotter than SPY’s. Delta, gamma and decay all hit the P&L β€” the cheaper and shorter-dated the option, the higher the leverage and the heavier the decay. Overnight IV changes at the open (vega) are not modeled. Wrong-direction calls are the mirrored move (put-call symmetry), and losses are capped at the premium. Leveraged ETFs apply their multiple to a single overnight move (no multi-day reset drag). Friction per trade: 0.05% for shares/ETFs; for options, a fixed spread cost (β‰ˆ$0.06 round trip on real SPY quotes) β€” which means the cheaper the contract, the bigger the bite as a percentage of premium.

What’s still simplified β€” read this part too. IV changes at the open (vega) aren’t modeled β€” roughly offset by charging slightly generous overnight decay. Your win rate is constant across time; real signals have hot and cold regimes. No earnings calendar or day-of-week structure. Fills are assumed near the mid of the opening auction; fast opens can be worse. All strikes share one IV β€” the real vol smile makes short-dated OTM wings trade even richer than modeled, so cheap-delta results here are, if anything, generous. Taxes, assignment, and margin calls are ignored. And the deepest one: the future can print a night history never has.

This is a hypothetical. The win-rate dial is a what-if you control β€” it is not a claim about any signal’s performance, including ours. Past gaps don’t predict future gaps. Simulated results are illustrative, not investment advice. Options and leveraged products can lose their entire value.