0DTE CFD options vs US options: we measured the real cost of every trade

Every options trade pays a hidden fee: the gap between the price you can buy at and the price you can sell at — the bid/ask spread. On same-day (0DTE) options you pay it on every single trade, and it's often the difference between a strategy that makes money and one that doesn't. So we measured it, precisely: every trading day for three years, from far out-of-the-money to deep in-the-money strikes, on the two venues where a European trader can deal S&P 500 daily options — a CFD dealer and a US exchange. This article is the fee schedule nobody publishes, including one result we didn't expect.

Two shops selling the same option

The same S&P 500 daily option — same strike, same 4pm New York settlement — is sold in two places. On a US exchange, where market makers compete in an order book. And at a CFD dealer, where one firm quotes you its own bid and ask, take it or leave it, nearly 24 hours a day. Every trader "knows" the exchange is cheap and the CFD dealer is expensive. We wanted the actual numbers — and the actual numbers turned out to be more interesting than the folklore.

The cost below is what one round trip across the spread took, in dollars, for the same position size on both venues (a $100-per-point exposure — one exchange contract, or the equivalent CFD stake). Measured at 10:00 New York, six hours before settlement, roughly 730 trading days per strike per venue.

What the bill says

For the strikes most people trade — out of the money and at the money — the exchange is much cheaper. It charges $10–$20 where the CFD dealer charges $70–$100: five to seven times less, on the median day. And on both venues the spread scales with the option's value — cheaper options, tighter spreads.

Deep in the money, the story flips. This is the part nobody tells you. At 90-delta — options trading around 100+ points — the CFD dealer's spread actually falls to ~$61, its cheapest price anywhere on the curve. The exchange's spread explodes: a $100 median, a 90th percentile of $566, and single prints far beyond. Across our three years of data, deep in the money, the CFD dealer was usually the cheaper venue. We won't claim to prove the mechanism from price data alone — most trading interest sits in out-of-the-money options, which may leave exchange market makers little reason to quote the ITM side tightly, while the CFD dealer prices the whole chain from one engine — but whatever the cause, the pattern held throughout.

The day-to-day risk is different in kind, not just size. Look at the bands, not just the lines. The CFD dealer's band is a narrow ribbon: on 9 days out of 10 you pay within ~$20 of the median, and even its worst 1% of days stays around $200. The exchange's band is a funnel: beautiful on a calm Tuesday, then $439 at the money in its worst 1% — the stressed mornings when a premium seller most needs to trade. One venue sells you a price; the other sells you a price range.

So — should we discard trading with CFD options?

If cost were the only thing that mattered, this article would end with "open a US brokerage account." It doesn't, for five reasons.

1. The markup buys certainty. The CFD dealer's price is known before you trade, essentially the same tomorrow as today, and it was still there — bounded — through the August 2024 volatility spike and the 2025 tariff panic. The exchange's discount is a market price: it was 5× cheaper most days, yet at the money it exceeded the CFD dealer's price on about 8% of days, clustered exactly in the periods when short-premium traders were forced to act.

2. The clock. CFD daily options trade nearly around the clock; the exchange opens at 9:30 New York. In our entry-time study, some of the best risk-adjusted entries were at 11:30pm and 4:30am New York. Those trades don't exist on the exchange at any price. Part of the CFD dealer's spread is rent for the other two-thirds of the day.

3. Access, margin and tax. No US broker account, no currency conversion, no pattern-day-trader rule, and position sizes small enough to trade a strategy rather than bet the account. Margin requirements on CFD daily options are typically far lower than on US-listed options, so the same strategy ties up less capital — which changes the return on capital, not just the cost per trade. And for UK residents, spread-bet profits are currently tax-free (as ever: your circumstances, and rules change). For many European traders the realistic comparison isn't "dealer vs exchange", it's "dealer vs not trading 0DTE at all".

4. Sometimes the CFD dealer is simply cheaper. The deep-ITM inversion above isn't a curiosity — certain positions (an in-the-money call instead of its out-of-the-money mirror, for instance) genuinely cost less to put on at the CFD dealer. Knowing the cost curve means you can choose the strike where the venue works for you.

5. Real strategies survive the markup. In our fill-price experiment, a plain short strangle on these same CFD options still earned a Sharpe ratio of 1.46 over two years after paying every spread in this article. The spread doesn't forbid an edge; it sets the hurdle the edge must clear. What kills accounts isn't the $90 — it's not knowing about the $90 until the live results come in under the backtest.

So no — keep CFD options. Just never trade them without knowing what they cost. Backtest your strategy on the real quoted prices first: if the edge survives the spread in the backtest, it can survive it live. If it doesn't, you found out for free.