|
PureBytes Links
Trading Reference Links
|
Of course, system testing is a confirmer of results that would have been
obtained in the past. Running the test without stops and see what the system
will do with no restraints will bring about "trust".
Every so often I see people using optimization, I see little value in
distorting the results by using this feature. It might be useful to get in
the ballpark for period adjustments in research but has no place in a active
system.
Real time charts will give quick confirmations of ideas, but here it is rare
to find software to test what the results would be over time. I like to use
5min charts. If I use my favorite 89 period MA that is 7.4 hours, the 1/2
cycle is 44.5 or 3.7 hours. In an active, volatile, stock you have a trade a
day on average. Only daytraders can benefit from the small % moves.
Richard Estes
-----Original Message-----
From: Steven Buss <sbuss@xxxxxxxxxxx>
To: Metastock-list <metastock-list@xxxxxxxxxxxxx>
Date: Saturday, December 13, 1997 1:36 PM
Subject: The Most Valuable Thing about Real Time Quotes...a Question...and
the Market...
>I thought it might be that I could make more money if I had real time
>quotes. But the most valuable thing about getting real time quotes is not
>that I can make more money (although I'm finding out that I can given my
>"trading personality").
>
>The most valuable thing for me about getting real time quotes has to do
with
>something Mr. Estes has said a few times recently. That one can build a
set
>of indicators (a "system" if you will) and then LEARN TO TRUST IT.
>
>I'm finding that LEARNING TO TRUST the indicators takes less time with real
>time quotes than with end of day data. Watching 1 minute bars flash by
>accompanied by simple 1, 5, and 60 minute default parm (5,3,3) stochastics
>is interesting because it provides the basis for an intensive and
>interactive learning experience. One is prompted to guess where the next
>bar will be every minute. And you find out 1 minute later if you got it
>right. And you notice things about how the indicator values change in more
>subtle ways when the next bar doesn't go the way you thought it would. In
>short, I think it's possible to shrink the time that it takes to make the
>indicator your own. This isn't a "technical" issue really; it has to do
>with what the psychology of learning is about...(i.e., increasing the
number
>of instances that I must "trust" to find out if the indicator(s) steer me
>wrong or not provides a broader and deeper psychological basis for the
>"trust".)
>
>I remember reading a chapter in one of Schwager's Market Wizard books about
>a guy who spent months and years perusing charts to ascertain the "price
>patterns" and indicators that might help him to become a profitable trader.
>(This was before the days of the software we now have available.) I
>understand that story now. This guy made the price significant patterns
and
>indicators "his own" by living with them day and night.
>
>* * *
>So, my question is this: Are there other things that relative newcomers
can
>do to shrink this "learning to trust" the indicators interval?
>* * *
>
>Regarding the market, the daily default parm stochastics (5,3,3) are all or
>mostly below 20 but the slope on the line is still negative. The weekly
>stochastics on the Dow and S&Ps are still above 80 but have turned down.
>Weekly stochastics on the Nasdaq 100 and the Nasdaq Composite haven't been
>above 80 since mid-August and the first of October respectively. For these
>latter two indices, the stochastic crossed below the MA this last week.
>
>Significant also is the change in the look of the default parm weekly MACD.
>For all the indices mentioned above the MACD-Difference has decreased when
>it had earlier been increasing.
>
>Steven Buss
>Walnut Creek, CA
>sbuss@xxxxxxxxxxx
>
>
|