Hi all,
For the lack of a better subject line I am calling this "time-flexible bars" and refer to the following:
If I look at a 5-minute bar chart, the 5-minute period is set by default from say 9:35 to 9:40 and from 9:40 to 9:45 and from 9:45 to 9:50 etc. I wonder if there is a way to change this to for example 9:36 to 9:41 and from 9:41 to 9:46 and from 9:46 to 9:51 etc.
It may seem irrelevant but the challenge I find is that whatever happens in a 5-minute interval (or any interval) is to some degree random. For example, the Stochastics indicator may be below 80 because the close of the 5-minute bar happens to be 1 tick lower just when the bar closes but actually it was 1 tick higher 1 second before and after (which would have made Stochastics stay just above 80). Therefore, whether you get a valid signal or not is subject to a certain "randomness". I am not yet sure it matters or if any indicator/strategy will end up with some randomness somewhere, or if a strategy that is this "sensitive" can be a good strategy, but I hope you get my point.
So, I am thinking about a way to remove that randomness and capture signals "around" the 5-minute interval that I am looking at. And changing the default intervals may be a way to do that. For example, I could feed 3 different Stochastics figures from three different 5-minute bars (9:34 to 9:39, 9:35 to 9:40, 9:36 to 9:41) and then deal with the results.
Does anyone have thoughts on that? Thank you!
p
For the lack of a better subject line I am calling this "time-flexible bars" and refer to the following:
If I look at a 5-minute bar chart, the 5-minute period is set by default from say 9:35 to 9:40 and from 9:40 to 9:45 and from 9:45 to 9:50 etc. I wonder if there is a way to change this to for example 9:36 to 9:41 and from 9:41 to 9:46 and from 9:46 to 9:51 etc.
It may seem irrelevant but the challenge I find is that whatever happens in a 5-minute interval (or any interval) is to some degree random. For example, the Stochastics indicator may be below 80 because the close of the 5-minute bar happens to be 1 tick lower just when the bar closes but actually it was 1 tick higher 1 second before and after (which would have made Stochastics stay just above 80). Therefore, whether you get a valid signal or not is subject to a certain "randomness". I am not yet sure it matters or if any indicator/strategy will end up with some randomness somewhere, or if a strategy that is this "sensitive" can be a good strategy, but I hope you get my point.
So, I am thinking about a way to remove that randomness and capture signals "around" the 5-minute interval that I am looking at. And changing the default intervals may be a way to do that. For example, I could feed 3 different Stochastics figures from three different 5-minute bars (9:34 to 9:39, 9:35 to 9:40, 9:36 to 9:41) and then deal with the results.
Does anyone have thoughts on that? Thank you!
p
Comment