|RS vs. S&P500 Exploration by Jim and Kerry Thorley
Re: How to do a large scale technical analysis with MS?
* To: metastock@xxxxxxxxxxxxx
* Subject: Re: How to do a large scale technical analysis with MS?
* From: Jim and Kerry Thorley <muffie53@xxxxxxxxxxxxxx>
* Date: Fri, 28 Sep 2001 08:31:27 +1000
* Reply-To: metastock@xxxxxxxxxxxxx
* Sender: owner-metastock@xxxxxxxxxxxxx
> Since I started with mechanical investing first, and then bought MS, I was *very* surprised that it is not possible to analyse hundreds or even thousands of stocks in comparison to each other.
> For example:
> Take all NASDAQ stocks, => about 6000
> select those 10% of the stocks with the best ROC(C,252,%), => 600
> take those with the best LinRegSlope(log(C),126) minus 2*projected Stdev, => 60
> take those that had a HHV(H,10) => ?
> check those for some trailing buy stops to determine if they are a buy...
> Take your sell rules...
> Take both and do a system-test for the latest 5 years, to evaluate this approach.
> Take above mentioned variables (10%, ROC-lookback, etc.) to do an optimization.
> Perhaps this is possible, but I cannot see???
I thought any one with MetaStock would appreciate further details of a technique that could be applied to ANY equity investment. I have personally found that using this can result in some very rewarding investments.
We took at look at equity investments that are different than both common stocks and mutual funds - SPYDRS and WEBS. Your chairman, using the well-proved technique of testing relative strength of those equities against the S&P 500, picked out the strongest performing of both that group and various sectors. I used the various Fidelity sector funds for sector analysis. A printed out list of those best performers was passed out to the group. We reviewed the MetaStock Professional Version 7.0's explorer function in general and more specifically demonstrated how to use explorer to pick sectors, SPYDRS, WEBS, and stocks that have a better than average chance of increasing in value.
I stated "... the well-proved technique of testing relative strength of those equities against the S&P 500 ...", because over a long period of time of testing various indicators and trading techniques, I have found that relative strength as defined above has proved to be the best of ANY indicators and/or trading techniques that I have found for picking investments that have the greatest chance of yielding a good profit.
NOTE - THIS IS NOT THE RELATIVE STRENGTH INDICATOR THAT IS BUILT IN TO METASTOCK!! - that indicator simply indicates a comparative relative strength of a security vs. ITSELF at various times.
The MetaStock Explorer settings follow, followed by an elaboration of both the formulas and technique.
MetaStock for Windows
RS vs. S&P500
Select price line on S&P500 chart and run exploration
RS vs. S&P500 Exploration
C / P
Column B: -5 days
Ref(C , -5) / Ref(P , -5)
Column C: -30 days
Ref(C , -30) / Ref(P , -30)
Column D: 5-0 Days
(C -Ref(C , -5)) / (P-Ref(P , -5))
Column E: 30-5Days
((Ref(C,-30))-(Ref(C,-5))) / ((Ref(P,-30))-(Ref(P, -5)))
Column F: Tot5,30
((C-Ref(C,-5)) / (P-Ref(P,-5)))+(((Ref(C,-30)-Ref(C,-5))) /
Filter Enabled: Yes
In the above exploration, I wished to find equities that have shown a high degree of relative strength over both the last 5 days and the proceeding 25 days of trading. (It was interesting to find that at the time the "best" one was the Sweden WEB symbol EWD on the American Exchange.)
You start the exploration by opening the chart of the index or other that you want to use as the standard to compare the other equities against, in this case the S&P 500. You then left click on the price plot of that chart. This "chooses" it and causes small black squares to appear along that plot.
You then run the exploration by choosing the explorer icon (a pair of binoculars) and choosing the exploration, then the explore button. I will leave it to you to pick which securities you wish to test with this exploration.
Explanation of the formulas: (All assume that daily plotted data is used, an assume that the latest data is loaded.)
Column A: --- C / P
C is the last date available close of those securities which you are testing. P refers to "a special data array identifier that is used to reference any indicator or price plot" - In this case, this refers to the last date available close of the "chosen" S&P 500. The calculated figure is the relative strength of the tested security vs. the S&P 500.
Column B: -5 days --- Ref(C , -5) / Ref(P , -5)
Ref(C , -5) = The close of those securities which you are testing as of 5 days ago.
Ref(P , -5) = The close of the S&P500 as of 5 days ago.
This calculates the relative strength of the tested security vs. the S&P 500 as of 5 days ago.
Column C: -30 days --- Ref(C , -30) / Ref(P , -30)
In a similar manner to the Column B, this calculates the relative strength of the tested security vs. the S&P 500 as of 30 days ago.
Column D: 5-0 Days --- (C -Ref(C , -5)) / (P-Ref(P , -5))
(C -Ref(C , -5)) = The difference between the current close of those securities which you are testing currently minus that as of 5 days ago.
(P-Ref(P , -5)) = The difference between the current close of the S&P500 minus that as of 5 days ago.
The calculation results in the slope of the relative strength of the security being tested between 5 days ago and the current relative strength. The figures for relative strength cannot be compared between securities because they have different prices. However the slope of that line CAN be directly compared to see which one is the strongest. A steeper slope (a larger value) IS a comparatively stronger performer compared with the S&P 500. NOTE - THIS IS TRUE IN BOTH BULL AND BEAR MARKETS!!
Column E: 30-5Days --- ((Ref(C,-30))-(Ref(C,-5))) / ((Ref(P,-30))-(Ref(P, -5)))
Similar to Column D, This calculation results in the slope of the relative strength of the security being tested between 30 days ago and the relative strength as of 5 days ago.
Column F: Tot5,30 ---- ((C-Ref(C,-5)) / (P-Ref(P,-5)))+(((Ref(C,-30)-Ref(C,-5))) / (Ref(P,-30)-Ref(P,-5)))
This function has no real physical/mathematical meaning, but is simply a total of the values of Columns D and E. If the slopes of both are high, it follows that the relative strength is good both in the recent past and the slightly earlier recent past.
In the report, I sorted results using this column specifying descending results. This placed the strongest securities at the top of the report.
FILTER SOURCE I use no filter on this data. If you are only interested in strong performers, you might include a function as follows:
((C-Ref(C,-5)) / (P-Ref(P,-5)))+(((Ref(C,-30)-Ref(C,-5))) / (Ref(P,-30)-Ref(P,-5))) > 0
This will eliminate those that have a negative slope in Column F. Useful if you have many securities to test and wish to make the report smaller. Personally, I include all values so that I might know which securities and/or sectors that should be avoided.
The data in columns A, B, & C are actually superficial to the exploration and could be eliminated. I used these primarily to illustrate the concept of relative strength.
I used the time frame of the last 5 days, and the preceding 25 days to pick up those that are showing CURRENT strong relative strength. I then checked the results of the exploration and made composite charts of promising securities by dividing their close by the close of the S&P500. I they examined both these composite charts and the charts of the specified security for further information before deciding on the purchase.
You can compare securities with ANY other chart by use of the same technique. I normally compare securities with both the S&P 500 and the NASDAQ composite. I have 2 separate explorations (with identical formulas), one named RS vs. S&P500, and one named RS vs. NASDAQ. This way, I can create and keep separate reports on each comparison.
I hope this helps you make good investments. Happy investing
| Source / From: