# Why 0DTE Backtests Fail: Slippage and Bid-Ask Drag

Source: https://cluenex.com/learn/technical-analysis/backtested-0dte-options-strategies-fail-live/
Publisher: Cluenex (https://cluenex.com)
Updated: 2026-11-03

## Short answer

Backtested zero-day-to-expiration (0DTE) strategies fail in live trading because theoretical models assume mid-market fills and smooth stop-loss exits. In reality, crossing wide bid-ask spreads on multi-leg trades and suffering slippage during afternoon gamma spikes routinely turns a backtested 75% win rate into a net cash loss.

## Key takeaways

- Mid-price fill assumptions in backtests hide a 5% to 15% execution penalty on premium received.
- Extreme afternoon gamma causes option prices to double in seconds, triggering stop-loss slippage of $0.50 to $2.00 per contract.
- Low-resolution 5-minute or daily data bars mask intra-minute price spikes that breach stops in live accounts.
- Multi-leg setups like iron condors compound spread friction across four separate option legs upon entry and exit.
- Realistic backtests must deduct full bid-ask widths, broker transaction fees, and conservative adverse fill estimates.

## Why do paper 0DTE backtests look so profitable?

Paper backtests of zero-day-to-expiration (0DTE) strategies show glowing results because they assume frictionless trading at the exact midpoint of quoted prices. When a software simulator assumes you sell an S&P 500 index (SPX) credit spread at the precise midpoint and buy it back at the midpoint, it eliminates the single largest expense in ultra-short-term trading: market friction.

Between 2022 and 2026, daily expiring options on the SPX index grew to account for over 50% of total SPX options volume, according to market data from Cboe Global Markets. This surge fueled automated backtesting engines that promised annualized returns topping 40% with win rates near 80% on delta-neutral iron condors and credit spreads. To understand how these strategies compare over longer holding periods, review whether [0DTE options beat weekly options for selling premium](/learn/technical-analysis/0dte-options-better-than-weekly-options/).

These simulations suffer from lookback bias and frictionless execution. A standard backtest assumes that an order placed at 9:35 AM ET fills instantly. It also assumes that if a trader sets a 200% stop loss, the order executes cleanly at exactly three times the collected credit. In real live trading with actual market makers, neither of those assumptions holds true.

## How does the bid-ask spread create an invisible drag?

Bid-ask spread drag is the difference between what a contract is nominally worth and what a market maker will pay you to trade it. For SPX 0DTE contracts, the bid-ask spread rarely sits tighter than $0.10, and it frequently widens to $0.30 or $0.50 per contract, even during normal morning trading hours.

Consider a trader testing a 10-point wide SPX vertical credit spread: selling a short strike and buying a protective long strike. On paper, the model calculates the midpoint of the short option at $2.50 and the midpoint of the long option at $1.50, generating a theoretical net credit of $1.00, or $100 per spread. In real life, market makers quote the short strike at $2.40 bid and $2.60 ask, and the long strike at $1.40 bid and $1.60 ask. To receive an immediate fill, the trader must sell the short leg at $2.40 and buy the long leg at $1.60, collecting just $0.80.

| Execution Metric | Paper Backtest Assumption | Live Market Reality | Dollar Impact (per 10-contract order) |
| :--- | :--- | :--- | :--- |
| **Entry Net Credit** | $1.00 mid-price fill | $0.80 natural fill | -$200 immediate drag |
| **Exit at 50% Profit** | Buy to close at $0.50 | Buy to close at $0.60 ask | -$100 additional friction |
| **Round-Trip Commission and Exchange Fees** | $0.00 | $2.60 to $5.20 total | -$26 to -$52 fees |
| **Net Realized Profit** | $500 theoretical gain | $148 actual gain | **69.6% theoretical gain erased** |

When a strategy trades once or twice per day, repeating this round-trip drag every session turns transaction friction into an insurmountable hurdle. Even on days when the underlying index finishes within the expected range, bid-ask friction siphons off more than half of the expected edge. Reviewing your overall [risk-reward ratio](/learn/portfolio-management/risk-reward-ratio/) shows how quickly fixed friction undermines a thin statistical edge.

## What does extreme gamma do to stop-loss orders?

Gamma measures the rate of change of an option's delta for every one-point move in the underlying asset. On the day an option expires, gamma spikes to extreme levels because the time value (theta) of the contract approaches zero, leaving only pure moneyness risk.

During the final two hours of the trading session (2:00 PM to 4:00 PM ET), an out-of-the-money contract with a delta of 0.10 can surge to a delta of 0.70 in less than five minutes if the S&P 500 drops 15 points. In an automated backtest, a computer detects that a stop-loss threshold was crossed at $3.00 and logs an exit at $3.00. In reality, market orders triggered by automated stops hit the book just as the quote spread blows out from $0.20 wide to $1.50 or $2.00 wide.

Because the underlying index is moving rapidly, liquidity providers pull back their limit orders to protect themselves from adverse selection. A live stop-loss order placed to buy back a spread at $3.00 often fills at $3.80, $4.50, or worse. A single stop-loss event with $1.50 of adverse slippage erases the profits of three to five successful winning sessions. Cluenex provides an interactive options guide to [shortlist strategies based on time horizons and risk limits](https://cluenex.com/guides/pick-a-strategy/), helping investors avoid structures vulnerable to sudden gamma decay.

## Why do minute-bar backtests hide intraday stop triggers?

Most commercial backtesters evaluate options strategies using 1-minute, 5-minute, or end-of-day bar data. These aggregation windows conceal high-frequency price paths that trigger real-world risk management rules.

Imagine a 0DTE put credit spread with an automated stop-loss rule set at twice the premium received. On a 5-minute chart, the opening price is $1.20, the high is $2.10, the low is $0.90, and the close is $1.10. The backtester checks the closing bar, observes $1.10, notes that the stop level of $2.40 was never crossed, and marks the position as healthy. 

However, in tick-by-tick live trading, a rapid order imbalance might have spiked the ask quote to $2.60 for four seconds at minute 2 before snapping back. In a live trading account, your broker's server reads $2.60, fires a market order, stops you out at a steep loss, and leaves you watching from the sidelines while the theoretical model claims the trade ended as a full profit. The rules governing order routing and [stop-loss execution methods](/learn/portfolio-management/stop-loss-methods/) explain why software stops behave differently from live conditional orders.

## When can backtesting 0DTE strategies still provide value?

Backtesting 0DTE options is not useless, but it requires conservative parameters that reflect real exchange execution rather than mathematical ideals. If you want backtest metrics that survive live execution, apply these three modeling rules:

1. **Model fills at the bid and ask, never the midpoint.** Force the backtest engine to sell at the posted bid and buy at the posted ask. If the backtest cannot turn a profit under full bid-ask spread costs, the strategy has no live edge.
2. **Include an adverse slippage penalty on exits.** Add an automatic $0.20 to $0.50 penalty per contract to every stop-loss execution to simulate afternoon gamma spikes and exchange liquidity evaporation.
3. **Test with tick data or 1-second bars.** Avoid 15-minute or 5-minute bars for same-day expiries. Only high-resolution data captures the violent wicks that trigger automated stop orders.

Strategies that rely on collecting tiny fractions of premium, such as 5-delta iron condors collecting $0.40 on a 5-point wide wing, rarely survive these adjustments. Viable strategies generally feature wider strike widths, deliberate sizing, and defined risk that does not depend on millisecond-level stops.

## What this means for you

If you manage your own portfolio and trade options, treat any 0DTE backtest claiming a win rate over 75% or smooth exponential equity curves with skepticism. Theoretical backtests represent the absolute best-case scenario in a clean laboratory environment, while live options trading occurs in a noisy, costly auction.

Before allocating real capital to an intraday quantitative options model, paper trade it live for at least 30 trading days during market hours. Compare your live execution prices against the theoretical midpoint fills generated by your backtesting tool. Keep a trade log tracking slippage on every fill. If the difference between your paper model and your real broker fills exceeds 10% of your gross premium, the strategy lacks the margin of safety required to survive in live trading. You can also inspect terms in the [Cluenex options and equity glossary](https://cluenex.com/glossary/) to ensure you understand every metric on your trading screen.

Finally, control your total exposure. Because zero-day options can lose 100% of their value in minutes, sizing each trade appropriately is essential. Review standard rules for [position sizing and account risk allocation](/learn/portfolio-management/position-sizing/) so that a string of live slippage events cannot impair your long-term capital.

## Questions people ask

### Why is slippage so high on 0DTE options?

Slippage is elevated because 0DTE options carry extreme gamma, meaning option deltas and prices swing violently within seconds. When price moves trigger market stops, market makers widen their spreads to avoid adverse selection, forcing retail stop orders to fill far below or above the previous trade.

### Can you use limit orders to avoid 0DTE slippage?

Limit orders prevent bad fills by capping the price you are willing to pay or receive, but they carry non-execution risk. If an underlying index moves rapidly past your limit price during an adverse move, your stop-limit order will not execute, leaving you exposed to a catastrophic maximum loss.

### What is the typical bid-ask spread on SPX 0DTE options?

Near-the-money SPX 0DTE contracts typically feature bid-ask spreads of $0.10 to $0.30 during calm morning hours, but spreads routinely widen to $0.50 to $2.00 during fast market selloffs or within the final hour before the 4:00 PM ET close.

### Why do backtesters assume midpoint fills?

Most retail backtesters use midpoint fills to simplify calculations and show smooth historical equity curves. However, this assumption introduces lookback bias by ignoring the market maker's edge and the real costs required to cross the spread in live trading.

## Sources

- [Retail Investors in Options: Trading Behavior and Performance](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4404704)
- [FINRA Investor Insights: Zero-Day-to-Expiration Options](https://www.finra.org/investors/insights)

Cite as: Cluenex. "Why 0DTE Backtests Fail: Slippage and Bid-Ask Drag." Cluenex Learn, updated Nov 3, 2026. https://cluenex.com/learn/technical-analysis/backtested-0dte-options-strategies-fail-live/