Re: Short Term High's/Low's

["Sean Taylor" <sean.taylor@xxxxxxxx>]

Dave,
Thanks for the HHV suggestion.

Allan,
Using Stoch() also sounds good.

Both worth a try.

However, what I was trying to do was implement Larry Williams definition of
a short term high/low.
Take Short term low.  A new low is defined as having higher lows on each
side. But, you ignore inside days!!!
Quote "An inside day means the market has entered congestion, the current
swing did not go further, but then again it did not reverse, thus until this
condition is resolved, we must wait and not use the inside day in our
identification process." - identification of a short term swing point..

I'd like to try to code this, if nothing else, display on an expert.

Thanks,
Sean


----- Original Message -----
From: "David DeFina" <ddefina@xxxxxxxxxxxx>
To: <metastock@xxxxxxxxxxxxx>
Sent: Wednesday, December 13, 2000 6:20 PM
Subject: RE: Short Term High's/Low's


> Using the HHV function you could code hhv(h,7) which would give you the
> highest high for the last 7 days, including today, or ref(hhv(h,7),-1) for
> the highest high excluding today.  And likewise the LLV function for
lowest
> low.  I don't know if this is what you're trying to accomplish, but this
is
> a good formula for identifying support/resistance points and for
determining
> stop placement.
>
> Dave D.
> ddefina@xxxxxxxxxxxx
>
> -----Original Message-----
> From: owner-metastock@xxxxxxxxxxxxx
[mailto:owner-metastock@xxxxxxxxxxxxx]On
> Behalf Of Sean Taylor
> Sent: Dec 13, 2000 1:45 PM
> To: Metastock List
> Subject: Short Term High's/Low's
>
> Hi,
>
> Does anybody have any suggestions as to how I might code an expert to
> identify short term high's and low's.
> This is using the simple premise that inside days are ignored.  i.e.
> previous non-inside day H is lower than current H which is higher than the
> next non-inside day H (for High's - reverse for Lows).
>
> At a guess, I need to use PREV to count the number of past consecutive
> inside() days to get to the previous valid bar.  Not sure how to identify
> the next valid bar or how to hang the lot together.
>
> Perhaps I'm completely wrong in my approach and there is a simple route.
>
> Thanks,
> Sean
>
>