Event calendar

This week's event calendar

Which S&P 500 names report earnings this week, and how much extra dispersion we measured after that kind of event. It says how much a stock tends to move, not which way.

Dispersion, not direction

The multiplier measures how much more than usual a stock moved after an event of that type. It does not say whether it goes up or down: predicting direction sits in our graveyard of signals, and it stays there.

Average over a 5-session window

It is the average realised volatility over the [t0, t0+5] window divided by the name's own prior volatility. It is not how much a specific name will move today: it is a cohort average over a window.

Information, not advice

The options market already prices this volatility; there is nothing to buy with it. It helps you know what is on the agenda before you decide. It does not decide for you.

Backtest results (simulation), not real trading. Not investment advice. Past performance does not guarantee future results.

Earnings scheduled this week

S&P 500 names with an earnings release in the coming days, per the calendar. Measured multiplier for earnings: ×1.089 (average over the 5-session window, not a per-day or per-name value).

Data as of 2026-08-20

DayBefore the openAfter the closeTime not confirmed
Monday, August 24
0 names
Tuesday, August 25
1 name
INTU
Wednesday, August 26
9 names
SJMWSM
ACRMCRWDHPQNVDASNPSVEEV
Thursday, August 27
8 names
BBYDGDLTRHRL
ADSKMRVLULTAWDAY
Friday, August 28
0 names

Open windows

Names with an event already published whose 5-session window we measure is still open.

AMDExecutivesAONExecutivesBAExecutivesBAXExecutivesGEHCExecutivesHONExecutivesINTCExecutivesLHXExecutivesLHXExecutivesLHXGuidanceLMTExecutivesUNHExecutives

The three event types that move volatility

Seven types measured over 5,211 news items with a catalyst (2026-06 → 2026-08). Ratio = realised volatility over [t0, t0+5] divided by the name's prior volatility; days with no event: 0.964. p from a date-clustered bootstrap; Benjamini-Hochberg deflation at 10%.

TypenRatiopReading
Measured and significant
Earnings1,4491.089< 0.0001Quarterly results release. This was the sanity check: it had to come out elevated, and it did.
Executives2671.115< 0.0001CEO/CFO changes or statements. Without filtering news attribution the effect fades, so the noise was noise.
Guidance521.1540.0141Guidance revisions. The largest effect, but also the smallest sample: the least precise of the three.
Measured and discarded
Analyst ratings2,9720.9141.000Noise. It is 42% of classified news and does not lift volatility above no-event days.
Mergers & acquisitions2010.8780.996Reduces volatility: after an offer the stock anchors to the offer price and stops moving with the market.
Regulatory2170.9200.929Indistinguishable from a no-event day.
Capital actions530.9620.515Indistinguishable from a no-event day.

Multipliers are published at their pre-registered values even though the attribution-filtered subgroups measure slightly higher: they err on the conservative side, and tuning them to the same data they were measured on would be cheating.

Get the week's calendar every Sunday, with the measured multiplier.

Free. No direction, no advice.