Re: Conventional Indicators Do NOT Work. I have the proof right here.
Lawyerkill
Steve wrote:
> Let me put something to rest: the question of whether typical technical
> analysis indicators work to forecast the future or not.
>
> I will prove, vividly, without a shadow of doubt, right here, that all
> conventional indicators do not work as forecast tools. Ok.
> 1,5,8,5,4,8,1,7,6,3,5,4, What comes next?
>
> What comes next is anything. Each number has precisely the same chance
> of turning up next. So no matter what common indicator is used, it is
> completely (100%) useless.
[snip]
LK
I agree with you, that's why I design my own indicators.
Lawyerkill
PS My Indicators said this would me a down day, and we are up 51
points, go figure.
Bill Reid
Steve <philh...@aol.com> wrote in message
news:1155233199.5...@p79g2000cwp.googlegroups.com...
> Let me put something to rest: the question of whether typical technical
> analysis indicators work to forecast the future or not.
>
Braise Jebus, we are saved!
> I will prove, vividly, without a shadow of doubt, right here, that all
> conventional indicators do not work as forecast tools.
Let me put on my thinking cap so I can grasp the intricacy of your
"proof"...
> Ok.
> 1,5,8,5,4,8,1,7,6,3,5,4, What comes next?
>
9, because it didn't show up before so it must be "due".
> What comes next is anything.
DAMN!!! My "Vegas System" crapped out again...
> Each number has precisely the same chance
> of turning up next. So no matter what common indicator is used, it is
> completely (100%) useless.
>
Maybe I didn't need that thinking cap after all...
>
> So there you go. Anyone labeling himself as an indicator expert or guru
> is really a fool of indicators.
>
Indicators have made many a man a fool, just like women and booze...
> Also, I've run extensive computer programs to backtest dozens of
> conventional indicators over the years to confirm that typical T.A.
> indicators indeed do not work.
>
Please describe these programs and show the statistical results.
> I've alpha-listed some of the indicators I've tested to be complete
> failures below. This is just the short list:
>
> Alligator indicator
> Accumulation / Distribution
> ADX
> BOI (Breakout Index)
> Bollinger bands
> Candle stick formations
> CCI, CCI Avg
> Chaikin Oscillator or any oscillator based indicators
> Coppock Indicator
> Cycle tracking
> DMI
> Elliott wave counting
> Fibonacci based indicators
> Gap indicators
> GANN indicators
> Linear Regression
> MACD
> OBV and all volume based indicators
> ROC
> RSI
> SMI
> STIX
> Stochastic indicators
> StochRSI
> Trendline based indicators
> TRIN (ARMS Index)
> Volatility
> Williams %R
> .. and this list goes on and on.
>
What about the Reverse Triple Salchow?
> Don't mess with any conventional TA indicator. You'll be wasting
> precious time.
>
Kind of like reading this group and finding some guy has posted the
same message eight times with the subject "Conventional Indicators
Do NOT Work. I have the proof right here."
> Stay clean. Life is too short to be belittled with meaningless
> trivialities. Keep all forms of noise out of your life.
>
You are a fluke of the universe. You have no right to be here.
> Hope this will save you months of your time (and life).
>
Where were you 25 years ago when I really needed you?
Of course, I can't help but notice that after stating that WE can't
predict stock prices, you then claim in other posts that (surprise!)
YOU can, using "arcs" the way only YOU can do.
Feel free to tell us exactly how to use "arcs" to predict the market,
so that we will not continue to flog ourselves using "conventional
indicators". We'll leave the light on in Usenet for your response.
In the meantime, I welcome you to the group. You fit right in, and
I shall think of you as "Steve, The 'Conventional Indicator' De-Bunker,
Multiple-Message Poster, Master Of 'Arcs'"...
> Good luck,
That's MY system!
---
William Ernest Reid
Post count: 418
Bill Reid
>
> Bill Reid wrote:
> > Of course, I can't help but notice that after stating that WE can't
> > predict stock prices, you then claim in other posts that (surprise!)
> > YOU can, using "arcs" the way only YOU can do.
>
> (who's we?)
> take your meds AFTER reading this next time
>
"WE" is everybody who does not know the secret of using
"arcs", at least one of which is I.
I'll have you know I'm totally drug-free and am also alchohol-free
since...well, since last night, I have been doing a little bit more than
usual recreational drinking lately, but if you'll just go ahead and
answer this:
> > Feel free to tell us exactly how to use "arcs" to predict the market,
> > so that we will not continue to flog ourselves using "conventional
> > indicators". We'll leave the light on in Usenet for your response.
I promise that I'll be all sobered up by the time I read it. Anxiously
awaiting...
---
William Ernest Reid
Post count: 419
LuckyDucky
> When I correctly locate the trendlines and fit the arc, I can forecast
> a top or bottom precisely (to the bar). When I make a miniscule error,
> the forecast can be 100% wrong.
And of course you prolly didn't make any of those 'miniscule errors' when
you did your extensive testing of TA , which spurted forth your 'proof' it
doesn't work, eh.
Ok bwana ,so next step is... toss us some forecast... ' to the bar' as you
say.
and how about posting an example pic of your ' arcs' , so we can be sure
you're not just rehashing Gann Lines or Fibo-time series.
sarp
Steve wrote:
> Let me put something to rest: the question of whether typical technical
> analysis indicators work to forecast the future or not.
> Good luck,
> Steve
A guy named Michael Parness wrote "Rule the Freakin Markets."
He turned $33,000 into $7 million.
He recommends these two books:
"Technical Analysis of Stock Trends," by Robt. D. Edwards and John F.
McGee
and
"Beyond Candlesticks," by Steve Nison
---- sarp
Bill Reid
> > Bill Reid wrote:
>
> > > > Feel free to tell us exactly how to use "arcs" to predict the
market,
> > > > so that we will not continue to flog ourselves using "conventional
> > > > indicators". We'll leave the light on in Usenet for your response.
>
> nothing in return to disclose what I know is one of the most
> revolutionary forecast tools thus far, perhaps the only forecast tool
> that will ever work and of which took me years to get to my current
> stage.
>
OK, I really only need a shove in the right direction, I have a veritable
arsenal of computerized tools available to research any strategy.
> Ok, "arcs" are parabolic shapes you find in all forms of (random) data.
Uh, hold the fort, right there. It is ABSOLUTELY IMPOSSIBLE to
predict any future data point in a "random" time series based on existing
time series patterns of any sort. Or rather, you can predict it, but your
prediction will be no better than "random" itself, and thus useless.
I thought that was the nature of your "proof" that "conventional" indicators
do not work. It IS a fact.
Now, if you think you can usefully predict future time series data points,
the time series is not truly "random". You CAN'T have it both ways.
> You'll notice them
> immediately when you look at a chart. For instance, pull out a
> longrange Nasdaq chart in year bars. You'll notice the chart forms a
> parabolic top in 2000 .. on both sides of this top.
>
Sure. But is this a "random" formation or not? (Personally, I absolutely
do not believe it is "random", and I DO believe I can "predict" the market.)
> Ok, the process: First you have to build a single arc that fits all
> data.
What is "all data"? Every market average? Every individual US stock?
Every individual stock in the world and every market average in the world?
> That is the first mountain (not hill of beans) .. took me many
> hours of trial and error.
Once programmed, my computer could do it in about an hour or so,
depending on exactly what "all data" means...
> Some arcs will worked here and there, they
> are wrong.
>
Mmmmmm, so your "arcs" work 100% of the time?
> Once you have the magic arc (which I have), the next mountain to climb
> is to practice placing it for a perfect fit.
This is actually pretty simple for me. I can find all the stocks or
averages
with any specified type of chart pattern fractal in a relatively short
amount
of computer run time. Been doing it for years.
And that's all you're talking about, really...just a particular type of
chart pattern, like a "cup and handle", "rounded bottom", etc.
> Arcs are TRENDLINE based
> so you have to first find the three TLs that comprise the price
> channel. The center TL is where you place the arc.
Uh, yes, and if there is just too much variation in the channel
(particularly
a certain type of noise) my software rejects a particular candidate time
series for a chart pattern match...
> Then you have to
> angle the price to a certain angle (I won't say the degrees).
>
"Angle the price to a certain angle"? Whaaaaaa?
> Then comes the biggest hurdle of all .. to make sure you haven't made
> the slightest error when forecasting. This is the stage I'm currently
> at.
>
Everybody makes mistakes. That's why I like to let the computer do
most of the work, at least they make consistent mistakes...
> The arc I'm referring to is a "time arc" and only reveals WHEN the next
> top or bottom is, not the magnitude. However I'm sure I will create a
> "magnitude arc." Right now, I cheat and simply reverse the time arc to
> look backwards to determine the magnitude.
>
Mmmmmm, strangely, I do a certain amount of this already, been
doing it for years.
One of the things I like are "matching/non-matching arcs", where I'm
comparing the dynamics over time of presumably linked time series,
such as corporate earnings, interest rates, et. al., and stock prices.
I'm looking for the areas outside the curve and fitting the curves
as markers for outsized predictable market moves.
> When I correctly locate the trendlines and fit the arc, I can forecast
> a top or bottom precisely (to the bar) .. sometimes many bars into the
> future. When I make a miniscule error however, the entire forecast can
> be 100% wrong.
That sounds a little like "chaos theory", "multiplication of errors". It
also sounds a little like "BS"...
> It takes hours of practice. I'm sure there's a much
> simpler approach. Right now this is what I'm doing.
>
Well, good luck with that...
> Another hint: price channels rarely start from the last top or bottom.
>
<taking notes>
> Tradestation has automatic arcs, but they do not work because TL is
> very primative and won't allow users to manipulate the price enough. TS
> is where I first got the idea of arcs from. I believe the French
> programmer (Pierre something) TS still works for TS ie (the Cruz
> brothers). I think he made the arcs for TS. I talked to Pierre, a
> bright guy. He also has a website. You can find other arc sources too.
> Scientists have studied these parabolic shapes in data for decades. You
> can find sites online of such studies from the 1920s and older. I'm
> just a rung on an old ladder, but one near the top.
>
Nothing new under the sun. You might even call "arcs" "conventional"
if you wanted to, since they are just another "chart pattern"...
---
William Ernest Reid
Post count: 422
Blash
> It is ABSOLUTELY IMPOSSIBLE to
> predict any future data point in a "random" time series based on existing
> time series patterns of any sort. Or rather, you can predict it, but your
> prediction will be no better than "random" itself, and thus useless.
Isn't this true of ANY "system"????
BrunoR
> One more thing, PRICE CHANNEL TRENDLINES do work much of the time.
>
> Actually there is a solution to this dilemna without using arcs. Can
> anyone figure it out? > Good luck, Steve
Did anyone see "this dilemma" or any other dilemma? If trend lines work
much
of the time where is the problem? Some stocks will overshoot in any
direction,
they just don't know about arcs or they don't care about indicators.
JMHO! ;-)
BrunoR
> I'll only disclose hints to get you or anyone going since I'm getting
> nothing in return to disclose what I know is one of the most
> revolutionary forecast tools thus far, perhaps the only forecast tool
> that will ever work and of which took me years to get to my current
> stage.
In one of Shakespeare's plays there is one line that fits this paragraph.
Do you care to guess which line would apply? The hints are obvious.
Bill Reid
Blash <bla...@comcast.net> wrote in message
news:C103C9FC.3F462%bla...@comcast.net...
But not all "systems" are "random". And even a random system
can be predicted, just not by looking at the "time series".
For example, you could have a pipeline with varying amounts
of water flowing through it. The amount of water flowing is determined
by which one of dozens of valves is being held down in a control room
a mile away. A monkey is in the room, and got into some whiskey,
and is staggering back and forth drunkenly, hitting the valves at "random".
You could sit at a point a mile down the pipeline and measure the
amount of water every second, perform all sorts of furious "calculations"
to analyze this data, and never be able to predict better than random
chance how much water will be coming at any point in the future.
On the other hand, if you were sitting there looking at a video feed
from a camera in the control room, you could predict this "random"
time series PERFECTLY.
Now stock prices are not perfectly "random" judged by certain
tests of "randomness" that mathematicians use, but they're pretty
close. And stock prices are not set purely at "random", but rather
more by what is called "chaos" by mathematicians, which unfortunately,
for practical purposes, are kind of close to random for the purposes
of predicting future prices based on the time series of past prices.
It's kind of like the weather: you can't perfectly predict EXACTLY
what the weather will be EXACTLY where you'll be standing tomorrow,
but the good news is, you can FORECAST well enough to know
whether you should carry an umbrella. The classic confusion in the
stock market arises because people are so unable to deal with an
irreducible level of uncertainty that they wind up carrying umbrellas
in the summer and walking around in shorts in the winter, "BECAUSE
NOBODY CAN PREDICT THE WEATHER, RIGHT??!!!???!!!"
---
William Ernest Reid
Post count: 423
Blash
> And stock prices are not set purely at "random", but rather
> more by what is called "chaos" by mathematicians, which unfortunately,
> for practical purposes, are kind of close to random for the purposes
> of predicting future prices based on the time series of past prices.
If EVERY system in use(charts, T/A, etc.) say XYZ should move up and
1(ONE) fund with a huge position has a clerk with a BAD hangover........His
nose hits the sell button when he nods off during the day.....
That is what is called "chaos" by mathematicians, chart readers, T/A
geeks, etc. They all think that someone knows something that they don't and
start selling too......
Herd instinct just overpowered the computers!!!.......How do you factor
that into a "system"??? There are just too many variables involved in stock
pricing to ever say "THIS ALWAYS WORKS"......
LuckyDucky
BrunoR <Bru...@nyc.rr.com> wrote in message
news:CTHDg.19094$u05....@news-wrt-01.rdc-nyc.rr.com...
" The fault, dear Bruno, lays not in our stars, but in our posters, that
they are underlings "
BrunoR
> they are underlings " or, the ramblings of a (bored!) fool. ;-)
Spydertrader
> I'll only disclose hints to get you or anyone going since I'm getting
> nothing in return to disclose what I know is one of the most
> revolutionary forecast tools thus far, perhaps the only forecast tool
> that will ever work and of which took me years to get to my current
> stage.
Once. Just once. I'd like to open Xnews and read ....
"I basically stumbled across this rather boring tool which took me all
of seven seconds to figure out. I plan on giving the tool away because
as a trader, I'd prefer to make money trading instead of selling
signals."
Alas, no such luck today.
- Spydertrader
P.S. Interestingly, I named an indicator sequence after you - in
reference to your discourse with one Jack Hershey years ago. "The 'Bruno
R' Set Up" actually performs quite well.
Bill Reid
> Bill Reid wrote:
> > > > It is ABSOLUTELY IMPOSSIBLE to
> > > > predict any future data point in a "random" time series based on
> > existing
> > > > time series patterns of any sort. Or rather, you can predict it,
but
> > your
> > > > prediction will be no better than "random" itself, and thus useless.
> > >
> > > Isn't this true of ANY "system"????
> > Well, yeah, it's a tautology: random is as random does...
> >
> > But not all "systems" are "random". And even a random system
> > can be predicted, just not by looking at the "time series".
>
> First thanks for the extensive response, but where did you find your
> false premise about random data?
Uhh, well, I read a book, several in fact, stayed awake in school,
you know, that sort of thing...
> I'm telling you right now, I can
> forecast random events 100% accurate, provided the arc is overlaid
> right ..
This is nonsense. If you can predict future time series data points
from past time series data points (with 100% accuracy, no less!), the
time series is not "random", but "deterministic".
You clearly don't understand the words you are typing on your
keyboard, or rather don't have a clue as to the underlying concepts.
You use the word "random" to mean anything that varies, even things
that vary in a totally predictable fashion. That's not what the word
"random" refers to.
> this can take ten minutes in itself and if both the data and
> arc is not placed 100% right, the forecast will be off .. often
> bigtime. One day I'll write code to make this maneuver automatic.
>
I already have the code. In order for you to make these claims
of 100% predictability, you should have undertaken computerized
back-testing of your "system". Since you now admit that you
have not, I think it's safe to assume that it is not 100% guaranteed
that you can predict the stock market with 100% accuracy.
>
> > Now stock prices are not perfectly "random"
>
> I say b.s., but neither me or you can prove it ..
Nah, mathematicians, who came up with the concept of "random",
have three tests of the raw data to determine if it can possibly be
called "random".
But again, what's funny is that the data can kind of pass these three
tests, and still not be "random" at all, and be 100% predictable! There
was an embarassing situation a few years back in some gambling machines
in Canada. Like all gambling machines, the software in the machines
used an algorithm called a "PRNG", or "PSEUDO Random Number
Generator", to generate the game results.
Now although the numbers that come out of a PRNG are all "mixed-up"
and seemingly vary at "random", they are still the result of a totally
DETERMINISTIC mathematical formula. You actually keep getting
the same numbers over and over again, although it may be a million
numbers before the sequence starts repeating itself (this is called
the "period" of the generator; the longer the "period", the more
"random" the numbers).
Now the usual drill when programming a game machine is to
occasionally "re-seed" the PRNG , or start the sequence over again
starting with a new number, based on some external, hopefully
asynchronous event. But the idiots who programmed these
machines forgot to do that, and some sharpies figured that
out, and since they knew the PRNG algorithm they could
predict the sequence of numbers EXACTLY, so they won
several million "loonies" before the casino caught on to the
problem.
Now you are telling us you've discovered the "PRNG" that
determines stock market prices...well, you're not the first, and
you won't be the last, but excuse us if we remain just a tad
skeptical...
> unless one of us are
> market makers manipulating stock prices. But this doesn't concern me
> because even if there's manipulation, my arcs still work.
Except of course, when they don't.
As I said, you wouldn't be the first to say that: "my system works
100% of the time, except for the theivin' market makers, and effin
e-Trade, too, the bastards!"
> As long as
> the "insiders" don't read my forecasts first though lol.
>
What'd I tell ya?
> > It's kind of like the weather
>
> Nah, unless you are a market maker "insider," price movement is pretty
> much random and 100% unpredictable unlike the weather.
Huh? You just said the stock market was 100% predictable, now
you are saying it is 100% unpredictable. Make up your mind.
> However, you are
> right in a way, because if you have good forecast tools, you can
> forecast both quite accurately.
>
Note very, Very, VERY carefully that weathermen don't try to
predict how many rain-drops will fall on you head based on an
"arc" of the rain-drops that have previously fallen on your head.
Rather, in the last few decades, they rely on satellites that image
storm systems as they move across the oceans, coupled with many
Earth-bound instruments of several types, and giant honkin'
computers that crunch gigantic amounts of all this data, and
they STILL get it wrong many times...but they get it a LOT better
than they used to.
This should be the real lesson about the value of trying to make
TRULY valuable predictions based on price movements alone, but
it's just so damn easy to sit there and see all kinds of pretty "patterns"
in the data, and blame the "thievin' market-makers" when your
"picnic" gets "rained out"...the point is to always try to "do better"
even if it is "really hard"...
---
William Ernest Reid
Post count: 424
BrunoR
Hey Spydertrader, that first paragraph is NOT my quote! I had answered
to that
paragraph instead. ;-). Jack Hershey, is he still posting somewhere?
Not that I
want to read it; For laughing out loud. It's much too complicated for
my brain.
Spydertrader
> Spydertrader wrote:
>> BrunoR wrote in misc.invest.stocks:
>>
>>> I'll only disclose hints to get you or anyone going since I'm
>>> getting nothing in return to disclose what I know is one of the most
>>> revolutionary forecast tools thus far, perhaps the only forecast
>>> tool that will ever work and of which took me years to get to my
>>> current stage.
>>
>> Once. Just once. I'd like to open Xnews and read ....
>>
>> "I basically stumbled across this rather boring tool which took me
>> all of seven seconds to figure out. I plan on giving the tool away
>> because as a trader, I'd prefer to make money trading instead of
>> selling signals."
>>
>> Alas, no such luck today.
>>
>> - Spydertrader
>>
>> P.S. Interestingly, I named an indicator sequence after you - in
>> reference to your discourse with one Jack Hershey years ago. "The
>> 'Bruno R' Set Up" actually performs quite well.
>>
>> http://tinyurl.com/m2fsx
>
> Hey Spydertrader, that first paragraph is NOT my quote! I had
> answered to that paragraph instead. ;-).
I apologize. I may have snipped too much off the post and attributed the
post to you rather than the reply.
> Jack Hershey, is he still posting somewhere?
> Not that I want to read it; For laughing out loud.
> It's much too complicated for my brain.
Yes, Jack still posts on occasion over at the Elitetrader.com web site.
He currently uses the handle Grob109, but has posted in the past as Jack
Hershey and Bubba7. I translated most of Jack's posts into an easier to
read (and follow) format. Now, if only I can convince Jack of the many
benefits of brevity, we should be all set. :)
The indicator sequence which bears your name continues to provide
profits. Each day someone posts their list of "'Bruno R' Set Ups" for
review. Most recently, GROW and NEU each signaled their run higher in
advance by showing a 'Bruno R.'
The Indicator Sequence even has made its way into the Futures Trading
Discussions. Jack even calls the set up a 'Bruno R' now. The last few
pages of my Journal (linked below) have several references to the 'Bruno
R' Set up. Had I realized sooner you continued to post on USENET, I
would have pointed out your ongoing fame much sooner.
A pleasure to have made your (electronic) acquaintance.
http://www.elitetrader.com/vb/showthread.php?s=&threadid=56555
Good trading to you.
- Spydertrader
Bill Reid
Spydertrader <sp...@billis.com> wrote in message
news:Xns981FAC7098A53sp...@216.196.97.142...
one, I want one!
What's his lousy indicator anyway, some type of moving average
crossover? Big deal! Jack Hershey blathers about what is essentially
just volume reversal, he gets an indicator named after him, some
guy says, "I buy when the stock crosses over the XX-day moving
average", he gets an indicator named after him. What do I have
to do to get one named after me?!!??!!!
Look, I've said several times one of my favorite setups involves
a breakout with volume from a low-level long base in a stock that
is seriously undervalued. Can't you name that setup after me?
Or better yet, call it:
"The WilBilMobile Rocket Ship To Da Moon" indicator!!!!
Thanks for your consideration...
---
William Ernest Reid
Post count: 426
Spydertrader
> Hey, no fair! I want an indicator named after me! If he can have
> one, I want one!
>
> What's his lousy indicator anyway, some type of moving average
> crossover? Big deal! Jack Hershey blathers about what is essentially
> just volume reversal, he gets an indicator named after him, some
> guy says, "I buy when the stock crosses over the XX-day moving
> average", he gets an indicator named after him. What do I have
> to do to get one named after me?!!??!!!
>
> Look, I've said several times one of my favorite setups involves
> a breakout with volume from a low-level long base in a stock that
> is seriously undervalued. Can't you name that setup after me?
> Or better yet, call it:
>
> "The WilBilMobile Rocket Ship To Da Moon" indicator!!!!
>
> Thanks for your consideration...
Nice to see you again William, but No 'moving average crossover'
indicators here. Sorry Bill. Perhaps, I wasn't clear enough though. The
'Bruno R' Set Up isn't an indicator. A 'Bruno R' stock simply shows the
indicators (in addition to price and volume) "setting up" in a certain
way. This "set up" often provides a clue to a change in trend. As such,
the trade signal provides significant profit. I named the sequence after
Bruno R after reading the following USENET (M.I.T.) post from 2002.
http://mailgate.supereva.com/misc/misc.invest.technical/msg15804.html
Unfortunately, we already have a 'Rocket' Set up, but perhaps, we could
have the "WilBilMobile Space Shuttle to the Van Van Allen Radiation
Belts" Indicator once an available slot opens in the future.
Good Trading to you.
- Spydertrader
OpEd
>
> Bill Reid wrote:
>> > > It is ABSOLUTELY IMPOSSIBLE to
>> > > predict any future data point in a "random" time series based on
>> existing
>> > > time series patterns of any sort. Or rather, you can predict it, but
>> your
>> > > prediction will be no better than "random" itself, and thus useless.
>> >
>> > Isn't this true of ANY "system"????
>> Well, yeah, it's a tautology: random is as random does...
>>
>> But not all "systems" are "random". And even a random system
>> can be predicted, just not by looking at the "time series".
>
> forecast random events 100% accurate, provided the arc is overlaid
> arc is not placed 100% right, the forecast will be off .. often
> bigtime. One day I'll write code to make this maneuver automatic.
So, Steve, can you overlay your arc on the Virginia pick three daily
lottery? I'd love to know what number is going to come up tomorrow. Heck,
it really is random (or as random as ping pong balls can make it) so, since
you can predict it with 100% accuracy, we can make some big bucks!
OpEd