Fundamental analysts often state that technical analysis is never profitable, regardless of how many technicians show otherwise. It creates an interesting situation, where ultimately you can make technical trading decisions without any fundamental information, but you cannot make a fundamental investing decision without using technical information.

If a stock's fundamentals state that the assets are higher than what the stock is currently trading at, then it is considered to be a potentially good investment. If a stock's fundamentals have shown to produce a 20% annual growth, then it is another potentially good investment. Either way, it's all dependent upon a stocks current price or movement. Both are technical information.

Most long-term investors spend many hours going over annual reports, research reports, news stories, broker recommendation lists, and a myriad of other information to see which stocks are potential candidates. The time spent on such activities can be equal to a part-time job!

This article will show two similar methods for stock selection that can be used to limit the number of securities to consider for long-term investment purposes. The basic premise for each is identical, but the indicators used are different so as to pose alternatives and to suggest ideas for customization. The examples will be shown using the MetaStock formula language, but most other software programs with search abilities will be able to accomplish the task.

The Premise

Ideally, we want to look at a stock that has shown an INCREASE of growth over the last year, and is making new 52-week highs. But what defines an increase of growth? Our definition should be prices with a steady upward movement, and has shown its most significant upward movements recently.

The First Method

Our first system uses a Rate-of-Change analysis. The Rate-of-Change function returns the increase or decrease in price over a predetermined amount of time. This can be displayed in point movement or as a percentage. Assuming that we are looking at one year of history (which is logical for long-term analysis), we want the most recent 3-month period to have had more growth than the previous 3-month period.

Note that we are looking at the growth as a percentage, not as points. Looking at point movement can return false results. A point move from 40 to 60 would be the same as a point move from 60 to 80 - 20 points. This would appear to be consistent growth. But in reality a point move from 40 to 60 is a 50% increase while a move from 60 to 80 is only a 33.3% increase. The growth is less when moving 20 points from 60 to 80.

The following formula, written in the MetaStock formula language, searches for consistent growth in stocks that are at the highest high value for the previous year.

The best way to explain the function is to view the MetaStock code and explain each part. This formula would be used in an exploration filter.

ROC(C,63,%)>=Ref(ROC(C,63,%),-63) AND (Ref(ROC(C,63,%),-63)*2)>=Ref(ROC(C,126,%),-126) AND Ref(ROC(C,126,%),-126)>0 AND H>Ref(HHV(H,251),-1)

The first line, ROC(C,63,%)>=Ref(ROC(C,63,%),-63), is interpreted as "the current 3-month percentage Rate-of-Change must be greater than or equal to the 3-month percentage Rate-of-Change of 3-months ago." In other words, the current 3 months of growth must be the same or greater than the previous 3 months of growth.

The second line, (Ref(ROC(C,63,%),-63)*2)>=Ref(ROC(C,126,%),-126), is interpreted as "two times the 3-month percentage Rate-of-Change of 3 months ago must be greater than or equal to the 6-month percentage Rate-of-Change of 6-months ago." We double the 3-month period to extrapolate what a 6-month Rate-of-Change would be (we want to simulate the same number of periods being analyzed).

This way, the groups of data being looked at would look like this:

|----------6 months----------|---3 months---|---3 months---|

to equal 1-year of data.

The third line, Ref(ROC(C,126,%),-126)>0, requires that the first six months of data being looked at are positive growth. Without this, we may be looking at stocks that are slowing down on their downward moves, instead of growing in value.

The last line, H>Ref(HHV(H,251),-1), requires the last price available to be the highest high for the previous year.

The last line can be changed to look for stocks that have made a new high in the last week by using this line in its place:

Alert(H>Ref(HHV(H,251),-1),5)

Using these formulas, we have a fairly robust screening method.

The Second Method

The second method is virtually identical to the first, except we are using a statistical method known as a Linear Regression Slope instead of a Rate-of-Change. We use the Linear Regression Slope to analyze all of the data to determine the movement instead of just comparing the current point value to a point value of x-number of days ago. This formula would be used in an exploration filter.

LinRegSlope(C,63)>=Ref(LinRegSlope(C,63),-63) AND (Ref(LinRegSlope(C,63),-63)*2)>=Ref(LinRegSlope(C,126),-126) AND Ref(LinRegSlope(C,126),-126)>0 AND (C>Ref(C,-63) AND Ref(C,-63)>Ref(C,-126) AND Ref(C,-126)>Ref(C,-252)) AND H>Ref(HHV(H,251),-1)

The only real difference between this formula and the previous one is the fourth line: (C>Ref(C,-63) AND Ref (C,-63)>Ref(C,-126) AND Ref(C,-126)>Ref(C,-252)).

Since the Linear Regression function doesn't isolate the starting value and the ending value, we need to add a function for this purpose. We want the close today to be higher than the close of 3 months ago, the close of 3 months ago to be higher than the close of 6 months ago, and the close of 6 months ago to be higher than the close of 1 year ago. This way, we determine that the prices are increasing while the Linear Regression Slope assures us that the trend direction is also increasing.

You can modify either of these filters to fit your trading style. You can add in a line like:

AND C>20

so that it only shows stocks that are over $20 in price (if using U.S. data).

Short-term Trading

The short-term technical trader is not left out in the cold with these types of formulas. All you need do is change the periods that are being examined. A shorter-term look can be taken by changing the values of 63 and 126 to 16 and 31 respectively, giving you a 3-month analysis instead of a 1-year analysis. You could also replace the numbers with 5 and 10 respectively to return a 1-month analysis. The fundamentalists can even use these shorter-term results as great hop-on-board candidates for research.

Conclusion

Initially, the results from this type of exploration show that an incredible amount of growth has already happened from most of the stocks returned, but these types of starting patterns are consistent with some of the greatest long-term investment successes in history.

While we don't have any direct buy or sell signals here, we do have something that tells us what stocks are showing significant increases in growth. An exploration of this nature can give you a great list to look at for fundamental investments and research, rather than looking through many reports to find the same types of companies.