Event nights: what options price, and what happens

Earnings straddle through the report

Is the options-implied earnings move mispriced, and does trading it pay after costs?

Run once on 3 Nov 2026, pre-registered, 9,402 earnings events from 2021 to 2026. Each straddle was opened at the close before the report and closed at the close after it, filled at the quoted bid and ask.

Even filled at mid, selling averages -2.9%, though the median trade makes +9%: rare large moves outweigh the many small wins.

Implied vs real move, by year

Takeaway. Use "implied is rich" as a starting assumption for how big a move will be, not as a trade signal. Short-volatility ideas stay defined-risk. Iron condors are tested in the next section.

Earnings iron condors through the report

Does capping the tail keep the implied-move premium after costs?

Run once on 4 Nov 2026 on 9,854 earnings events from 2021 to 2026. Each condor was sold at the close before the report and bought back at the close after it. All four legs were filled at the bid and ask, with $0.0065 per share per leg each way.

The history run is descriptive only. The verdict is forward-only, judged once after 2027-04-10. The history file reports bid/ask fills only; it has no mid-fill run. The same condor on a no-earnings night lost about as much, so most of the loss is the round trip's cost, not the earnings move.

Takeaway. The overpricing is measured, but selling it with condors has not been shown to pay after worst-price fills.

FDA decision days (PDUFA)

Do out-of-the-money wings pay into FDA decision dates, where a binary approval or rejection can move the stock?

Run once, pre-registered on 29 Oct 2026, on 830 frozen FDA decision events from 2018 to Oct 2026. Bought the 0.20-delta strangle or one wing at the ask before the decision and sold at the bid, with $0.0065 per share each side.

Takeaway. The market priced a bigger move than the decision delivered: median implied 14.7% against a realised 7.2%, and the real move beat the implied on only 26% of 753 events. Buying wings into PDUFA dates lost about half the premium on average.

The skew is positive, so the wings do produce rare large winners, but the mean after costs is deeply negative. n >= 30 held everywhere.

Election nights

How SPY moves from the close on election day to the next day's close, and what happens overnight.

SPY's Nov 4 straddle priced a ±1.2% move at the Nov 3 close, about three times the 0.4% median day of the last 65 sessions. Overnight, SPY fell to -1.4% around 10 pm ET and was back to -0.3% by 1 am ET.

Takeaway. The reliable part is a big move: about 2% in four of six elections. Only 2016 had a truly dark night, and 2022 fell the next day. Six events cannot support a conclusion, and three of them were midterms.

Overnight figures are SPY's after-hours and pre-market trades (SIP), which are thinner than futures. The 2016 low is a thin pre-market print.

Dark night, then a rally? CPI and jobs-report mornings

When SPY has a rough night before a CPI or jobs report, does it recover during the next day's session?

Run once on 4 Nov 2026, pre-registered. 147 CPI and jobs-report mornings from April 2016 to June 2022, each compared with the same trade on ordinary days in the same years. SPY was bought at the 9:30 open and sold at the 4:00 close, with 0.025% cost each way.

How often a down open recovered

Takeaway. Tested, no edge after costs. Buying every event-day open lost a little, worse than ordinary days. Buying only after a gap down had 26 trades, below the 30 the test needs, so it is not measurable at this n; the study still fails its gate. A down open on report day closed above the prior close less often than an ordinary down open. 2016's election-night rally is the pattern people remember, and it did not repeat on these mornings.

Lower bound: the one-sided 95% bound of the mean, resampled by month. The gap-down trial had 26 trades, below the 30 the test needs. Neither trial passed the first stage, so later years (2022 to 2026) were never tested. Recovery counts cover April 2016 to July 2024; recovered means it closed above the prior close.

This week's earnings calls

Pre-event calls made before each report, graded after it. A condor is an iron condor that sells the implied move with capped risk.

The Nov 5 calls were made at 3:39-3:46 PM ET at worst-price fills. Condors are unvalidated; the forward verdict on them is after 2027-04-10.

Next studies

Three studies are moving. Each step is marked done, active or next.