The Fed blackout week: why dollar pairs can feel different before FOMC

[Fairman]

The Fed blackout week: why dollar pairs can feel different before FOMC

About ten days before each FOMC meeting, Fed officials stop giving public comments on policy. People call it the blackout period. It runs until the day after the decision.

Why care as a scalper? Because a lot of dollar movement on an ordinary week comes from Fed speakers. Someone says something mildly hawkish at a lunch event, yields tick up, the dollar follows. During blackout that source goes quiet. The dollar then trades mostly on data and whatever's happening elsewhere.

What I see, roughly, is two things. Data surprises can get a bigger reaction because there's nobody to talk them down or up afterwards. And on days with no data, the dollar pairs can be sleepier, with ranges that don't go anywhere. Mean reversion setups tend to behave better then, breakout setups less so. Not always. Enough that I check where we are in the cycle.

Then there's the last day or two before the decision, which is its own thing. People flatten positions, ranges compress, and anything that looks like a breakout often isn't.

I keep the meeting dates in my calendar with the blackout start marked. Nothing clever, just context for the week.

[Fairman]

Blackout covers Fed officials only, so other central banks keep talking and can move the dollar indirectly through EUR or JPY. An ECB speaker during Fed blackout can end up being the main driver of EURUSD for the day. And the press still writes Fed stories during blackout. Sometimes a well sourced article about what the Fed is thinking does the job a speech would normally do. Those can hit without warning and with no calendar entry, which catches people out. So blackout doesn't mean nothing will move the dollar. It just means one of the usual sources is gone, and the others matter a bit more.

[Fairman]

If you journal by market condition, tagging the blackout weeks is cheap and might show something. I wouldn't expect a dramatic difference, and with only eight meetings a year the sample builds slowly. Still, after a year or two you'd have enough to see whether your breakout trades really do worse in those weeks, or whether it's just a feeling. I've started tagging them but don't have enough yet to say anything with confidence. If anyone has tracked this for longer, I'd like to hear whether the difference holds up or fades once the sample gets bigger. My guess is it's real but small, and probably swamped by the data calendar most weeks.