SPXPlay › Research

What 1,051 Historical SPX Scenarios Reveal About 0–3 DTE Vertical Spreads

VantureCap Research · Published July 21, 2026 · Report 01

A paired, descriptive study of four defined-risk SPX verticals priced on the same 1,051 historical setups. The headline is less comfortable than a win-rate pitch: the two credit structures won often, yet their occasional large losses pulled average one-contract P&L below zero in this scenario population.

Executive summary

• Credit spreads won frequently but had negative average P&L. Bull put credits were profitable in 81.4% of observations and bear call credits in 77.0%; their means were −$45.41 and −$65.44 per contract.
• The median and the mean told opposite stories. Median credit-spread results were positive, while mean results were negative. A small set of much larger losing outcomes outweighed many smaller credits.
• Debit spreads displayed the reverse payoff shape. They lost more often. The bull call debit spread was the only one of the four with a positive full-sample mean, +$17.03, despite a 43.5% win rate and a negative median.
• Regime labels changed the mix, but did not create a universal winner. Average results varied sharply across VIX and trend buckets, and several intuitive directional relationships appeared. Sample sizes also shrank quickly inside those buckets, especially for 0DTE and panic observations.
• Losses are not omitted. There were 1,729 losing trade observations out of 4,204, including 1,400 that lost at least 90% of their listed maximum loss. The four outcomes on a day are paired alternatives, so these are outcome counts, not 4,204 independent trading days.

Result 1: Win rate alone hid the payoff asymmetry

On a simple profitable-versus-losing count, the credit spreads dominated. The bull put credit finished above zero 856 times and below zero 195 times. The bear call credit finished above zero 809 times and below zero 242 times. Neither series had a zero-P&L observation after commission.

That frequency is real, but incomplete. A credit spread receives a relatively small premium up front and keeps it when the short strike is not breached at expiration. When the underlying travels through the width, the loss can be many times the initial credit. A high proportion of small profitable outcomes can therefore coexist with an unattractive average result.

Result 2: Average P&L reversed the win-rate ranking

The bull call debit spread had the lowest positive-outcome rate among the three higher-ranked structures, yet it produced the only positive full-population mean: +$17.03. Its median was −$811.55. That gap is the debit-spread version of skew: frequent premium losses with occasional gains large enough to lift the arithmetic mean.

The negative credit-spread means do not prove that “selling premium does not work.” They say something narrower: these particular, systematically selected, unmanaged one-contract spreads, held to expiration on this eligible historical population, did not convert their high win rates into positive average P&L after the documented commission.

The complete outcome table

One contract per outcome. P&L includes a $1.30 per-trade commission and assumes no early management. “≥90% max loss” is a reporting threshold, not a separate exchange outcome.

Result 3: The losing outcomes were large enough to matter

The average losing bull put credit outcome was −$1,825.95, versus a positive median across all bull put outcomes of +$277.45. The corresponding bear call figures were −$1,622.92 and +$256.20. This is why replacing a P&L distribution with one win-rate number removes the part of the result that controls risk.

Debit spreads reached the near/full-loss threshold much more frequently: 47.1% for bull calls and 58.6% for bear puts. That is consistent with purchased verticals expiring below breakeven often enough to lose most or all of the debit. Their attraction is not loss avoidance; it is the possibility of a larger payoff when direction and distance are right.

Volatility-regime comparison

The report uses the trainer’s entry-time VIX bucket, computed from the prior close. The table shows average one-contract P&L, not a forecast and not an instruction to trade a bucket. The same four structures are evaluated within every scenario in each row.

Three contrasts stand out. First, the bull put credit’s loss frequency rose from 10.3% in calm observations to 31.7% in stressed observations. Second, bear put debit spreads had a positive mean in the stressed bucket (+$210.06) but not in the panic bucket (−$231.29). Third, the panic bucket produced a positive bull call mean (+$248.91) while its bear call credit mean was −$245.96. These are descriptive interactions in uneven samples, not proof that a regime label alone supplies an edge.

Trend-regime comparison

Trend labels use only information available by entry: recent returns, moving averages, ATR, and the entry level. Directional alignment appeared in several averages — bull structures in strong-up observations and bear puts in down observations — but it was not monotonic.

For example, bear put debit spreads averaged +$110.55 in the 168 down-trend observations, yet −$23.43 in the 136 strong-down observations. Strong-down setups were also where the bear call credit posted its highest near/full-loss frequency, 27.9%. A label describes the tape coming into entry; it does not guarantee continuation during a very short hold.

DTE composition and comparison

The 41 same-day observations are too few to carry the same descriptive weight as the 461 two-DTE observations. The builder selects a common qualifying expiration nearest the midpoint of the 0–3 DTE window, which is why 2 DTE is the largest group. This is a property of the scenario construction, not the natural market-wide frequency of each DTE.

How the study was constructed

1. Start with eligible SPX sessions between August 2021 and June 2026 that have sufficient pre-entry SPX and VIX history.
2. Choose one 0–3 DTE expiration that can price all four vertical types, preferring the expiration nearest 1.5 DTE.
3. Select the credit-spread short strike near 1% out of the money and target a width near 0.5% of spot. Select the debit-spread long strike near at the money and target a width near 1% of spot.
4. Require positive max-profit/max-loss values, keep the risk cap at $5,000 per contract, and require credit shorts to begin out of the money.
5. Hold every selected spread to expiration. Compute one-contract settlement P&L with the 100× multiplier and subtract $1.30 commission. No stops, rolls, early exits, slippage model, or discretionary filtering are added.
6. Aggregate only the resulting outcomes. The report does not expose historical quote rows, individual entry dates, or licensed price series.

See the full methodology and metric definitions for the inclusion rules, regime definitions, reproducibility contract, and data publication policy.

Limitations

• Constructed choice set: one rule-selected spread represents each structure. Other deltas, widths, expirations, and fills can behave differently.
• Simplified execution: historical quoted premiums and settlement labels are used with a flat commission. The study does not model bid/ask slippage, legging risk, taxes, margin changes, or market impact.
• Held to expiration: active management, stops, profit targets, rolling, and closing before settlement are outside scope.
• Selection effects: a day enters only when all four structures share a valid qualifying expiration and the underlying/VIX history is complete.
• Paired observations: 4,204 outcomes come from 1,051 days. The four alternatives on the same day are dependent, and adjacent market days are also serially related. No p-values or confidence claims are made.
• Regime sample imbalance: VIX buckets range from 90 panic scenarios to 462 elevated scenarios; DTE groups range from 41 at 0DTE to 461 at 2DTE.
• Historical and in-sample: the analysis describes this fixed period and this builder. It is not an out-of-sample forecast, and future market structure may differ.
• Calendar caveat: some late-2025 macro release dates reflect eventual shutdown-related delays that were not all known at the earlier entry date. Those event annotations do not alter settlement P&L, but they limit event-context interpretation.

Reproducibility and publication policy

The report is generated from the checked-in SPX trainer bundle using a deterministic aggregate script. Build-time validation requires exactly four vertical outcomes on one common 0–3 DTE expiration for every scenario. The page publishes counts, rates, percentiles, and grouped P&L summaries only. It does not publish licensed raw option chains, raw OHLC series, individual scenario dates, or a downloadable row-level dataset. Viewing, citation, and reuse are governed by the SPXPlay Research Aggregate Data License.

Research note: amounts are nominal U.S. dollars per one-contract outcome and are not normalized by margin, max loss, or capital. A positive historical mean is not a recommendation; a negative mean does not prove that every implementation of the structure is unprofitable.

Playbook Trainer is an educational game built on historical market data. Nothing on this page is investment advice or a recommendation to trade. Options involve substantial risk; defined-risk spreads can lose their full maximum loss. Scenario dates are masked, and prices reflect historical option quotes with simplified fills.