Stock Market Might Fall After New Year - The Author's Answer

2 views
Skip to first unread message

pro...@uclink4.berkeley.edu

unread,
Dec 31, 1995, 3:00:00 AM12/31/95
to
It does look like a lot of people have words about forecasting the financial
market movement. Good or bad comments are all very constructive, I wish
I could come up with the statistical statement as "It is in 95% confidence
interval that the market will ... ". People has proven again and again that
statistical forecast of the market tells zero information. Every thing is noise
signal. And that is why I did not use conventional statistical forecast method
in generating this forecast charts. A forecast is just a forecast, not a fact
casted in stone. To believe or to comment is one's own choice. But I assume
people have the open mind to accept different concepts and ideas.
For the ones that do not like my idea, please stop reading now. For people
that would like to explore new era of artificial intelligence and to taste its
benefit in making investment decisions, please read the following important
notes. Thanks.

People has asked me if the forecast chart is indicative of future s&p 500
movement. The chart was done in the candlesticks format. The long white
bar acts as a divider, separating history information from future prediction.
The daily chart forecasts the next five days movement, the weekly chart
forecasts the next two weeks movement, and the monthly chart forecasts
the next two months movement. The movements of these three different
time periods are not necessarily synchronized or correlated. Following is
some tips on reading these charts.

Tips on reading the chart:

1. The chart points out directions mainly. It is also trying to find flunctuations but only could
do it to a certain degree.
2. S&P 500 is highly correlated with the US t-bonds, if S&P
keeps going higher and t-bond fails to follow, it might be better to do arbitrage than
straight investment in stocks. (i.e. buy bonds and sell s&p index)
3. The three charts
represent three different time lengths. It is not necessary that they are correlated.
Computer is only a machine, it can only tell so much. There certainly are times that
computer does not know where the market is moving either. But when this embarrassing
situation happens, it usually means that the market is ready for a big move. (in the next
N days, which will be pointed out by the forecast charts in the following weeks.)
4. MOST IMPORTANTLY!!! THE FORECAST CHART IS ONLY A PRODUCT OF
ARTIFICIAL INTELLIGENCE COMPUTER PROGRAM, IT CAN BE USED AS
REFERENCE MATERIAL ONLY!!! PLEASE BASE ON YOUR EXPERIENCE AND
JUDGEMENT WHEN YOU INVEST.

Hi Investors,

The recent forecast of the s&p 500 futures contracts shows that it is
going higher again. However, I must point out that what you see is
only a neural net computation result, and your judgement should
precede the machine's decision.

By the way, if you are an arbitrager or a spread trader, you may want
to send an e-mail to the following address to receive the up-to-date
result.

To receive the arbitrage report, send mail to:

pro...@uclink4.berkeley.edu

To view the s&p 500 forecast charts (short, mid, long term trends),
go to URL:

http://www.geopages.com/WallStreet/2022/index.html

wizardland

Larry Gorman

unread,
Dec 31, 1995, 3:00:00 AM12/31/95
to
In article <4c6qds$f...@agate.berkeley.edu>,

<pro...@uclink4.berkeley.edu> wrote:
>It does look like a lot of people have words about forecasting the financial
>market movement. Good or bad comments are all very constructive, I wish
>I could come up with the statistical statement as "It is in 95% confidence
>interval that the market will ... ". People has proven again and again that
>statistical forecast of the market tells zero information.

Wrong Wrong Wrong - Rather there is a mountain of evidence that returns are
very predictable. Pick up any recent finance journal published in the last 10
years. e.g. Fama/French 1992 Journal of Finance.

Your comment that "a forcast of the market tells zero information" is
confusing cause with effect. It is information that yields the forecast not
vice versa as you state. This is the difference between conditional and
unconditonal expectations. Maybe you've heard of it - the idea (rational
expectations) jsut won the Nobel Prize in Economics (Lucas).

> Every thing is noise

What ??? Either you (1) don't understand forecasting very well or (2) you
need to understand and define noise, or (3) both (1) and (2).

This is exactly my earlier point (in previous post under the title " the
market COULD _______ ") That is, so many people are out in the guru industry
telling stories - using ad hoc models - making incorrect statements in the
face of mountains of contrary scientific findings. Not being familiar with
basic findings - which any MBA or CFA worth his salt *should* know - is
inexcusable. You seem to fit the category quite well - which is unfortunate
for both you and those who follow your advice.

>signal. And that is why I did not use conventional statistical forecast method
>in generating this forecast charts.

Based on your commentary thusfar is seems likely - at least to me - that your
avoidance, or more accuartely - your failure, to compute confidence intervals
is due to a lack of understanding of the required methodology. Don't pawn it
off on some lame excuse such as 'noise'. If it is noise then it's very easy
to define the confidence interval - once you define what you mean by noise -
which you've yet to do.

A forecast is just a forecast, not a fact
>casted in stone. To believe or to comment is one's own choice. But I assume
>people have the open mind to accept different concepts and ideas.
>For the ones that do not like my idea, please stop reading now. For people
>that would like to explore new era of artificial intelligence and to taste its
>benefit in making investment decisions, please read the following important
>notes. Thanks.
>

>wizardland

Larry Gorman

unread,
Jan 1, 1996, 3:00:00 AM1/1/96
to
In article <4c6qds$f...@agate.berkeley.edu>,
<pro...@uclink4.berkeley.edu> wrote:
>It does look like a lot of people have words about forecasting the financial
>market movement. Good or bad comments are all very constructive, I wish
>I could come up with the statistical statement as "It is in 95% confidence
>interval that the market will ... ". People has proven again and again that
>statistical forecast of the market tells zero information. Every thing is noise
>signal.

Wrong wrong wrong - you got it backwards too. Information yields forecasts
not vice versa. AND it is well known (via refereeed Journals in Finance) that
there is substantial predictability in returns. 20 years ago this was thought
to be false - today it is know to be true. If you doubt this then you should
read the literature that tests your claim. You'll find it is rejected.


Michael Walsh

unread,
Jan 2, 1996, 3:00:00 AM1/2/96
to
I'm not familar with anyone who has been able to consistently
pridict market behavior using a model (other than strong-form
efficiency) published in any financial journal. Perhaps you could
enlighten us.

I do know of a few who lost their shirts trying to promote the wisdom
they published or found published in refreeed journals. They to
chanted "wrong, wrong, wrong!" but only for a brief period.
Wake-up calls always come too late....

P.S. Anyone know how Joseph Grainville is doing?

Larry Gorman

unread,
Jan 2, 1996, 3:00:00 AM1/2/96
to
Some sources documenting predictability in returns include

Explaining the predictabilty of asset returns. Ferson and Harvey. 1993 in
Research in Finance Volume 11 pg 65-106, A book isbn 1-55938-651-7

Sources of Predictability in portfolio returns Ferson and Harvey, May-June
1991 Financial Analysts Journal pg 49-56.

Predictabilty Risk and returns in emerging markets. Harvey 1995. Review of
Financial Studies pp 773-816.

The risk and predictabilty of international equity returns. Harvey. Review of
financial studies 1993 Vol 6 pp 527-566.

Wayne Ferson is a Finance Prof at U of Washington
Campbell Harvey is a Finance Prof at Duke U.

Both write extremely clearly. Thats why I cite their papers. If you cant
follow a paper by Ferson/Harvey then its unlikely you'll be able to follow
other academic papers.

Hope this helps

Larry G

In article <4c9tbn$k...@news4.digex.net>,

Rodger Frego

unread,
Jan 6, 1996, 3:00:00 AM1/6/96
to

It is good to remind ourselves that although past performance does
not guarantee future results, it can be a very helpful guide to how a
fund manager balances risk and reward.

Sarge


Gregory L. Meadors

unread,
Jan 10, 1996, 3:00:00 AM1/10/96
to
We are still leading this 3 year contest. One more
year to go. For more info. and FREE sample Newsletter
see our website:
http://ison.com/mktsys/

===============================================
METHODOLOGY SHOWDOWN STOCKMARKT TRADING CONTEST
1/1/94 to 1/1/96
===============================================

Profit/Loss Contestant Methodology

+$289,651..... Greg Meadors......... Harmonics
+$191,326..... Glenn Neely.......... NeoWave
+$105,791..... Joe Armenio.......... Turning Points
+$ 11,731.....*George Lane.......... Stochastics
+$ 10,219.....*Ron Jaenisch......... Andrews/Babson
+$ 2,584.....*Gary Smith........... Tape Reading
+$ 1,285.....*Benjamin Crocker..... Crocker Charts
$ 0.....*Andrew Cardwell...... RSI
$ 0.....*Bob Veres............ Fibonacci

-$ 2,582.....*Gerald Appel......... MACD/Time Trend
-$ 11,011..... Stephen Porter....... Trading Bands
-$ 15,468.....*Phyliss Kahn......... Gann
-$ 17,879..... Sherman McClellan.... McClellan Oscillator
-$ 25,964.....*Don Wolanchuk........ Elliott Wave
-$ 42,061.....*John Hussman......... Fundamentals
-$ 44,543.....*Mike Drakulich....... Momentum
-$ 47,074.....*Arch Crawford........ Astrology
-$ 50,000..... Doug Sooley.......... Market Profile
-$ 50,000..... Gary Wagner.......... Candlesticks
-$ 50,000..... Harry Schiller....... Consensus
-$ 50,000..... Peter Eliades........ Cycles

* Resigned from active competition.
Starting with 21 contestants only 5 are left after 2
years of trading. This 3 year S&P trading contest is
sponsored by Traders Catalog and Resource Guide with
real time trades placed through AUDITRACK. Each
contestant started with $50,000 in simulated funding.

Market Systems
2761 Mansfield Drive
Burbank CA 91504

Voice (818) 247-1133
Fax (818) 247-5310
Email mea...@ix.netcom.com
Internet http://ison.com/mktsys/

Disclaimer:
As required by the CFTC: HYPOTHETICAL OR SIMULATED
PERFORMANCE RESULTS HAVE CERTAIN INHERENT LIMITATIONS.
UNLIKE AN ACTUAL PERFORMANCE RECORD, SIMULATED RESULTS
DO NOT REPRESENT ACTUAL TRADING. ALSO, SINCE THE
TRADES HAVE NOT ACTUALLY BEEN EXECUTED, THE RESULTS MAY
HAVE UNDER-OR-OVER COMPENSATED FOR THE IMPACT, IF ANY,
OF CERTAIN MARKET FACTORS, SUCH AS LACK OF LIQUIDITY.
NO REPRESENTATION IS BEING MADE THAT ANY ACCOUNT WILL
OR IS LIKELY TO ACHIEVE PROFITS OR LOSSES SIMILAR TO
THOSE SHOWN.

Ken Grogan (klg@iti-oh.com)

unread,
Jan 10, 1996, 3:00:00 AM1/10/96
to mea...@ix.netcom.com
mea...@ix.netcom.com (Gregory L. Meadors ) wrote:
>We are still leading this 3 year contest. One more
>year to go.
>
<snip>

>
>Profit/Loss Contestant Methodology
>
>+$289,651..... Greg Meadors......... Harmonics
>+$191,326..... Glenn Neely.......... NeoWave
>+$105,791..... Joe Armenio.......... Turning Points...

Congrats on your results so far.

Allow me to ask a few questions...

- Are you trading a single contract in the contest? If your using multiple
contracts, what is your strategy (or philosophy) for adding/subtracting
contracts?

- Glenn Neely seems to be gaining on you; Are you getting nervous yet?
(Do you have any idea what NeoWave is?)

- Any insight on why so many of the other contestants did so poorly (other
than the obvious SuperGenius vs ScumOfTheEarth angle)? Some might find it
distressing that so many disparate systems and methods seemed to fail so
misrerably.

- Did your Harmonics catch yesterday's (Jan.9) selloff?

Good Trading,

Ken

--
The opinions and views expressed here are my own and
they do not reflect the views or opinions of my employer.


Gregory L. Meadors

unread,
Jan 10, 1996, 3:00:00 AM1/10/96
to
In <30ee81ef....@news.space.net> rfr...@blitz.de (Rodger Frego)
writes:

Hi Sarge, I agree. Here are the two year results of the
Methodology Showdown Trading Contest. I am still
leading the contest with over 450% returns in 2 years.
For FREE sample Newsletter and more info. see my website at
http://ison.com/mktsys/

Greg, Market Systems Newsletter/S&P & OEX Hotline

===============================================
METHODOLOGY SHOWDOWN STOCKMARKT TRADING CONTEST

===============================================

Profit/Loss Contestant Methodology

+$289,651..... Greg Meadors......... Harmonics
+$191,326..... Glenn Neely.......... NeoWave

Ron Rowland

unread,
Jan 10, 1996, 3:00:00 AM1/10/96
to
In article <4cvb8k$r...@ixnews2.ix.netcom.com>, mea...@ix.netcom.com (Gregory L. Meadors ) says:

>Profit/Loss Contestant Methodology
>
>+$289,651..... Greg Meadors......... Harmonics
>+$191,326..... Glenn Neely.......... NeoWave
>+$105,791..... Joe Armenio.......... Turning Points


As I recall, a year ago you had a much bigger profit.
How much ficticious money did you lose in 1995?

Gregory L. Meadors

unread,
Jan 11, 1996, 3:00:00 AM1/11/96
to
You are referring to month to month posted equity highs.
Year end returns were up from last year.

BTW, I have not been trading Auditrack has agresssively has
last year since there has not been much competition. Glenn
is working higher, I might have to put on a few trades. <g>

Reply all
Reply to author
Forward
0 new messages