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

Re: Gann Swing concepts (Fibonacci trader) in Metastock? #2



PureBytes Links

Trading Reference Links



Thomas Winklhofer wrote:

>
> Thank you Ken, for sharing your work. I will look into your Indicators and
> System tests. If I understood your idea correctly, you a buying dips. That is
> not the approach of Krausz. In addition to that I would like to say that these
> 'short term' or 'minor' Retracements are very dangerous to buy, because the do
> not hold. IMHO retracements with time scales of  >1-2 months represent good
> buying/selling signals on the 50 and 61.8 level (golden ratio).

Thanks for pointing that out.  I'm not endorsing wholesale buying of fib
retracements. That's plain crazy.  Seldom, if ever will I buy a .382 retrace on a
daily chart, but occasionally I will buy retracements at the 50 or .618 level
provided there is a confluence of indicators to suggest that it's a low risk
opportunity, such as previous sup/res, key moving averages, a strong oversold
indication, price action, voices from beyond, etc.  The simpler the indicators the
better.  Actually Fibs, as far as entries go, generally play a minor role in my
trading even though there is some validity to them.  I do take notice.  You
mentioned time scales.  FWIW, if you want to wait at least 1 month before buying,
that can be added to the test that I sent.    Just add the condition to the most
recent peak BarNumber.  That is, tell it not to buy unless a retracement occurs >=
the most recent BarNumber of Peaks + 20 or something similar.


> Interesting would be - Automatic recognition of valleys/peaks of swing channels
> and fibo
> lines 38.2 and 61.8
>  (parallel to channels, that's not possible with metastock).

As I said, I have not read the article, but when you said recognition of 38.2 and
61.8 fibo lines, I thought of my indicators/tests and simply thought that you might
find it useful and would be able to adapt it to your needs.  If not, toss it!  :-)


> n Metastock the ZigZag Indicator would be the appropriate one programming
> plotting swings.
> But the function parameter do not fit. Problems:
> - DATA ARRAY can take only O,H,L OR C. We need to plot from High to Low.
> - A minimum change in $ or % is not a suitable condition. We would need a free
> programmable condition

The minimum change simply defines the size of the swing from point A to point B.
Some people feel that a 1% swing is appropriate.  If it seems too minor of a swing,
change it to a higher value.  Since a Zig Zag indicator of the Highs alone cannot
define a swing from price high to price low, you need to use two Zig Zags, one that
uses the highs and one that uses the lows.  The test that I sent does this, and it
uses the correct chronological order to do the calcs for entries.  Of course
Metastock's System Tester can only calculate entries from the O,H,L, and C, but
that's another story.  Zig Zag can only calculate price swings that meet certain
minimum criteria.

>Kraus identifies up/down Swings by two consecutive
>'higher highs'/'lower lows'
>after a Low/High.

If you are basing swings on the price difference of a pattern formed by the highs
(and conversely the lows), you should be able to program MS to plot the value when a
high pattern occurs, when a low pattern occurs, and then derive the length of the
swing from that, and using that value, figure any fib levels needed, and not use Zig
Zag.  Since I have not read the article, much if not all that I have said may be
irrelevant, so it seems silly for me to comment further.
As I said at the beginning of this e-mail, you mentioned fib retracements and I sent
what I had, but I certainly can't offer much beyond that without reading the
article.  Hopefully by the time I get around to reading the article you will have
the solution. and share it with us.  :-)

Regards,
Ken
--
mailto:divenfish@xxxxxxxxxxxxx