|
PureBytes Links
Trading Reference Links
|
The first 1- 1.5 hours.
The last 1.5 - 2.5 hours (the closing of US being a key factor).
Definitely NOT the "noon balloon": 12- 2 Eastern.
Among other things, the timeliness factor that you mention was one of the
variables in a program marketed by a fellow who used to be in the NY/NJ area by
the name of Moore, who, as I recall, divided the trading day into something
like 45 minute segments and then had various rules for trading each of those
different segments.
"Shraga(Feivi-Philip)" wrote:
> Hi everyone,
>
> I was testing some S&P-mini daytrading systems, "tweaking" here and there,
> when an idea popped into my head. I added a line to the code which would
> only let the system trade for two hours a day, and tested different 2-hour
> periods. What I got is really interesting. For one year, the last two hours
> of the session gave 90 procent of the profits. Then I checked the year
> before, and those two hours didn`t give good results. However, the FIRST
> two hours, improved the results for that year by over 20%. Now, what this
> tells me is that certain time periods (year?) have very specific trending
> "hours". If so, I wonder about the possibility of simply studying several
> months, and trying to identify the recent "hours" which seem to trend more,
> and only trade during those hours. Problem is, you can`t really test and
> verify such an approach. Has anyone else noticed such a tendency for the
> S&P market to trend more during certain hours for an extended period?
>
> Philip
|