The Stock Selection Section 4 

What we’re going to cover

In this section, we’re going to look at Section 4 of the Stock Selection Guide.  Section 4 is pretty much the target—figuring out if the stock is fairly priced. All the work in sections 1 through 3 converge into part 4 of the Guide.  We want to make sure that the company is likely to double in value in the next five years.  This criterion usually forms part of a screen that most members of BetterInvesting apply when deciding if they want to purchase a company at the current price.

Section 4 of the Stock Selection Guide

Section four of the Stock Selection Guide is an evaluation of risk and reward.  It has five subsections:

A.   Determination of a five-year future high price

B.   Determination of a five-year future low price

C.   Figuring out where today’s price is in relation to the projected high and low prices in A and B

D.   Comparison of the potential gain of investing with a loss

E.   Looking at projected price appreciation and figuring out the expected overall growth.

We’ll handle each of these sections in turn.

A. Determining a five-year high price

This part is pretty straightforward.  Most of the heavy lifting is done already!

We have calculated two things that we’re going to need: a high PE ratio (we found that in section 3) and a growth rate for the earnings per share—calculated in section 1 (or section 2 with the preferred method).  We are going to use the growth in EPS to calculate what the earnings could look like in five years.  Here’s how.

First of all we have a growth rate, an EPS growth rate, which was calculated in the last chapter.  We’re going to use the EPS growth rate based upon the preferred method—that rate is 17.4%.  For completeness, let’s see what value we get for the EPS if the projected value of 15% were used.

We know what the earnings per share were for the last fiscal year—they’re in Value Line!  The reported EPS was $1.06—look up at the section 3 of the SSG or refer to the Value Line survey for Clayton Homes.  When we completed section 1 in Chapter 1, the EPS growth rate was projected to be 15% annually.  A slower rate than observed historically—a conservative move.  So, we projected that the current EPS of $1.06 would grow at 15% per year for the next five years.  In other words, at any given point in time, the earnings per share twelve months in the future will increase by 15%.  Let’s run through what that means.

The EPS in 1999 is $1.06; therefore, the EPS in 2000 should be 15% greater.  To calculate 15% of a number, you multiply that number by 0.15, so

0.15 x $1.06 = $0.159 ~ $0.16

The EPS will grow, approximately, an additional $0.16.  In 2000, we expect the EPS to be equal to the EPS for 1999 plus $0.16.

EPS₂₀₀₀ = $1.06 + $0.16 = $1.22

There is a ‘2000’ subscript on the EPS to show to which year I am referring.  Keep in mind, we’re not trying to predict EPS for the year 2000.  We’re tying to estimate earnings in year 2004; we calculate 2000 as a stepping stone along the way.  Great, we’ve calculated the EPS for 2000, how about 2001?  Same deal—just increase the EPS for year 2000 by 15%.  Easy.

0.15 x $1.22 = $0.183 ~ $0.18

We project that the EPS will grow by 18¢.  In other words, the EPS in 2001 will be eighteen cents higher than those in 2000, or

EPS₂₀₀₁ = $1.22 + $0.18 = $1.40

We’re on a roll.  Increase the EPS for year 2001 by 15% and you get the EPS for 2002

0.15 x $1.40 = $0.21

The EPS increases by 21¢, to the EPS for 2002 becomes

EPS₂₀₀₂ = $1.40 + $0.21 = $1.61

If you repeat this process, you will find that the EPS for 2003 will be $1.85; the EPS in 2004, the year that interests us, is projected to be $2.13.

We’ve just calculated the projected EPS using the traditional method.  Earlier, the EPS was calculated using the preferred method, the result was a higher value of $2.37.  The preferred method gives a higher EPS for the year 2004 than was obtained using a simple projection based on a 15% annual growth.  In fact, as was mentioned previously, the preferred method projects earnings to grow at 17.4% annually.

We’ll use the results obtained from the preferred method in this analysis; however, it is always useful to show the ‘traditional method.’  The preferred method appears to be getting more popular among NAIC people.  I am sure part of the reason is increased teaching by chapters; while a major reason is the popularity of programs such as Toolkit, which make these once difficult calculations so easy.

One of the advantages of using five years is that earnings might not rise in a nice straight line.  Things happen.  Companies take charges.  However, using a five-year horizon helps to average out some of these year-to-year effects.  The most important thing is that sales keep on growing.  If sales start to falter, it’s time for a major investigation—it could be a natural hiccup—but it is time to investigate what’s going on.

So, using the preferred method, we determined that the EPS in 2004 will be $2.37.  That’s the number we’ll use.

Just one final comment on the EPS calculation.  Very often, by the time you get Value Line, there may well be a quarter or more of information for the next fiscal year.  In the Clayton Homes example, quarters for fiscal year 2000 are being reported.  Keep in mind that it is best to stick to a fiscal year approach.  If you decided to use EPS for the most recent four quarters, you would then be projecting into the future for those four quarters—not the fiscal year.  Whatever you choose to do, be consistent.

What we are going to do is to combine a PE ratio and an EPS to get a price.  Since price divided by EPS gives the PE ratio, the PE ratio multiplied by the EPS gives a price.  If we know what the highest EPS is likely to be, and we multiply that high EPS by the highest PE ratio we expect, we should come up with a high price—that’s the thinking.  That’s what we’re going to do, but more about the high PE.

In section 3, we determined that the average high PE for the past five years was 19.8 while the average low was 11.4.  The method use gives a flavor of the PE range, but keep in mind there’s no way to predict the future with 100% accuracy.  Ultimately, we want an answer that will guide us in the proper direction.  So approximations here and there are not a huge concern.

I noted earlier that there was a trend in the PE ratio.  I mentioned too that PE ratios tend to move in cycles.  When interest rates are low, or people are bullish (very upbeat and positive) about the prospects for the market, PE ratios tend to expand.  So how do we capture this trend in PE ratios and benefit—rather than suffer—from it!  One method is to use a weighted average rather than the regular average used in section 3.

If you haven’t realized it, I am a true blue conservative when it comes to investing J.  But I firmly believe in pleasant surprises rather than nasty shocks.  So, when I have a choice, I select the more conservative route.

When it comes to PE ratios, when I see a trend, I tend to do the opposite of most other people.  The trend must be clear; it must be discernible.  Now, if you look at section 3 of the SSG, you will see that the PE ratio is slowly falling.  I will weight more recent results more heavily.  Here’s how it’s done.

The most recent PE is given the heaviest weight; the size of the weight decreases on going from most to least recent.  The weight is multiplied by the PE to give a composite.  The composites are added up.  Then the sum of the composites is divided by the sum of the weights to give the weighted average.

The PE using the straight average was 19.8; the weighted average, since it puts more emphasis on recent numbers, results in a lower PE, 18.4.  If a similar procedure is carried out for the low PE, the result is

Again, this method gives a lower PE (10.8) than the straight average.  This is the one I’ll use in the analysis that follows.

When the PE trend is the other way, highest most recent and lowest five years ago, I do something a little surprising.  If the expanding PE is part of bull market—market indices such as the S&P 500 and the Dow are moving up—I weight again, but put more emphasis on earlier results.  In examples above, if the PE’s were lowest in 1995 and highest in 1999, I would weight 1995 the highest and 1999 the lowest!  It’s a conservative stance and tends to be punishing towards fast growing companies.  More on this point later.  If the ROE and PTP are also expanding, then this behavior is a growth company in growth mode!  I tend to use a standard average in this case.

There is a criticism against using weighted PE’s.  The issue is that the weighting of the PE either pulls both the high and low down or pushes them both up.  It’s like getting a pay raise to match the cost of living—nothing really changes.  Some people are inclined to weight the high PE and take a straight average for the low—sounds conservative to me!

So, we have a high PE and a projected EPS.  The EPS was determined using the preferred procedure, ($2.37) and the high PE was determined using a weighted average.  The high PE is 18.4.

Since the PE is the price divided by the earnings per share, multiplying the earnings per share by the PE gives a price!  That’s exactly what has happened above.  The projected high price is 43.6.  The thinking is that we have a range for PE and we believe that the EPS will grow steadily (on the average) for five years.  If the PE range is valid and the EPS is reached in year five, the high price should be reached at some time in year five.  Hence, the price is projected to reach $43.6.

For completeness, let me show you what would have happened had I used the average high PE (19.8) and the EPS for 2004 based on an annual growth rate of 15% ($2.13).  Had I used this method, the projected high stock price would be $42.2.  This projection is not a whole lot different than the projected price in the last paragraph.  The important message is that the G in SSG is a guide.  It gets you in the neighborhood, and that really completes your journey!

The high price we will use in the analysis that follows later is $43.6

B. Determining a five-year low price

You’ve probably noticed that while there is only one method for projecting the high price, there are four for the low price.  A lot of time was spent calculating the high price; don’t expect four times that for the low prices!

Method (a) is something you’ve seen before.  This method is similar to the calculation for the high price.  This time we take the low EPS and multiply it by the low stock price.  Since we accept that we have a growth company, we anticipate that the EPS will grow year after year.  If this assumption is correct, then the most recently reported fiscal year should be the lowest EPS observed in the next five years.  Note, the most recent fiscal year result is used rather than the sum of the past four quarters.  The low price for a fiscal year is being projected.  Since the lowest EPS is that for the most recent fiscal year, the number $1.06 from 1999 is taken.

When the average high and low PE’s were calculated, these were the averages observed in the past five years.  The assumption used in this analysis is that this observation is representative of the behavior of the PE ratio; it will occur again in the future.  Hence, if the low PE were to hit today with the current EPS, we’d get the low price for the next five years.  So

11.4 x $1.06 = $12.1

This projected low price is based on the average PE; if the weighted average is used—it was calculated in the last section, then the calculation is

10.8 x $1.06 = $11.4

We have an opportunity for a ‘sanity check.’  We have a current PE of 8.0 based on the last four quarters of earnings.  We also calculated a projected PE ratio of 6.8.  Since the current PE and the projected PE are both lower than the projected low PE, there is good reason to use our future estimate.  This PE of 6.8 projects a very low price

6.84 x $1.06 = $7.25

Once again, since projecting lower is conservative, that’s what I will choose to do.  Judgement has been applied here.  (It’s a weak example.)  But the sanity check used was to look at today’s PE and to compare it to the average low PE.  The low price chosen with this method is $7.25.

Method (b) is very straightforward.  It’s the average of the low price of the past five years.  This price was calculated in the previous chapter.  Here’s the calculation again.

The sum of the past five years is $45.8; dividing by 5 gives an average of $9.16, which is rounded to $9.2.

Interestingly, the low price showed a trend until this year—it declined after four straight years of an increase.  For growth companies, the low price often tends to increase year after year.

There is probably not a lot of value in calculating a weighted average here.  This method is meant to create another option for a low price—a real average.

Method (c) involves looking at the past two or three years and selecting a low price.  In this case, the low price was observed in 1999: $8.3.  This number is called the ‘recent sever market low.’  It is generally not that beneficial to go back more than three or so years.  Perhaps if the past three years had very similar prices, then you might search further back for a lower price.

Methods (b) and (c) are complementary in some respects.  They are really designed for a cyclical company: a company that follows the business cycle.  If there is a downturn in the economy, the GDP is declining, you would consider using method (c).

During a recession, recent low prices are often revisited.  “Investors,” (in this context, I say this word with scorn) tend to steer clear of the market—they don’t invest—and prices tend to be suppressed.  Interest rates are often low during a recession—the Fed tries to get the economy moving, but it takes a time for people to start buying back into the market.  Using method (c) as the economy slows down (particularly for cyclical companies) gets you ready for surprises.

If the economy is moving out of recession, method (b) could be used as an estimate for a low price.  Recessions rarely extend beyond two quarters, so averaging in five years of information covers the business cycle.

Traditionally, a recession occurred every five or so years.  So plotting ten years of information in section 1 almost always covered at least one recession—you got a chance to see if you are studying a real growth company.  Section 3 should touch on a recession too.  However recently, the economy has been in a sustained expansion, with no sign of a slow down.  This fact doesn’t challenge the validity of the approach here, but it does give a context as to why some of the approaches in the SSG are used.

Finally, method (d) is geared for those who invest for income (dividends) primarily.  In section 3, we calculated the dividend yield for the past five years.  If you recall from the last section, the dividend yield is the dividend divided by the stock price.  It’s the return that an investor expects to receive from dividends alone.

The rational here is that if all else fails, people will eventually be attracted to the stock for the dividend alone.  In this sense, the dividend will act as a brake on a price drop.  If you don’t believe it, let’s look at an extreme.  It would be unbelievable if the price dropped to a nickel!  It sure would!  The reason is that the dividend is six cents J.

So, if the dividends act as a brake on a price drop, we can use the dividend yield to determine when that brake has triggered in the past.  Since a high dividend yield occurs when there’s a low price, the high dividend yield during the past five years should be determined.  This calculation is already carried out in column H of section 3.  The highest dividend yield, 0.7%, occurred in 1999.

If we know two things in this relationship the third can be calculated.  We have a value for a dividend yield (remember, we choose the high dividend yield because that occurs when the stock price is lowest.)  We also know the dividend.  The recent dividend is $0.06.  The dividend divided by the dividend yield gives a stock price.

Since we’re doing a mathematical calculation, the dividend yield must be express as a decimal rather than as a percentage.

  The dividend yield is, therefore, 0.007.

This method gives a low price of $8.60.  So, the four methods give the following estimates for a low price:

(a)          $7.25

(b)          $9.5

(c)          $8.3

(d)          $8.6

A copy of an SSG for this section can be found a few pages back.  The work is not yet completed.  The big challenge lies in actually selecting the low price.  One thing that is clear: the temptation to average these four results should be resisted.  There is a clear rational for each of the four methods.  They are, to some extent, not compatible one with the other.  The first method is for growth stocks; the middle two—(b) and (c)—are more suited to cyclicals; the final method is for those focusing on income as the prime criterion for selection.

The low price selection method that most NAIC members find themselves using is method (a).  This method is appropriate for Clayton Homes.  Clayton Homes is a growth company, as revenues grow steadily regardless of the overall economic environment.  When analyzing a growth company—a company without a cyclical character in the revenue stream—method (a) is appropriate.

The low price of $7.25 is selected of the analysis in section 4.

A word on rapid growth companies

Rapid growth companies can create a problem.  After all, if the stock is growing by 30 to 50% a year, and the low price is moving quickly too, using method (a) can appear to be too conservative.  The full ramifications of this issue will be apparent when we zone.

Now keep in mind that method (a) for a growth company helps to minimize your risk.  Using the method I am about to discuss adds risk.  The method is based on the fact that if a company is growing rapid, the stock price is expected to track the earnings.  Hence the high and low stock price should both increase from year to year.  If you have five years of history, you can look back and see if this behavior is observed.

Next look at the recent stock price for the past three or so club meetings—that is, the past three months.  Take the average and then discount (reduce) this price by either 20% or the sustained EPS growth rate, whichever is larger.  Note, the earnings per share should be growing steadily.  It should be a well behaved growth as evidenced by a reasonably straight line in the SSG.

Using Clayton homes as an example, the stock price for the past three months (in this example) was 9.00, 9.56 and 8.44.  The average is

The average turns out to be $9.

Discounting this figure by 20%—that is, reducing the number by 20% gives $7.2.  (9.00 – 0.2 x 9.00 = 7.20).  If the EPS growth rate were 32%, then I would reduce the recent average price by 32% and use the resulting number as a low.  For Clayton Homes, the discount was the greater of the growth rate (17.4%) or 20%.  Since 20% is larger, that was used as the discount factor.

This method is designed for rapidly growing companies where there is clear evidence that the growth has been sustained.  But keep in mind the following—rapid growth is unsustainable.  Rapid growth companies get severely punished when they cannot deliver on (unreasonable!) expectations.  So, be prepared for a rocky time.  Keep in mind that this method does introduce some additional risk into your analysis.