NQ Week Ahead (Sep 14–18, 2026): FOMC, the Dot Plot and Quarterly Expiration

Week five of the series: where the market stands, what's on the calendar, and the levels I'm working with going into the week.

NQ has spent almost four weeks inside a range a little under 900 points wide, and this is the week the calendar hands it a reason to leave.

The E-mini Nasdaq-100 (NQ) closed Friday back inside the 2026 high-volume node, the band of prices where most of this year's volume has traded, after a round trip from above the prior week's value area to the CPI lows and back. The 5-, 20- and 50-day simple moving averages sit stacked inside that node, the daily Bollinger Bands (two standard deviations either side of the 20-day average) keep narrowing, and the options market has price pinned between a put wall near 28,800 and a call wall at 30,000. Three ways of measuring a market that has stopped moving. On Wednesday at 2:00 PM ET / 20:00 CEST the Federal Reserve is expected to raise rates for the first time since 2023, and on Friday the September contract expires into quarterly options expiration. Compression plus catalyst is the setup.

Last Week, Scored

The Sep 7–11 map had one level I cared about most, the shelf at 29,450–29,500, and one tell above it at 29,600. The tell went first. An early-week push above 29,600 found no acceptance, the buyers who chased it were trapped, and price rotated back into the shelf. Then the shelf broke, 29,400 went with it, and the rotation ran to the 29,150–29,200 area I'd marked as the prior-week point of control by Thursday's open, with a retest in Globex (the overnight electronic session) on Friday before CPI. 29,000 was never touched. From there the market did what it has done every week since August 19: it went back to the year's point of control (POC, the single price with the most volume traded in 2026) and closed near it.

That makes three weeks running that the downside path in these posts has traded to its level and the upside path has stopped one rung short. Weekly value still migrated higher for the third consecutive week, because sellers keep getting the break lower and each one has been bought back to the POC by the close. I read that as a range that is tightening, not as a bias.

What the Market Took Into the Weekend

The pattern was the one this market has run since June: firm data pushes yields up, higher yields take risk off, and oil decides whether the day closes red or green. Brent went through $100 early in the week and above $108 by Thursday as the tanker blockade in the Persian Gulf dragged on, and US diesel reached $6 a gallon. Thursday's PPI landed on top of that, the 10-year jumped 11 basis points in a session, and the losing streak reached four days, the Dow's longest since April. The Treasury's first expanded buyback, on Friday, took in only $5.2 billion against a $6 billion cap, a small number with a large message about demand for duration.

Friday's CPI had every ingredient for a fifth down day, with core a tenth hot and the probability of a hike at the September 16 meeting up from about 70% to roughly 90% on CME FedWatch, which reads the odds off fed funds futures. Stocks rallied about 1% instead, because oil fell 2.5% off its highest level since May and because four sessions of selling had already priced the hike. Michigan sentiment at 47.8 with one-year inflation expectations at 4.6% went by unnoticed, which is its own tell about what the market cares about this month.

Underneath the index the split I keep coming back to held. The companies building AI infrastructure get bought and the companies it might replace get sold. Oracle reported Thursday night with $664 billion of contracted backlog and gained about 7% on Friday, while semiconductors and memory led into midweek and were the first thing sold on Thursday's PPI.

Wednesday at 2:00 PM ET

The decision comes at 2:00 PM ET / 20:00 CEST from the Federal Open Market Committee (FOMC, the Fed's rate-setting body), with Chair Kevin Warsh's press conference at 2:30 PM / 20:30 CEST. A 25-basis-point hike to 3.75–4.00% is about 90% priced. September is also a projections meeting: alongside the statement the Fed publishes its Summary of Economic Projections and the dot plot, each policymaker's anonymous forecast of where the rate will be at year end and beyond. In June, nine of nineteen dots showed at least one 2026 hike; in July, three voters dissented in favor of hiking. With the September move itself no longer information, the dots and the press conference are what the bond market will trade, and NQ follows the bond market.

A constructive reaction, to me, looks like the 10-year rejecting 5%, the 30-year failing to hold above last week's 5.36%, WTI staying under $100 and semiconductors leading the bounce. The other version is the 10-year accepting above 5%, the 30-year extending its breakout, oil turning back up and the leaders getting sold. Neither version is confirmed at 2:00 PM on Wednesday, or at 4:00 PM. The first move after an FOMC statement is the one that often reverses in the press conference, and the press conference move can reverse again overnight. What I score is Thursday, and whether the European session accepts price where Wednesday left it or rejects it. Never traded through an FOMC afternoon? The post on trading around news events lays out the three ways people handle it, flat, fade or follow, and what each one costs.

Levels

All levels are on NQ 12-26, the front month since the September 10 roll. The balance is the frame. Since August 19 price has failed to accept above the 2026 POC on every attempt and has been bought back into it on every break. Reading from the top:

Below 28,800 the references thin out to 28,565, 28,360, 28,300 and 28,000, where a weekly POC from the week of July 27 has never been retested and where the balance would have expanded by roughly its own height. Nothing in the current tape argues for that. It is on the chart because the hedging that has been dampening moves inside the range flips sign below the put wall, and a four-week range tends to travel farther than its own height once that happens; the gamma post covers why.

The level I care about most this week is 29,660, the top of last week's value: it is where buyers were trapped a week ago, it is what sellers have to hold, and acceptance above it would be the first thing in a month that changes the map. The tell below is 29,350.

The Calendar

Everything on the week, including the second-tier prints, is on the calendar page. What can move NQ, in ET and CEST:

Expiration Friday

NQ 09-26 and MNQ 09-26 stop trading at 9:30 AM ET / 15:30 CEST on Friday and settle in cash to the Special Opening Quotation, an index value built from each Nasdaq-100 stock's opening print that morning. If you rolled on September 10 there is nothing to do; December has been the market for a week. What changes is the tape around it. Friday is quarterly options expiration: index options settle in the morning, single-stock options at the close, the futures in between, and a month of positions that have pinned price between 28,800 and 30,000 gets closed or rolled. A Friday move on its own tells me little. Where price is accepted on Monday, September 21, with FOMC and expiration both behind it, tells me a lot.

How I'm Trading It

Window: 09:00–15:30 CEST, 3:00–9:30 AM ET, as always.

Monday and Tuesday are inside the balance with nothing scheduled that can break it; the 20-year auction prints after my window. I'm looking for the trade that has paid in this range for a month: a marked level breaks, the break fails, the reclaim is the entry. That's Key Level Reversal (KLR), my setup. Candidates are the edges of last week's value, 29,350 and 29,660, and the shelf at 29,480–29,500 between them. Momentum through the middle of a pre-FOMC range is the trade I skip.

Wednesday my session ends six hours before the decision. Retail sales at 14:30 CEST is inside the window, so I'm flat from 14:25 by rule and the first rotation after it is for watching. A pre-FOMC European morning is usually the tightest of the week, so the realistic expectation is a small range and maybe no setup. I don't trade the decision itself. On a prop-firm evaluation account the tape from 20:00 to 20:45 CEST is a coin flip with a wide spread, and an account whose loss limit trails its equity high can't afford coin flips.

Thursday is the session I care about. Asia and the overnight session will have had their say, and the 09:00 CEST open is the first European read on whether price accepts where Wednesday left it; the overnight high and low are the first two levels on the chart. The 14:30 stack (claims, housing, Philly Fed) is second tier, so I pause for it rather than close the book. Acceptance above 29,660 or below 29,350 that survives the European session is worth more to me than anything printed on Wednesday afternoon.

Friday the Bank of Japan lands before my window and Governor Ueda's press conference at 08:30 CEST lands inside its first hour. A hike to 1.25% is priced; a hawkish surprise and a sharp yen rally would hit NQ through the carry trade (positions funded with cheap yen) before New York wakes up, so the first hour is for reading. September expires at 15:30 CEST, the minute I close, and I'm on December anyway. Industrial production at 15:15 is noise.

Journal tags for the week: the gamma regime each morning, the range of my window on Monday through Wednesday against Thursday and Friday, and one word on Thursday's session: accepted or rejected.

Nothing here is financial advice. This is my own weekly preparation, published as-is: levels are references, not signals.