Does buying options before earnings pay? Across 459 US reports, a call lost 42.8% on average at real prices

How this study happened

Before an earnings report, a stock's options already put a number on the move: a call and a put at the share price cost, together, about as much as the move the market expects. I wanted to know what someone who bought a call before the report actually got back.

I wrote the rules down first, then ran them on every US report in my own record of option prices, from late June to early October 2026.

Wojciech, 12 October 2026

The findings at a glance

−42.8%

average per trade: a call with its strike as far above the share price as the move its options priced in, bought before the report and sold at the next close, at real prices, 459 US earnings reports. Margin of error −66.6% to −19.1%

−9.0%

the same trades at the midpoint price, halfway between bid and ask: too little to tell apart from no loss at all. Most of the loss is the spread

17%

of the 459 calls made money. The median trade: −90%

34%

of the calls found no buyer at the close after the report: everything paid was lost

Claims and verdicts

Three claims and what the data says

Every claim from every study, in one list →

Claim tested	Finding	Verdict
Buying a call before a US earnings report and selling it at the next close pays	−42.8% on average at real prices, 459 reports; margin of error −66.6% to −19.1%. 17% of the calls made money	Does not hold
Selling the call in the morning after the report beats holding it to the close	The main rule (sell when the bid reaches 1.5 times the price paid): −0.7 percentage points (pp) against holding, margin of error −10.3 to +9.0 pp. The other two rules: much the same	No improvement
Calls on companies that usually beat analysts' profit estimates do better	+14.7 pp against the other companies, margin of error −50.7 to +80.1 pp: no difference these trades can show either way	Can't tell yet

1 · The main test

At real prices the call lost 42.8% on average; at the midpoint, 9.0%, a result luck alone could produce.

First, the main test. For each report I worked out the move the stock's own options priced in: what a call and a put at the share price cost together, as a share of the share price. The median was 9.1%. Then I took the call with its strike that far above the share price, on the nearest expiry after the report, bought it at the ask at the last close before the report and sold it at the bid at the first close after it. The ask is the price sellers want; the bid is the price buyers offer.

Stocks that report on the same evening move together, so I average each evening first; a report released the next morning before the open counts with the evening before. 29 evenings had at least three reports, 430 of the 459 reports, and each evening counts the same. Across them the call lost 42.8% of what was paid, on average, with a margin of error from −66.6% to −19.1%, the range luck alone could move the average within. That's a loss even at the ends of that range.

Priced at the midpoint instead, halfway between the bid and the ask, the same trades lost 9.0%. So most of the loss, about 34 percentage points of it, is the spread. A buyer can't count on the midpoint, though: you pay the ask and get the bid unless someone takes your limit price.

Average return per trade on the call: at real prices −42.8% (margin of error −66.6% to −19.1%), at the midpoint price −9.0% (−42.2% to +24.3%). They are the same trades: the gap between them, about 34 percentage points, is what the bid/ask spread cost. — WB Trading Labs · Fig. 1A call with its strike as far above the share price as the move the stock's options priced in. Real prices: bought at the ask at the last close before the report, sold at the bid at the first close after it. Midpoint: halfway between bid and ask, −9.0% with a margin of error from −42.2% to +24.3%. Average of the 29 evenings with at least three reports (430 of the 459 reports), each evening with the same weight; thin line: margin of error (95%). US earnings reports from 24 June to 9 October 2026. Data: my own record of US option prices, my calculations.

The average also hides how lopsided a bought call is: the median trade lost 90%. Fig. 2 shows how all 459 calls ended; a call with no bid at all at the close after the report counts as a total loss.

100 squares, each 1% of the 459 calls: 34% (157 calls) found no buyer at the close after the report, a total loss; 49% (224) lost part of the price paid; 17% (76) made money. 2 calls broke even exactly. — WB Trading Labs · Fig. 2All 459 calls of the main test, bought at the ask at the close before the report and sold at the bid at the first close after it. No buyer: no bid at all, counted as −100%. US earnings reports from 24 June to 9 October 2026. Data: my own record of US option prices, my calculations.

2 · Other strikes

Every strike distance lost on average, and the further out, the more.

These are side results: I wrote them down to see the whole curve, and none of them tests anything on its own.

With a loss that size, I wanted to know whether the strike itself was the problem. So I tried other distances on the same reports: the strike at the share price, half the priced move, the whole move, one and a half times it and twice it. And a call with a delta of 0.25, one whose price moves about a quarter of a dollar for each dollar the stock moves; it's a common choice for a call with its strike above the share price.

Every one of these calls lost on average, and the further the strike from the share price, the larger the loss: from −32.8% with the strike at the share price to −64.7% at twice the priced move.

I tried the other side too. A put with its strike the same distance below the share price lost 27.2% on average, but its margin of error reaches +2.0%, so by the rules I wrote down first, these reports can't settle the put either way. A call and a put bought together, which gain on a big move in either direction, lost 41.1%.

Average return per trade at real prices by strike: at the share price −32.8% (margin of error −53.2% to −12.3%), half the priced move −37.9% (−56.6% to −19.1%), the priced move −42.8% (−66.6% to −19.1%), 1.5 times the move −60.3% (−82.2% to −38.4%), twice the move −64.7% (−91.8% to −37.6%). At the midpoint: −17.2%, −17.8%, −9.0%, −30.0% and +2.4%. — WB Trading Labs · Fig. 3Strike above the share price, in multiples of the move priced in by the stock's options. Bought at the ask at the close before the report, sold at the bid at the first close after it; average of the evenings with at least three reports. 27% of the calls made money with the strike at the share price, 8% at twice the move. The 0.25-delta call: −43.4%, close to the main test. Each row leaves out the reports with no listed strike close enough to its target, so the rows cover different reports (358 to 465); the row at the share price leaves out stocks whose listed strikes are far apart, so it leans to the most-traded ones. US earnings reports from 24 June to 9 October 2026. Data: my own record of US option prices, my calculations.

3 · Selling in the morning

Selling in the first two hours of the session after the report did no better than holding to the close.

What if you don't wait for the close and sell into the morning move instead? Before looking at the prices, I wrote down three rules: sell when the bid reaches 1.5 times the price paid; sell when the stock reaches the call's strike; or simply sell about half an hour after the open, whatever the price. Once a rule fires, the sale goes through at the bid about 24 minutes later (my morning prices are recorded about every 12 minutes, 15 minutes delayed), and a call whose rule never fired is sold at the last price recorded that morning.

None of them did better than holding to the close. The main rule, 1.5 times the price paid, came out 0.7 percentage points worse than holding, with a margin of error from −10.3 to +9.0 percentage points. My morning record covers about two thirds of the trades, and only the first two hours of the session.

Calls bought before US earnings reports: selling in the morning after the report against holding to the close

Sold	Against holding	Margin of error
When the bid reaches 1.5 times the price paid (main rule)	−0.7 pp	−10.3 to +9.0 pp
When the stock reaches the call's strike	−0.4 pp	−9.8 to +9.1 pp
About half an hour after the open	−1.1 pp	−10.4 to +8.2 pp

WB Trading Labs · Table 1The same calls and the same purchase price as the main test; only the moment of sale changes. Return when sold in the morning minus return when held to the close, in percentage points (pp), averaged over the 24 evenings with at least three such reports. In bold: the rule I chose as the main one before counting. Margin of error: 95%. Data: my own record of US option prices, saved about every 12 minutes in the first two hours after the open, 15 minutes delayed; my calculations.

4 · Earnings-beat history

Whether a company usually beats its profit estimates made no difference these trades can show.

Some companies beat analysts' profit estimates almost every quarter, and it's tempting to think calls on them do better. So I split the calls by the company's record: a company goes in the first group if its earnings per share beat the estimate in at least 9 of every 10 of its recent reports (its last 10, or all of them if it had fewer), and every other company in the second. Compared evening by evening, calls on the first group came out 14.7 percentage points ahead, but the margin of error runs from −50.7 to +80.1 percentage points. That includes zero by a wide margin, so these trades show no difference either way.

5 · Weekly and monthly options

Weekly and monthly calls both lost.

This is a side observation, not something I set out to test. The rule takes the nearest expiry after the report. For most companies that's the Friday of the same week, a weekly option. Companies without weekly options list only monthly ones, so there the call still had two to four weeks to run. Both kinds lost much the same: weekly calls 41% on average, monthly calls 48% (287 and 172 calls).

What this can't answer

One earnings season, mostly volatile stocks.

This is one stretch of one earnings season, 24 June to 9 October 2026. The stocks were mostly volatile ones: the volatility priced into the calls bought, how big a yearly swing the option prices assume, averaged 99%. Other seasons, calmer stocks and other markets may look different. A loss as large as the main test's shows up clearly with this many reports; the put and the beat-history results don't.

Next comes a separate study of selling options before reports.

Method

How I counted

Rules written down first: on 12 October 2026, before computing any result of this study, I wrote down what to buy, at which prices, how to treat a call nobody would buy, how to average, and how I'd read each possible answer.
Data: my own record of US option prices (the bid and ask of every listed contract, saved after each US close, and about every 12 minutes in the first two hours after the open, 15 minutes delayed); the stocks' closing prices; the report dates and whether each company reported before the open or after the close. For section 4: Yahoo Finance's history of past earnings-per-share estimates and results, counting only reports before each trade. I publish aggregates only.
The priced move: the midpoint prices of the call and the put with the strike nearest the share price, added together and divided by the share price, from the prices before the report, on the trade's expiry.
What was bought: the listed call with its strike nearest the share price plus the priced move, kept if it was within a quarter of the priced move of that target (and never less than 1% of the share price); otherwise the report drops out of that row. With the strike at the share price, the allowance is 1% of the share price. The 0.25-delta call: the one whose delta, worked out from its own implied volatility, was nearest 0.25, if it was between 0.20 and 0.30. The put: the same, below the share price.
When: bought at the ask at the last close before the report (for a report after the close, that day's close; before the open, the previous day's close), sold at the bid at the close of the first session that contained the news. The return is the bid received minus the ask paid, divided by the ask paid. Commissions aren't included. No split or large dividend fell inside any trade.
Expiry: the nearest listed expiry not before the session that reacted to the report: for most companies the Friday of that week, for companies with monthly options only, the monthly one.
No buyer: a call with no bid at the exit counts as −100% and stays in the results.
Evenings, not trades: stocks that report on the same evening move together, so their results aren't independent; a report released the next morning before the open counts with the evening before. I average each evening first, then the evenings, each with the same weight. Only evenings with at least three reports count: 29 evenings and 430 of the 459 reports in the main test. The median and the shares use all 459 calls. The margin of error is a 95% range, from how much the evening averages differ from each other.
Selling in the morning (section 3): my morning record covers about two thirds of the trades. In the morning it only saves strikes within 15% of the share price, so when the stock was far below the strike, the recording didn't follow the call. Those trades count as held to the close, since no morning price was recorded for them, and with the stock that far below the strike the price rules could not have fired. Trades on stocks not recorded that morning at all are left out of this comparison.
Earnings-beat history (section 4): a company goes in the first group if its earnings per share beat the estimate in at least 9 of every 10 of its recent reports (its last 10, or all of them if it had fewer), counting only earlier reports. The groups are compared evening by evening, only on evenings with at least two trades in each group (17 evenings).
Reports left out: of 562 reports, 459 made it into the main test: 2 were recorded twice, 6 had no option prices, 1 had its after-report price more than 3 days late, 9 had no matching expiry at the exit, 34 had no readable priced move (no bid or ask on the options at the share price), and 51 had no listed strike close enough to the target.

I post new studies on X first.

Follow @WojciechBernPL

All studies

Does buying options before earnings pay? Across 459 US earnings reports, a call lost 42.8% on average at real prices