[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: Swinging...



PureBytes Links

Trading Reference Links



> -----Original Message-----
> From: Ian Waugh [mailto:ianwaugh@xxxxxxxxx]

> I'm trying to define a swing point, let's say a high, but it's not like 
> a "standard" high that has a lower high on either side, I'd like to 
> define a Swing high like this:
> 
> A bar makes a higher high. The next bar can have the same high or a 
> lower high but not, of course, a higher high. Then take the lowest low 
> of these two bars. Now, in order to actually qualify that first bar as a 
> swing high, we need a bar that *closes* lower than either of the first 
> two bars.
> 
> The minimum number of bars required to define a high is, therefore, 
> three. However, theoretically, there's no limit to the maximum number of 
> bars although in practise it's never reached the limit (:-).
> 
> Having defined a swing high, we then look for a swing low which is 
> defined in exactly the same way but, er, the other way around. You 
> cannot have two SHs or two SLs in a row, the order must be SH, SL, SH, 
> SL, etc.


I did something like this recently for a client.  My solution to this
was (for Pivot High, same thing for Pivot Low):

1. When you've found the Pivot High Candidate:
    a. set a flag, bPHSetup, to true.
    b. Set a counter, iPHSetupLag to 0.
    c. Save lowest low of the two bars into iPHLowest.

2. Increment iPHSetupLag.  If iPHSetupLag > <some value> then reset
bPHSetup to false.  This sets a limit on number of bars you'll seek
to confirm a candidate.

3. If bPHSetup and C < iPHLowest then you've confirmed the Pivot.

I think that will do it for you.

Cheers,
Brendan
---
Brendan B. Boerner
brendan@xxxxxxxxxxxxxx
Karakhorum Ventures, Inc.
www.Karakhorum.com