|
PureBytes Links
Trading Reference Links
|
RE: How does your testing work for the April 2000 Nasdaq drop and the
Sept-Nov
Nasdaq drop?
Tom, I tested from September 1999 to present, and it avoided both the March
and September plunges almost entirely. I believe since the stocks I tested
were on the Nasdaq, they quickly under performed the S & P index during the
plunges, which caused the moving averages to cross soon after the drop
began. It was primarily these two events this year that caused me to be
impressed by this indicator set-up.
This isn't a recent example, but I'm at the office. The top section has the
traditional moving averages (10 and 30 exponential) based on the stock
price. The bottom section uses the same moving averages, but their based on
IBM's relative strength to the S&P 500. You can see the primary difference
in February - April 1999 where the stock whipsaws up and down, causing bad
signals in the conventional moving averages. The bottom window shows the
10-day moving average coming and barely touching the 30-day moving average,
eliminating some of the noise.
I've always hated moving averages because they lose all the money you made
with them when a stock consolidates. This indicator comes closer to solving
that problem in my mind. It is also just as good or better at signaling
changes in trends.
I'll send current examples tomorrow.
ddefina@xxxxxxxxxxxx
-----Original Message-----
From: owner-metastock@xxxxxxxxxxxxx [mailto:owner-metastock@xxxxxxxxxxxxx]On
Behalf Of Tom Sprunger
Sent: Dec 28, 2000 10:35 AM
To: metastock@xxxxxxxxxxxxx
Subject: Re: TC2000 Relative Strength Moving Average
I experimented with this and found a serious limitation in my opinion.
Since its based upon the stock's price performance relative to an index, it
fails when the stock and the index are moving in the same direction at about
the same rate. I.E., if the Nasdaq is in free fall and the stock is falling
at about the same rate, this approach may show that all is o.k., when in
reality you are losing your behind.
I think it may work o.k in an exploration, but may fail applied to a
specific portfolio that you may already have under conditions like described
above..
How does your testing work for the April 2000 Nasdaq drop and the Sept-Nov
Nasdaq drop?
Tom
----- Original Message -----
From: "David DeFina" <ddefina@xxxxxxxxxxxx>
To: <metastock@xxxxxxxxxxxxx>
Sent: Thursday, December 28, 2000 10:26 AM
Subject: TC2000 Relative Strength Moving Average
> Anybody tried using TC2000's new moving average based on relative strength
> in Metastock?
>
> I developed a good indicator to identify trends in stocks, but TC2000's
idea
> of using a 10 and 30 day moving average of relative strength (vs. SP500,
> NASDAQ, etc.), works as well or better than mine. Using moving averages
> based on the stocks relative strength to a broader market average
eliminates
> market gyrations from the equation and allows the moving averages to be
> sensitive only to stock performance relative to the index. I programmed
> this into Metastock by plotting a relative strength indicator in white
> (making it invisible on white background), and created 2 moving averages
> based on this indicator (10 and 30 exponential). When the 10-day crosses
> above the 30 day, this is assumed to be bullish, and you look to be
bearish
> when it crosses back below the 30-day line. Obviously, the two moving
> averages can be any number, but these seem to work well.
>
> I tested this setup on many Nasdaq stocks by programming an expert with my
> indicators (not TC2000) and comparing the trend changes to mine visually,
> and saw that they were very similar. My system is very accurate for
finding
> trends, and entry points, so I'm assuming this TC2000 setup is as
reliable.
> I haven't found a way yet to program relative strength into a system test,
> so if anybody knows how I'd appreciate learning the secret. I used the
> SP-500 as the comparison to Nasdaq stocks because it correlates well with
> the Nasdaq, but isn't as volatile.
>
> My trading system only uses the trend as one piece and several other
> indicators for entry and exit, but I believe this indicator is as close to
> perfect for determining trends as you'll find.
>
> Dave D.
>
|