|
PureBytes Links
Trading Reference Links
|
I have read and applied Bill Williams "Chaos" techniques (ahem, beware that
the "Chaos" and "Fractals" are used somewhat freer than a strict their
mathematical definition).
I do find his stuff interesting, and it does appear to have both merit and
potential, although I personally was not able to make it work for my
intraday trading style on the e-mini. That doesn't mean it is not workable.
For myself, I chose not to become enamored with an esoteric pursuit and
returned to the trading task at hand. However, I learned enough to "tuck
this one away" and play with it another day. I do not use it in my regular
trading.
I did find several of his concepts useful:
* His one and two bar interpretation of trend
* Metrics for identifying market squat conditions (and turn tendency)
* The "Awesome Oscillator" (AO) for wave count
* Rules for trading market structure (waves)
I used his trend definition to build a new type of trend indicator. I like
how it reacts quickly without lag, and is more in tune with price movement
than typical trend indicators. I also like have a non-parameterized
indicator ... no tweaking. Like price, it just "is".
In the book he claims that all turns happen within 1-2 bars of a squat
condition (higher volume and dropping MFI). My analysis shows he is
correct. Unfortunately, the appearance of a squat could also mean a pause
before the market resumes so the squat is not sufficient. I believe he says
as much. I find squat on tick bars is more effective than minute bars,
which is how I end up using it.
In my book, the AO is pretty amazing. Essentially he reworked the MACD and
gave it a different interpretation and usage. The AO indicator is for
finding wave count, NOT for market entry or exit. Here I find his analysis
spot on. I found the AO most accurate on tick bars, like 100, 200, or 400.
What I can say is that this is a very strange phenomena, since I could
observe wave count all the way down to 25 ticks (smaller becomes difficult
to interpret). So that "random breakout" at 6:43 is never random at all: it
ALWAYS falls within the wave count (like, I never found a counter-example
in several years of intraday data). Truly very, very weird.
Of course, there are downsides. The first is that even with AO it is far
too easy to get lost in real time (yes, I know there are proprietary tools
that do similar things). The second is that the AO formation will read best
in one time-frame ... and then that time frame changes. The reading when
"off" is difficult. The changing, tracking, and "self-similar" nature of
waves makes them very confusing to apply. I'm not the first to say that
waves and cycles are slippery beasts. For me, it simply took too much
attention away from the market. I find this syndrome consistent with wave
adherents ... wave study becomes the market, and not the other way around.
I'll comment on Williams' trading plan for market structure even though it
wasn't (exactly) my own. The plan accounts for wave mis-identification,
protective stops, and position scaling. However, it doesn't lend itself to
automation or easy back testing. In general, you can't count on visual back
testing because of the influence of wave patterns that have already
happened. I prefer to understand reward/risk ratios, expectancy, and many
other trading stats. Williams doesn't provide this, and the information is
hard to compile.
Like most things, I find I can walk away with some morsels, but for me this
was not a meal.
Kevin
At 03:23 PM 1/4/2004 -0800, mike ball wrote:
Has anyone that has read this book, been able to
convert the written words of Bill Williams into a
workable intraday trading system, such as what he has
fully attempted to describe?
|