Trading ES and NQ Futures on CPI and Jobs Report Days

CPI and the monthly jobs report both release at 8:30 AM ET, an hour before the regular US cash open at 9:30 AM ET. That puts a scheduled, binary-feeling event in a thin pre-open market. Spreads widen, stops can slip, and a position that looked fine at 8:29 can be a different trade by 8:31. The dates aren't fixed by a simple rule either. Check the Bureau of Labor Statistics release schedule at bls.gov instead of assuming second Tuesday or first Friday, because schedules shift.

Why these two reports matter to ES and NQ

Equity prices reflect expected earnings discounted by interest rates. CPI feeds expectations for Fed policy: a hotter print than expected can raise rate fears and a cooler one can ease them. The jobs report is murkier because the market's reading depends on the rate environment. Weak jobs might be read as good news if traders want cuts, or bad news if they're worried about growth. The same headline can produce opposite reactions in different periods, so the reaction matters more than your opinion of the number.

The usual sequence

Before the release, activity often thins out as participants step back. At 8:30, the first move comes in seconds, frequently from systems reading the headline figure. Then traders read the details, such as core CPI components or revisions and wages in the jobs report, and the first move can partly or fully reverse. That's the pattern to plan for: a fast spike, a reassessment, then a direction that may or may not hold into the cash open. I won't put a probability on any of those stages, because I can't source one. Observe it on your own charts and log it.

Three ways to handle the day

The lowest-risk approach is to be flat at 8:30 and wait. Let the reaction settle, then trade after the 9:30 open using the opening range you've defined and tested. You give up the first move but you also skip the spike and the reversal.

A middle approach is to enter on a breakout of a range that forms after the release, with smaller size and a stop placed according to the day's wider ranges. Define the range before the day starts so you aren't choosing a window after seeing it.

The highest-risk approach is to trade the 8:30 move itself. That calls for very fast execution, stops wide enough for the whipsaw, and an assumption that your fill resembles the screen price. In practice, market orders can fill far from the last quote, and stops can slip well beyond their level. Without evidence that you're consistently better than the spread and slippage, I'd treat that as a bad use of risk.

Resizing for a wider day

If wider ranges force a wider stop, hold dollar risk constant and cut contracts. Hypothetical: your risk budget is $200 per trade. On a normal day an 8-point MES stop costs 8 x $5 = $40 per contract, so you could trade 5. On a data day you need a 16-point stop, which costs $80 per contract, so $200 supports 2 contracts (2.5 rounded down). You aren't being timid, you're keeping the loss at the number you chose. Remember that a stop can still fill worse than planned.

Prop evaluations

If you trade an evaluation, read the rule document before the release. Daily loss limits, trailing drawdown and rules about trading around news differ by firm and product, and some have restrictions on holding through events. A single slipped stop at 8:30 can end an account. The simplest protection is to be flat before the release and to know exactly which drawdown measure the firm uses.

Where levels fit

KPL levels are fixed reference prices anchored to the regular-session opening print. The anchor only exists once the 9:30 AM ET open prints, so on a data day it already reflects whatever the 8:30 release did. That means the levels give you a post-reaction map instead of a way to trade the release itself. Details are on the KPL trading page. Whatever you use, test it. Log your CPI and jobs days separately for a few quarters, include slippage, and compare results with ordinary days before deciding the setup deserves capital. A handful of data days is a small sample, so treat early results as hints, not proof.