The first-move fake-out, spread and slippage
Lesson 17 · about 10 min
The seconds after a release are the most expensive time to trade. The price moves fastest, the book is thinnest, and the direction of the first move is, for many events, no better than a coin flip at predicting where the price will be an hour later. This lesson explains why, and what that implies for anyone trading the reaction.
What happens at 8:30:00
The number is released to the wire services and read by algorithms within milliseconds. The first trades are between machines. Market makers, who were quoting a tight spread at 8:29:59, widen or pull their quotes because they do not want to be run over. For a few seconds the visible book in ES might show a fraction of its normal depth.
Then three groups act in overlapping waves:
- Algorithms trading the headline surprise. They push price in the "obvious" direction. Fast, and often overshoot.
- Discretionary desks reading the internals. Slower, sometimes disagreeing with the headline. This is where a beat with bad revisions gets sold.
- Positioning unwinds. Whoever was positioned for the wrong outcome has to exit, and their exits can dominate everything else.
The "first move" is wave one. The "real move" is often waves two and three, which may reverse it.
The fake-out, in numbers
It is common on data days for the first one-minute move to be at least partly retraced within fifteen minutes, and for the direction at 9:30 to differ from the direction at 8:31 a meaningful fraction of the time. The exact fractions depend on the release, the regime and the instrument, and you should measure them in your own log. What is consistent across studies and desks is that chasing the first candle has poor expectancy once spread and slippage are included.
The mechanical reason is that the first move is driven by the headline, and the headline is the least informative part of the report (Module 3). The second reason is that the first move is where stops on both sides get hit, so the price is pushed through levels by forced flow rather than by information.
Key idea: The first move after a release reflects the headline and forced flow, not the full report. Its direction is a weak guide to the next hour, and the cost of entering during it is at its highest.
Spread and slippage during data
Three costs rise at the same time:
| Cost | Normal ES (regular hours) | Seconds after a major release | Notes |
|---|---|---|---|
| Bid-ask spread | 1 tick (0.25 points) | Often 2 to 4 ticks, occasionally more | Widens more in the pre-market 8:30 slot than at 10:00 |
| Depth at best price | Hundreds of contracts | Tens, sometimes single digits | A modest market order sweeps several levels |
| Slippage on a stop | Usually 0 to 1 tick | Several points is common; double digits on CPI or FOMC | Stops become market orders and fill at the next available price |
In FX, retail spreads on EUR/USD can go from a fraction of a pip to several pips at the print, and some brokers widen far more. In gold futures, the spread widens and the market can gap several dollars between prints. Bond futures widen less, because the participants are mostly institutional, but they move a lot.
For the trader, this means: a stop placed 10 points away in ES is not a 10-point risk on CPI day. It might be a 15- or 20-point risk. Size accordingly (Module 5, lesson 3), and never assume a limit order to exit will fill in a fast market.
Reading the first fifteen minutes
Instead of trading the first candle, use it as information. Watch:
- The high and low of the first minute. These become the reference levels. Price holding above the first-minute high after ten minutes is more meaningful than the first-minute move itself.
- The 2-year yield's direction versus ES's direction (Module 1). If they disagree with what the headline "should" do, wave two is disagreeing with wave one.
- Whether the initial move stalls at a pre-marked level (yesterday's high, the overnight range, a prior value area). Stalls at known levels are where reversals start.
- Volume. A first move on huge volume that then goes quiet has exhausted the forced flow; a first move that keeps building volume in the same direction is being confirmed.
By 8:45 you usually know whether the first move is being accepted or rejected. That is when a trade with a definable stop becomes possible.
The 10:00 releases are different
ISM and the housing data land at 10:00 a.m. with the cash market open and a full book. The spread stays tighter, slippage is smaller, and the fake-out is less violent. If you want to learn to trade reactions, the 10:00 releases are a gentler classroom than the 8:30 ones. FOMC at 2:00 p.m. is the exception: a fully open market but the largest event, with its own two-wave structure (statement, then presser).
Try it: For the next three 8:30 releases, do not trade. Record, for your instrument: spread at 8:29:55, widest spread you see in the first thirty seconds, the first-minute high and low, the price at 8:45 and at 9:30. After three, you will have a personal fake-out rate and a personal slippage estimate that no course can give you.
Recap
- The first seconds after a release are traded by algorithms and forced flow; the first move reflects the headline, not the report.
- The first-minute direction is a weak predictor of the direction an hour later; measure the rate in your own log.
- Spread, depth and stop slippage all deteriorate at once; a 10-point stop in ES can be a 20-point risk at the print.
- Use the first fifteen minutes as information: first-minute range, the 2-year versus ES, stalls at known levels, and volume.
- 10:00 releases are a gentler place to learn reactions than the 8:30 slot.
See it drawn
Original diagrams for the ideas on this page. Illustrative, not real market data.