Most weeks spread their risk out. This one does not. The market is closed Monday for Labor Day, the middle of the week is nearly empty, and then the two inflation prints that matter land back-to-back at 8:30 a.m. on Thursday and Friday — after which the Fed meets on September 15-16 with nobody available to walk anything back, because the pre-meeting blackout already started on September 5.
Trading SPY before CPI is less a forecasting problem than a scheduling one. If you hold SPY overnight, the shape of the week matters more than any single call on the number. Four trading days produce four overnight sessions, and the gap risk in them is nowhere near evenly distributed. Two are close to inert. One is loaded. Here is the map before you size anything.
The setup you are carrying into Tuesday
Friday, September 4 changed the tone. August nonfarm payrolls came in at 162,000 against a Dow Jones consensus of roughly 53,000, with the unemployment rate steady at 4.1%. The index reaction was mild — the S&P 500 slipped 0.38% to 7,718.60, the Nasdaq Composite fell 0.29% to 26,506.99, the Dow lost 0.51% to 53,414.25 — but the bond market’s reaction was not. Treasury yields jumped, with the 2-year reaching its highest level since January 2025 as traders raised the odds that the Fed’s next move is a hike rather than a cut.
That reversal in polarity is the thing to internalize before Tuesday. For most of the last two years, strong data was tolerable and soft data was welcome. In a tape pricing hike risk, hot numbers cut the other way — and this week hands the market two of them in thirty hours.
The overnight map, session by session
Four nights, and they are not interchangeable:
| Overnight session | Hours closed | Scheduled catalyst | Relative gap risk |
|---|---|---|---|
| Fri 9/4 close → Tue 9/8 open | ~65.5 | None scheduled | Moderate — duration, not news |
| Tue 9/8 → Wed 9/9 | ~17.5 | NFIB optimism (6 a.m.), consumer credit | Low |
| Wed 9/9 → Thu 9/10 | ~17.5 | Oracle after the close; PPI, core PPI and jobless claims at 8:30 a.m. | High |
| Thu 9/10 → Fri 9/11 | ~17.5 | CPI and core CPI at 8:30 a.m.; UMich sentiment at 10 a.m. | Highest |
Night one: 65 hours of nothing in particular
The Friday-to-Tuesday hold is the longest closed stretch of the week — roughly 65 and a half hours versus the usual 17 and a half. Nothing is scheduled inside it. That does not make it safe; it makes it a duration bet — more calendar for something to happen in, and no session in which to adjust. The risk here is unscheduled and unforecastable, which makes it a position-sizing problem rather than a research problem.
The quiet middle
Tuesday night into Wednesday is the emptiest session on the board. NFIB small business optimism at 6 a.m. and July consumer credit at 3 p.m. are second-tier releases that rarely move index futures on their own. If you want a night to carry normal size, this is the one the calendar gives you.
Thursday’s open is the one people underestimate
Wednesday night stacks two unrelated things into one gap. Oracle is scheduled to report fiscal first-quarter results after the close, and in a market where AI-infrastructure spending is a primary index driver, a single large-cap guide can move futures well beyond its own ticker. Then at 8:30 a.m. Thursday, August PPI and core PPI arrive alongside weekly jobless claims.
PPI gets treated as CPI’s undercard, and it usually is. But it is released first, and in a week where the market is actively repricing hike odds, producer prices function as a preview that traders will extrapolate from — sometimes wrongly. A hot PPI Thursday morning can put SPY into Thursday’s session already priced for a hot CPI, which changes the risk of the Thursday-night hold before you even see the CPI number.
Trading SPY before CPI: what the nowcast is actually implying
The Federal Reserve Bank of Cleveland’s Inflation Nowcasting tool, as of September 4, had August headline CPI tracking at 3.38% year over year and 0.36% month over month, with core CPI at 2.38% annually and 0.20% sequentially.
The annual figure looks unremarkable next to July’s 3.4%. The monthly figure does not. July headline CPI rose just 0.1% month over month, and June was negative 0.1%. A 0.36% monthly print would be more than triple July’s pace. Much of that gap is energy — the energy index was up 14.7% over the twelve months ending in July, against 3% for food — which is why core is nowcast at a far calmer 0.20%.
A nowcast is a model, not a forecast of the print, and it is regularly wrong by a tenth or more. Treat it as a description of where the risk is skewed, not as a number to trade against.
Three branches into Friday’s open
- Core comes in at or below 0.2%. Headline can run hot on energy and still be dismissed as a supply story. This is the branch where a gap down gets bought quickly, because it hands the Fed a reason to sit still on the 16th.
- Core prints 0.3% or higher. The hike conversation stops being a tail scenario two sessions before the meeting, with no Fed speaker able to soften it. This is the branch with the widest downside distribution, and the one where a gap is least likely to be a fade.
- Both land near consensus. The most probable outcome, and the least interesting — the gap compresses toward noise and Friday trades on positioning into the FOMC rather than on the data.
Notice what these branches have in common: none of them are tradable at 8:31 a.m. if you are already holding. The number lands seventy minutes before the open, futures reprice instantly, and the SPY gap is set before you get a fill. That is worth remembering alongside the fact that the overnight futures print is not your gap — the ES quote at 5 a.m. is a live, thin market, not a promise about where SPY opens.
What to decide before Thursday’s close
The decisions worth pre-committing to are all size and structure decisions, and they are all cheaper to make on Tuesday than at 3:55 p.m. Thursday:
- Whether you carry any position into Thursday night, or flatten and let Friday come to you.
- If you carry, what fraction of normal size. A stop does not execute overnight, which is why position sizing is the real risk control on an overnight hold, not the stop level you wrote down.
- What you will do with a gap you did not want — including whether you believe the historical evidence on whether gaps actually fill, and at what size, well enough to act on it under pressure.
This is exactly the kind of week where a nightly pressure reading is useful as a structuring tool rather than an answer. SpyTomorrow’s readings describe the balance of overnight pressure heading into a session; they do not know what the Bureau of Labor Statistics will publish on Friday morning, and no model does. Their value on a week like this one is that they force the position to be evaluated before the catalyst, not after it.
The calendar cannot tell you which direction Friday opens. It can tell you which night deserves half your usual size — and that is the more useful of the two answers.
This article is for educational and informational purposes only. It does not constitute financial advice, investment recommendations, or a solicitation to buy or sell any securities. Past performance and historical data do not guarantee future results. All trading involves risk, including the potential loss of principal. Always do your own research and consult with a qualified financial advisor before making trading decisions.
