Mid-week reversals: does the weekly high or low form on Wednesday?

[FXS London Scalper]

There's a common idea that the weekly high or low often forms around Wednesday. It can be checked quickly with daily data.

Method:

• For each week, note the day on which the weekly high and low formed.
• Count by day, for several years and several pairs.
• Compare with what you'd expect by chance.

The chance part matters: in a random walk, highs and lows are more likely on the first and last days of the week, not in the middle. So even an equal distribution across days would say something.

What may explain patterns if they appear: data scheduling. US payrolls on Friday, central bank decisions mid-week, and Monday's low activity.

How to use it, if at all: as context rather than a signal. Knowing that a strong Monday-Tuesday move might be reversed later in the week can help with targets, but won't tell you when.

Has anyone done this count properly?

[Amanhasnoname]

I did a count on EURUSD daily bars a while back, mostly out of curiosity, and the result surprised me a little. Over about seven years, the weekly high or low landed on Wednesday less often than the middle-of-the-week story suggests, and Monday and Friday were more common, which matches your point about random walks. Wednesday got a bump only in weeks where a central bank decision fell that day.

What I found more useful was splitting weeks by how Monday and Tuesday behaved. When those two days moved more than one and a half times the average daily range in one direction, the weekly extreme was already in about seventy percent of cases by Wednesday's close. That is a different question, and it gave me something practical for setting targets on Thursday.

I would not trade a reversal on the strength of it. I only use it to stop myself from chasing a late-week continuation after a very stretched start.

[PTScalper]

There's a common idea that the weekly high or low often forms around Wednesday. It can be checked quickly with daily data.

Method:

• For each week, note the day on which the weekly high and low formed.
• Count by day, for several years and several pairs.
• Compare with what you'd expect by chance.

The chance part matters: in a random walk, highs and lows are more likely on the first and last days of the week, not in the middle. So even an equal distribution across days would say something.

What may explain patterns if they appear: data scheduling. US payrolls on Friday, central bank decisions mid-week, and Monday's low activity.

How to use it, if at all: as context rather than a signal. Knowing that a strong Monday-Tuesday move might be reversed later in the week can help with targets, but won't tell you when.

Has anyone done this count properly?

Hi,

interesting idea, thank you for sharing

From my own trading system i look for high and low from two types of charts:

1) Longer term, D1 charts, i check it for stronger supports and resistances.
2) M5/M15 depends on currency pair, in most cases i watch with that like 8-14 days and look for opportunities, how to scalp volatility
on that longer term S/P from shorter timeframe perspective.

I do not trade only 1 trade, i do like 10-50 trades in that zone and focus on scalping volatility + building better cost average price on that position.
Thanks to that i can have just smaller rebound and im will be out with profit

Take a care,
Have a great trades.