Selling puts before FOMC, CPI and jobs reports
Should you avoid selling puts into scheduled macro events? We tested the three big ones. FOMC was the opposite of what most people expect, CPI did nothing, and the jobs report was the one to respect.
How we tested it
On QQQ, TQQQ and SPY, 1,422 days each from late 2020 to 2026, we sold a 25-delta put on every day and split the days by calendar distance to the next scheduled event, using the published schedules (as scheduled, not as later revised). The measure was how often the strike was breached, scored out of sample on the last 40% of the data. Across all days the out-of-sample breach rate was 13.9%.
FOMC: the days before a decision were the safest
Puts sold within two days before a Fed decision breached 11.1% of the time, the lowest of any window, and the sign held on all three tickers (QQQ 11.1%, TQQQ 13.0%, SPY 9.3%, against 13.5%, 16.2% and 12.1% on all days). The likely reason is that implied volatility is bid up into the meeting, so a 25-delta strike sits further from the price.
A rule that stands aside before FOMC would have removed the best-behaved entries. Note the sample: 162 entries are only about 11 meetings, so this is a consistent sign, not a precise number.
CPI did nothing; payrolls is the one to watch
Every CPI window matched the overall breach rate within a fraction of a point. Entries within two days of CPI breached 13.8% against 13.9% for all days.
The jobs report was different. Puts sold within two days before a payrolls release breached 18.0% of the time out of sample, about 4 points worse than all days, and worse in-sample too, on all three tickers (QQQ 16.7%, TQQQ 19.2%, SPY 17.9%). A payrolls surprise is a gap with no volatility run-up beforehand, so the strike is not pushed out the way it is before a Fed meeting. The sample is about 30 releases, so the estimate is consistent but wide.
What it changes in practice
Do not treat every macro event the same. A blanket stand-down around FOMC is not supported by this data, and CPI timing does not matter for a 25-delta put held 30 to 45 days. If you want one event filter, payrolls is the candidate, at the cost of skipping about one entry day in seven. Like the 200-day gate, that is a modest trade-off, not an edge, and it does nothing for gaps that arrive between scheduled events.
Historical replays are not a forecast. Options can lose more than the premium collected. This is education, not a recommendation to trade anything.