Blog / Triple Witching Dates for 2026 and 2027, and What the S&P 500 Did on All 107 Triple Witching Days Since 2000

8 min readJorgAI TeamOct 23, 2026

Triple Witching Dates for 2026 and 2027, and What the S&P 500 Did on All 107 Triple Witching Days Since 2000

Red pushpin marking a date on a paper calendar, cover for the triple witching dates and data post

The 2026 triple witching dates are March 20, June 18, September 18 and December 18. The 2027 dates are March 19, June 17, September 17 and December 17. Two of those eight are Thursdays, not Fridays: June 18, 2026 and June 17, 2027, because the Juneteenth market holiday falls on the third Friday in June both years, and when the exchange is closed on the third Friday, expiration moves to the prior session. The only remaining triple witching of 2026 is Friday, December 18.

We also went back through every triple witching day since 2000, all 107 of them, in SPY's daily data. The short version: the day is louder than it is large. Volume ran a median 1.23 times the prior 20 sessions and above normal on 72% of the days, yet the price move was slightly smaller than an ordinary session, the index closed higher on 51 of 107 (47.7%) against 54.4% of all days, and the single biggest reason SPY looks like it drops on every triple witching in a raw price feed is a dividend, not the witching. Details and the table follow.

Triple witching dates in 2026

  • Friday, March 20, 2026: done. SPY closed -1.43% on volume 1.84 times normal.
  • Thursday, June 18, 2026: done, and a Thursday because Juneteenth (Friday, June 19) closed the market. SPY closed +1.04% on 1.42 times normal volume.
  • Friday, September 18, 2026: done. SPY closed +0.13% on 1.64 times normal volume.
  • Friday, December 18, 2026: the next one, and the last of the year. December is the heaviest-volume witching of the four (median 1.47 times normal since 2000).

Triple witching dates in 2027

  • Friday, March 19, 2027
  • Thursday, June 17, 2027: a Thursday again, because June 19, 2027 is a Saturday and the Juneteenth holiday is observed on Friday, June 18, when the market is closed.
  • Friday, September 17, 2027
  • Friday, December 17, 2027

The rule behind the list: triple witching is the third Friday of March, June, September and December, when stock index futures, stock index options and single-stock options all expire together (many traders call it quadruple witching, counting single-stock futures). When that Friday is a market holiday, the session before it carries the expiration. It happened on March 20, 2008 (Good Friday) and on June 18, 2026, and it will happen on June 17, 2027. What triple witching is and why it exists is covered in the September guide; this page is the dates and the data.

What the S&P 500 did on all 107 triple witching days since 2000

The dataset: every triple witching session from March 2000 through September 2026, measured on SPY with dividend-adjusted closes from Yahoo Finance. For each day we recorded the close-to-close return, the day's high-to-low range, volume as a multiple of the prior 20 sessions, and the return over the five sessions after. The comparison set is every trading day since 2000, 6,737 of them.

  • Direction: higher on 51 of 107 days, 47.7%. All trading days since 2000: 54.4% higher. Average close-to-close move -0.16%, median -0.09%; the average day in the comparison set is +0.04%.
  • Size: the average absolute move was 0.72% against 0.80% for all days, and the median high-to-low range was 0.93% against 1.05%. 53% of triple witching days closed within half a percent of the prior close, against 47% of all days. Triple witching is, on the median day, slightly calmer than normal.
  • Volume: median 1.23 times the prior 20 sessions, above normal on 72% of days, with a record 5.15 times on March 16, 2001 and 2.92 times on December 20, 2024. The typical trading day sits at 0.95 times its trailing average. This is the one thing about the day that is reliably unusual.
  • The week after: average -0.00% over the next five sessions, higher 42 times out of 107. The average five-session stretch since 2000 is +0.19% and higher 58% of the time, so the post-witching week has run a little below par.
Bar chart of SPY volume on each of the 107 triple witching days from March 2000 to September 2026 as a multiple of the prior 20 sessions, mostly below 1x in the 2000s and mostly 1.2x to 2x from 2013 on, with a 5.1x spike in March 2001 and 2.9x in December 2024

Why SPY looks like it falls on every triple witching

Pull SPY's raw closing prices and you will find it closed lower on 84 of the 107 triple witching days, a 78% hit rate that looks like a tradable pattern. It is not. SPY goes ex-dividend on the third Friday of March, June, September and December, the same day as triple witching, on all 107 days in this sample. The quarterly dividend, about 0.43% of the price on average, comes off the quote at the open. Adjust for it and the 84 down days become 56, and the average move goes from -0.59% to -0.16%. Any triple witching study on the S&P 500 ETF that does not say "dividend-adjusted" is measuring the payout.

The four witchings are not the same

  • March (27 days): higher 13 times, the widest ranges of the four (median 1.16%), median volume 1.07 times normal, and the week after averaged +0.49%.
  • June (27 days): higher 13 times, median volume 1.23 times normal, week after -0.39%, higher only 7 times.
  • September (27 days): higher 12 times, median volume 1.34 times normal, and the weakest week after of the four at -0.82%, higher 8 times out of 27. That is the September effect doing its work, not the expiration.
  • December (26 days): higher 13 times, the heaviest volume (median 1.47 times normal, with the 2.92 times of 2024 the biggest since 2001), and the strongest week after at +0.74%, higher 17 times out of 26, which is the run into the year-end rally.

The day has changed over the years

In the 2000s, triple witching volume was ordinary: median 0.99 times normal across 40 days, with price moves averaging 0.90% in a decade that included two bear markets. In the 2010s the median volume multiple rose to 1.27 while the average move shrank to 0.46%. Since 2020 the multiple is 1.42, the average move is back to 0.84%, and the index has closed higher on only 8 of 27 witching days. The growth in index-tracking funds and in options volume shows up in the volume line; the price line mostly follows whatever the market was doing that year.

The six largest triple witching moves since 2000

  • March 20, 2020: -4.31%. The pandemic crash, with the Fed's emergency measures a few days old. Volume 1.33 times normal.
  • September 19, 2008: +3.97%. The Friday after Lehman, when the Treasury's rescue plan and the short-sale ban were announced. Volume 1.54 times normal.
  • December 15, 2000: -2.26%. The dot-com unwind, a day after the election was finally decided.
  • September 15, 2000: -2.19%. Oil at a decade high and a run of earnings warnings.
  • March 21, 2003: +2.13%. The first days of the Iraq war, part of the sharpest rally of that spring.
  • March 20, 2009: -2.13%. Two weeks after the financial-crisis low.

Every one of them had a headline that had nothing to do with expiration. The expiration mechanics add volume and some intraday noise at the open and in the closing auction; they do not supply the direction.

What this means if you trade or automate around the dates

  • Expect volume, not a trend. The closing auction on triple witching is the biggest of the quarter, and index rebalances often land the same day. Fills near the close can be worse than usual on thin names and better than usual on index heavyweights.
  • Do not back-test SPY on raw closes. The ex-dividend overlap turns a coin flip into a fake 78% down rate. Use adjusted data, or measure on the index itself.
  • Know the Thursday dates. In 2026 and 2027 the June expiration is on a Thursday. A schedule, a bot or an options position keyed to "third Friday" will be a day late both years.
  • Stops still run. Wider opening noise can touch a tight stop that would have survived an ordinary day. The stop-loss guide covers how far from the price a stop should sit for the setup, and the expiration-Friday study covers the monthly expirations that are not triple witchings.

JorgAI's auto-trader treats triple witching like any other session: the member's stops, caps and daily loss limit apply, and no rule in the system buys or sells because a date is on the calendar. If you want rules enforced on your own account on every session, including the loud ones, the start page walks through connecting a broker.

Quick answers

When is triple witching in 2026? March 20, June 18 (a Thursday, because of the Juneteenth holiday), September 18 and December 18, 2026.

When is triple witching in 2027? March 19, June 17 (a Thursday, because Juneteenth is observed on Friday, June 18), September 17 and December 17, 2027.

Does the stock market go down on triple witching? Not reliably. Since 2000 the S&P 500 ETF closed higher on 51 of 107 triple witching days (47.7%), with dividends accounted for, against 54.4% of all days. Raw price data shows far more down days only because SPY goes ex-dividend the same day.

Is triple witching more volatile? Volume is, at a median 1.23 times normal. Price moves are not: the median range was 0.93% against 1.05% on an ordinary day.

What is the difference between triple and quadruple witching? The same day. "Quadruple" adds single-stock futures to index futures, index options and stock options; most people still say triple witching.

Data: SPY daily open, high, low, close, dividend-adjusted close and volume from Yahoo Finance, March 2000 to October 22, 2026; triple witching dates computed as the third Friday of March, June, September and December and moved to the prior session on exchange holidays (March 20, 2008; June 18, 2026); 2027 holiday dates per the NYSE holiday calendar. Returns are close to close; "the week after" is the fifth session's close against the witching close. Past patterns are not predictions.

Let the AI do the trading.

Set it up in minutes. 7 day free trial on Starter and Pro.

Get started free