Step 5, Part 1

I used my idea of examining 1, 5, 10, and 20 days moves and found that there is a relatively weak correlation between the magnitude of 5 day and 20 day moves, as well as 10 and 20 day moves. This relationship seems to exist for all 3 pairs I examined, which are AUD/USD, EUR/USD and GBP/USD. That being said the correlation is still relatively weak, probably 0.38 to 0.43, but that's considerably higher than the correlation the 1 day moves had with the other ranges, which was less than 0.20.


I'm wondering if my methodology is correct, as I took the absolute value of the moves and ran the correlation between the percentile of the various time ranges. So in the case of the 5 and 20 day ranges, the higher the percentile move of the 5 day, the higher the percentile on the 20 day. So if there is a big move on the 5 day, there is a good chance of a big move on the 20 day as well, or if there is a small move on the 5 day, there is a good chance of a small move on the 20 day as well.


Does this knowledge have any particular value? Seems sort of common sensical, no? The better question is: can this knowledge be used to create a system? Of course I would need to examine some other information, but this could lead to something I could build a system on.


That's all I have at present. I don't think I'm ready to leave step 5 yet since I'm not sure what those big moves have in common. I'll need to do some more research on that.