How to Trade a Fed Decision
To trade a central bank rate decision, start from what the market already expects, because that is priced in. The rate announcement is usually anticipated; the statement wording, the projections and the press conference are where the unexpected part arrives.
A rate decision is not one event. It is an announcement, often a set of projections, and then a press conference — and the market frequently moves one way on the first and back the other way on the third.
Before you start
The market-implied expectation before the announcement, because that is what is priced. Rate futures give a probability for each outcome, and it is public.
An understanding that the decision is usually the least surprising part. Central banks signal heavily in advance, so the number itself is often already known within a narrow range.
A decision about your position made the day before, not on the day. Flat, reduced, or held with a plan. Decided when there is time to think.
The steps
1. Find out what is already expected
Rate futures imply a probability for each outcome. A decision matching a high implied probability is close to a non-event for price.
2. Treat it as three separate events
The announcement, any accompanying projections, and the press conference some minutes later. Each can move price independently and they frequently disagree.
3. Read the statement’s changes, not the statement
The wording is mostly reused from the previous meeting. What moved is which phrases changed, and services that publish a comparison exist for exactly that reason.
4. Expect the first move to be unreliable
The immediate reaction is fast, thin and frequently reversed once the conference provides context. It is the part of the event with the least information and the widest spread.
5. Account for the spread all session
On this site’s shared series a round trip measures about 2% of the median bar range of 0.493, and on a decision day it is elevated from before the announcement until well after the conference.
6. Size for a fast move against you
Price can move faster than orders execute. The largest single bar on this site’s shared series measured 2.338 against a median of 0.493, and event sessions are where that tail lives.
7. Wait for a range before taking a directional view
Once the conference is over and a range has formed, there are levels to trade against. Before that there is a fast market and no structure.
How to tell it worked
The implied expectation was checked before the announcement, in every case.
Your position was decided at least 1 day in advance.
0 trades were taken in the first 5 minutes after the announcement.
And the statement was compared against the previous one, rather than read fresh.
Why the initial move reverses so often
Because the first reaction is to a headline and the second is to an explanation. Automated interpretation of the statement moves price in seconds; the press conference adds context that frequently changes what the statement meant.
And because liquidity is at its lowest exactly then. Participants step back before the announcement, so the initial move happens on a thin book and is easily reversed.
What happens to expected volatility
It is priced up in advance and collapses afterwards. Options across the event carry a premium for the expected move, and that premium disappears once the uncertainty is resolved.
Which means an options position can lose money on a correct directional call. The move happened, and the expected-volatility component that was paid for evaporated at the same moment.
On this site’s shared series an implied volatility of 7.9% reproduces both measured leverage results within 0.03 percentage points, which is the arithmetic linking expected movement to what a position actually experiences.
The simplest workable approach
Be flat through the announcement and the conference. No position, no exposure to the fastest and widest market of the week, and no stop being filled somewhere you did not choose.
Watch the session and mark the range that forms afterwards. By the time the conference ends there is a high and a low made on real volume, and those are levels rather than reactions.
Trade the following session against them, if your method has a setup there. The event has been priced, the spread has normalised, and the structure it created is now ordinary structure.
That approach gives up the event’s biggest moves entirely. It is a real cost and it is the trade being made — the moves given up are the ones that happen in the conditions where a stop is least reliable, and giving them up is why the approach is workable for anybody not set up to trade microseconds.
The original data
Of the 24,971 unique videos in research/search-study-corpus.jsonl, 7 mention central bank decisions
in the title, at a median of 1,723 views across 7 channels — and none of the 7 is instruction-shaped.
News trading appears in 1 at 42,108 and economic calendars in 0. The counts come from
site/corpus_count.py.
7 videos, none instructional, at a 1,723 median. The most heavily anticipated scheduled event in most markets has no instructional coverage in a 25,000-video corpus, which is a striking gap for something whose date is published a year in advance.
The answer to the question on that chart is that the rate was never the event. The statement’s wording or the projections changed — and since the decision itself was already priced, everything that moved came from the parts nobody quotes in the headline.
When it fails
The failure is trading the first thirty seconds, and it combines every disadvantage at once. The spread is at its widest of the week, the book is at its thinnest, the move is faster than orders execute, and the information being reacted to is a headline that the press conference may contradict within the hour. A position taken there is paying the highest cost available to act on the least reliable information in the session.
The second failure is ignoring the implied expectation. The decision is mostly priced.
A third is reading the statement fresh. The changes are the content.
A fourth is treating it as one event. It is three, and they can disagree.
A fifth is normal position sizing. The tail on these sessions is wide.
And a sixth is expecting an options position to pay on a correct call. The volatility premium collapses at the same moment.
Related
Implied volatility is how the expected move is priced beforehand. Trading sessions covers when these land relative to each market. And circuit breakers are what happens if the reaction is extreme enough.
The reversal is the part I had to learn the hard way. The initial move on the announcement is frequently undone within the hour once the press conference starts and the language gets explained, and trading the first thirty seconds means trading the half of the event with the least information in it.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.