|
PureBytes Links
Trading Reference Links
|
Thanks for sharing the logic on ranking, Taforme. It's not an easy
problem to solve. It is a major limitation to MS. I'm surprised no one
has written a dll for doing that kind of data manipulation. It would
be valuable in both stock and mutual fund trading.
At the bottom of this post, I've described my method for using
momentum in a system and what the system is based on. This method is
easily tested if the historical data is available.
There are multiple ways to get the data into the MetaStock to test
ranked lists. Unfortunately, testing over a long period of time, say
20 years, is really not feasible having to load and unload the data
unless someone has a lot of free time and patience.
In another post, I mentioned using the 2nd derivative of the price
curve as a measure of momentum. That particular method can be tested
because it's essentially the slope of the polynomial that defines the
price curve over the lookback period. I calculated it using two
methods. One was a simple technique using limits. I just broke the
curve up into small pieces, took the slope of each piece and added
them back together. The second method was the use of a dll that
actually calculated the slope of the polynomial. That was a little
more accurate than the estimated method, but in the end the estimated
values worked just as well on a filtering basis.
For testing the RSC, relative strength comparative, I used SpyGlass
but I put the S&P500 index into the population and then choose only
stocks that were outperforming the index by X% as the subgroup from
the original population. It's not exactly the same as the ranking
method, but it works pretty good for this type of momentum logic.
For testing the external relative strength, it is a straight forward
test of the stocks that are outperforming other stocks by X%, ie 70%,
80%, etc.
Testing some of the other momentum formulas is a bit tricker and isn't
easily done on a long term basis. To test those, I took random time
frame samples from the population and compared the performance of the
indicators to the performance of the ones like the polynomial slope
and the ERS. I also took time frame samples from the various types of
market conditions like downtrends and uptrends. This type of testing
is vary tedious.
The URSC is another type of program for finding RSC values.
I don't use RSC programs because I've found that I can get very close
to an identical ranking using a weighted ROC formula, and using it for
ranking suffers from the drawbacks I've already mentioned.
As I have written in several articles my favorite type of trading and
the highest return I've consistently found has been achieved using the
HHV method with a filter here and there similar to TechTrader. Darvis
systems, Turtle logic, and Donchian Channel systems are all based on
the same logic. Many of the commercial trading systems are built
around some variation on the Donchian Channel.
I have found that using momentum with the logic of the HHV system
makes a huge difference in the success of the system. In my case, I
use a quantitative pre-screen done by Value Line that incorporates
fundamentals and quantitative factors with momentum into the screening
model, so I don't have to apply momentum filters to the trading
formula in MS. The IBD, S&P Platinum and Fair Value lists work well
also as does Ford Equity Research Top 20. There are a number of ways
to use pre-screening effectively in a trading system.
I imagine it is possible to design some kind of MS system that applies
a set of momentum based trading rules to cut the population down to a
group of 100 to 200 stocks. From there it would apply the HHV logic to
that group to find trades. I tested a few systems of this type, but
found the ones I've described to be more consistent and profitable so
I stopped creating and testing them.
This is my experience. May be some of the logic will work for you and
may be it won't. A lot of new traders prefer to reinvent the wheel. I
did a lot of that when I was new. I polished a few old wheels along
with reinventing some. It was a lot of work.
Have fun!
Super
--- In equismetastock@xxxxxxxxxxxxxxx, "taforme" <taforme@xxx> wrote:
>
> I offer the following comments on external relative strength and
> Metastock.
>
> Assume a group of 100 stocks. In one column of the Explorer, do a
> price ROC based on percentage for a set time period. Then sort the
> stocks by this column. This should yield an ordered list that ranks
> the 100 stocks based on price strength so that the top stock had the
> most increase (or smallest decrease) in price and the last stock had
> the lowest increase (or the highest decrease) in price.
>
> As has been suggested before, this ordered list of 100 stocks can be
> exported to a spreadsheet where each stock can be easily assigned a
> rank number from 1 to 100.
>
> The problem is then how to get this rank information back into
> Metastock so that it can be accessed for testing, etc. The
> Downloader can be used to add this data for each stock into the open
> interest field for the respective stocks; open interest data is not
> normally used for stocks. An open interest indicator is built-in to
> Metastock and so the data contained in the open interest field now
> becomes available for most purposes.
>
> Yes, this would be a time consuming chore if each of the 100 stocks
> had to be updated manually. Fortunately, it can be automated. A
> macro program, e.g. MacroExpress (I use it and get no benefit for
> mentioning it), can be used to automate this process. I have used
> MacroExpress together with the Downloader to enter a current real-
> time price update into otherwise EOD data (20 minutes old) just
> before the market close so indicators can be generated on more
> current price data. If desired the rank data could be entered each
> day so that a continuous rank value would be available for back
> testing.
>
> Of course, any data that is not internally available within Metastock
> could be inserted into the open interest field of stocks and then
> accessed for testing within Metastock.
>
Yahoo! Groups Links
<*> To visit your group on the web, go to:
http://groups.yahoo.com/group/equismetastock/
<*> Your email settings:
Individual Email | Traditional
<*> To change settings online go to:
http://groups.yahoo.com/group/equismetastock/join
(Yahoo! ID required)
<*> To change settings via email:
mailto:equismetastock-digest@xxxxxxxxxxxxxxx
mailto:equismetastock-fullfeatured@xxxxxxxxxxxxxxx
<*> To unsubscribe from this group, send an email to:
equismetastock-unsubscribe@xxxxxxxxxxxxxxx
<*> Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/
|