Actually, no, this post is excellent, it's exactly WRONG. Backwards.
It be easily demonstrated even with Excel (maybe even BoWo could try instead of making useless sheets that show how he can pyramid his system) that longer-term (weekly - monthly) returns are significantly less leptokurtic than shorter term returns. Intraday returns have much more "abnormal" distributions and also show significantly more instances of being non-random. Lastly, leptokurtosis is absolutely NOT associated with trending behavior.
This is the problem when people don't do the work, don't understand the math (which is not difficult), and just parrot stuff they have heard elsewhere.
It be easily demonstrated even with Excel (maybe even BoWo could try instead of making useless sheets that show how he can pyramid his system) that longer-term (weekly - monthly) returns are significantly less leptokurtic than shorter term returns. Intraday returns have much more "abnormal" distributions and also show significantly more instances of being non-random. Lastly, leptokurtosis is absolutely NOT associated with trending behavior.
This is the problem when people don't do the work, don't understand the math (which is not difficult), and just parrot stuff they have heard elsewhere.
Quote from SarahNGuyen:
little reminder (back to initial topic - didn't read the entiiiire thread), intraday trends are actually very (very) close to random when it comes to data distribution.
Weekly and longer timeframes display "abnormal" distribution (leptokurtosis), clearly showing trends.
And yes, those (long term) trends can be used to make money.
