Some data

Capture

Roughly 10 years of data on Eur/Usd, measuring what hour of the day the high/low occurs in. So far, it doesn’t look right to me, so I’m checking data first thing tomorrow; Yet, an interesting result if true. I’m trying to recreate a statistic found in the thread which intuitively makes sense, however this current data set I’m looking at looks.. Unusual. Hmm…

Update: I added a bit of data, here’s what the numbers are showing:

set

If I assume that my calculations are correct (which I’m still not so sure about even though I checked the data), It’s showing that 2:00 EST marks the price extreme of the day (High AND Low) 8-10% of the time. This means it’s saying that 1 hour before LO has some significance.

Lets see what else the data is suggesting:

stat

Clarification: Between NYO/LC means the time period while both London and NY are active, 8:00-12:00 EST.

If we consolidate data into time periods, the data starts to look a bit more intuitive. The chance to make a price extreme during the London session is roughly 50%, and in between the two most liquid sessions, 60%. Therefore we can say 60% of the time after NY has closed, price will remain range bound until London begins again

What’s potentially flawed in the data? Barring calculation errors, the first thing to look at is the definition of the day. The current set-up is midnight est, in the middle of the Asian session. Results might be slightly different if I ran data LO to LO. Issues of DST may also be an issue, especially since I’m looking at years worth of data; The time for LO isn’t always the same.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s