Take Stock's Summary for ResMed (RMD) reports that "Earnings have grown historically at only -26.54%" and that "Recent earnings growth has been only -32.00%."

(D) In contrast to Take Stock's -20.5% EPS forecast, all the analysts are estimating about +20.00% EPS for the next 5 years with S&P, FactSet, Zacks, First Call exactly at 20.00%, Reuters at 19.50%, and Value Line low at 18.00% (with 95% Earnings Predictability). Take Stock is not structured to recognize these other estimates, but the large difference of some 40% between Take Stock (-20.5%) and the other analysts (+20.0%) demonstrates that Take Stock is unrealistically low.

Suggestions:

(1) Take Stock should define and eliminate obvious outliers in its EPS assessment. Eliminating outliers is fundamental to forecasting future growth rates as including nonrecurring amounts is a faulty basis for projecting the future. Moreover, Take Stock routinely eliminates high values in its assessment of historic and forecast High and Low PEs;

(2) In addition to eliminating outliers, Take Stock should also revise its automatic choice of a forecast EPS which is now always the lower of the Business Model or the Forecast EPS Growth Mode. For RMD, the Business Model led to EPS of $3.14 (29.8%) while the Forecast Growth Mode led to EPS of $0.27 (-20.5%).

Take Stock should be revised to accept the Business Model when it is substantially different from the Forecast Growth Mode (say 25% or more) or when the Forecast Growth Mode results in a negative EPS, subject to Take Stock's existing rule that it never exceeds a Forecast EPS of greater than 20.0%;

(3) Take Stock should also revise its lingo which seems misleading. For example: Earnings have grown historically at only XX.XX% [ADD] "for the last year" and Recent earnings growth has been only YY.YY% [ADD] "for the trailing four quarters".

Armin Fields

From the The Infirmary forum at StockCentral.com, armin fields writes:

Take Stock's Summary for ResMed (RMD) reports that "Earnings have
grown historically at only -26.54%" and that "Recent earnings growth
has been only -32.00%."

(A) The -26.54% is true only for the last year which isn't
representative of RMD's history. Earnings have grown 8.7% per year for
the last 5 years (using Hemscott data) and 20.7% per year for the last 10;

(B) The -32.00% is for TTM quarterly earnings and includes an obvious
outlier that distorts the outcome. The last 4 quarters of EPS were:
$0.37, $-0.20, $0.35, $0.33, for a TTM of $0.85 EPS. The $-0.20
quarter represents a -158.40% change from the prior year and, as a
result of including that atypical quarter, the TTM change was a
distorted -32.00%;

(C) That one very poor quarter is given too much weight by Take Stock
and misrepresents its historic EPS growth rate (-26.5% for one year)
which in turn distorts its forecast EPS growth rate (-20.5% for the
next 5 years);

(D) In contrast to Take Stock's -20.5% EPS forecast, all the analysts
are estimating about +20.00% EPS for the next 5 years with S&P,
FactSet, Zacks, First Call exactly at 20.00%, Reuters at 19.50%, and
Value Line low at 18.00% (with 95% Earnings Predictability). Take
Stock is not structured to recognize these other estimates, but the
large difference of some 40% between Take Stock (-20.5%) and the other
analysts (+20.0%) demonstrates that Take Stock is unrealistically low.

Suggestions:

(1) Take Stock should define and eliminate obvious outliers in its EPS
assessment. Eliminating outliers is fundamental to forecasting future
growth rates as including nonrecurring amounts is a faulty basis for
projecting the future. Moreover, Take Stock routinely eliminates high
values in its assessment of historic and forecast High and Low PEs;

(2) In addition to eliminating outliers, Take Stock should also revise
its automatic choice of a forecast EPS which is now always the lower
of the Business Model or the Forecast EPS Growth Mode. For RMD, the
Business Model led to EPS of $3.14 (29.8%) while the Forecast Growth
Mode led to EPS of $0.27 (-20.5%).

Take Stock should be revised to accept the Business Model when it is
substantially different from the Forecast Growth Mode (say 25% or
more) or when the Forecast Growth Mode results in a negative EPS,
subject to Take Stock's existing rule that it never exceeds a Forecast
EPS of greater than 20.0%;

(3) Take Stock should also revise its lingo which seems misleading.
For example: Earnings have grown historically at only XX.XX% [ADD]
"for the last year" and Recent earnings growth has been only YY.YY%
[ADD] "for the trailing four quarters".

Ellis:

Thanks for your point-by-point response to my latest critique of Take Stock. 

Your response made it very clear that you are entirely satisfied with Take Stock, and see no reason to make any changes whatsoever, as long as everyone understands that it is for new investors.  In short: “…we feel justified and perfectly content to keep it as it is”.

Here’s my reaction:

(1) Take Stock's methodology is flawed, I think, regardless of whether or not the stock is like RMD which had one atypical quarter in the last year.  Consider:

As anyone can see, Take Stock’s Forecast EPS is way, way less than any of the analysts:

For MDT, TS estimates 2.80% compared to Value Line’s 12.50% (100% Earnings Predictability and the lowest of the 6 other estimates);

For SKY, TS estimates 8.90% compared to Value Line’s 17.00% (100% Earnings Predictability and again the lowest of the 6 other estimates;

For JNJ, TS estimates 4.63% compared to Value Line’s 8.00% (100% Earnings Predictability and VL is again the lowest, but identical to S&P and FactSet); and

For ZMH, TS estimates 4.09% compared to Value Line’s 11.00% (80% Earnings Predictability and VL is again the lowest of the 6 other estimates).

To make this even more apparent, Take Stock’s estimate is still substantially less than the Reuter’s estimate less 2 Standard Deviations for ALL four stocks.  Reducing an average by two Standard Deviations is an exceptionally rigorous standard and used more often by university researchers than by investors.

And, Take Stock’s Quality rating is also substantially less than S&P’s Quality rating for ALL four stocks.

(2) About RMD, you wrote:

"Under what circumstances would you suggest that a new investor ignore recent history, eliminate the most recent year as an outlier, and use earlier years to estimate future growth? I would never!"

Maybe you might “never” ignore one recent bad quarter, but ALL of the other 6 estimates for RMD did just that and were consistently way, way higher than Take Stock’s forecast EPS, one of the several issues I pointed out initially but which you ignored.  For RMD, TS’s forecast EPS was a mind-boggling -20.5% compared to S&P, First Call, Zacks and FactSet (all at +20.0%), Reuters (19.50% and R-2SD at 17.76%), and Value Line (again the lowest of the 6 other estimates at 18.00% with 95% Earnings Predictability).  That’s a whopping 40% difference between Take Stock’s -20.5% to the other analyst’s +20.0   Moreover, Take Stock gave RMD a 1.6 Quality rating ….unacceptable….compared to B+ from S&P/NAIC.

"I'm perfectly content to discourage the new investor from investing in RMD or stocks with a similar scenario."

CONCLUSION:

If you truly want to discourage new investors from considering RMD and similar stocks, Take Stock should explicitly and unambiguously provide such a warning….without completing any analysis and forecasting an unbelievable and untrustworthy -20.5% EPS growth rate for the next 5 years.

I think one major flaw of Take Stock is to project a 5 year growth rate that is based on one poor quarter, especially when we are taught to eliminate unrepresentative or anomalous amounts so they don’t distort projections of the future.  Another big flaw is to ignore all EPS history prior to the most recent year: you wrote, with regard to RMD, that “the years of good growth are considered irrelevant….We will always take the more conservative course.”

The Take Stock flaws I’ve identified not only produce wacky forecasts, they also teach bad habits to new investors: that it’s OK to ignore the stock’s long-term history and OK to project the next 5 years based on one, poor quarter.

Armin Fields

Armin:

At 03:04 PM 12/3/2007, you wrote:

>Your response made it very clear that you are entirely satisfied with
>Take Stock, and see no reason to make any changes whatsoever, as long
>as everyone understands that it is for new investors. In short: we
>feel justified and perfectly content to keep it as it is.

Here's my reaction....

I would hate to be the closed-minded, smugly self-satisfied person you paint me as in your post. But, I suppose my comments which you cite would have me doing it to myself! 

I'm sorry if I came off as being so defensive that I am blind to legitimate criticism.

Armin, I would be pleased to advocate improvements to the software that might make it feasible for the software to be more "sensitive" to those exceptions. However, I would not want to sacrifice any of the conservatism that would prevent a single investor from investing in a company whose last year was messed up without going through the process of challenging that year's performance and satisfying herself that she could overlook it in favor of the prior history.

As it is, we created the software to be as "fail safe" as possible for any investor who, for whatever the reason, couldn't, wouldn't, or didn't do the necessary research to determine that such a current year was an anomaly and not a harbinger of things to come. But, we also provided the tools to allow that investor who did such research to override those judgments and use her own. Unfortunately, they are not yet present on the on-line version of the software.

We teach investors to disregard the herd and think for themselves. If the herd...or even a preponderance of analysts out there...say you can ignore certain data and estimate growth more aggressively, they it's a matter of your own judgment whether or not you want to take their word for it. I would suggest you get a dose of David Dremen before you rely too heavily on the analysts' estimates.

In any case, I'm wide open to any algorithms you might suggest to allow the software to take into consideration the things you mention but; I'll not compromise on fail-safe default for the novice. (Nor would you, if you were putting your software out there for the public to use and, in this litigious world we live in, could afford no more than zero tolerance for results that might err on the less conservative side.)

ET

Hey Ellis:

Did I wear out my welcome?  You gave no point-by-point response this time and said nothing about my critique of Take Stock’s analysis of MDT, SYK, JNJ, and ZMH (312 words, 2 pages) and nothing about my conclusions (163 words, 3 paras).

You did say: “I would suggest you get a dose of David Dremen [sic] before you rely too heavily on the analysts' estimates.”

A lot of people cite Dreman, but haven’t read him.  I’m sure you have, but maybe you forgot that he did not recommend disregarding reliance on analyst estimates, even after citing a bunch of horrible stats.  And, I bet he didn’t reduce those stats by 2 standard deviations like I did.

You can look up what Dreman did recommend as I gotsta-go….on vacation for 8 days from sunny San Diego to the East coast where it’s cold and maybe snowy (I haven’t seen snow or worn long pants in ~30 years).

A

From the The Infirmary forum at StockCentral.com, armin fields writes:

Did I wear out my welcome?  You gave no point-by-point response this
time and said nothing about my critique of Take Stock’s analysis of
MDT, SYK, JNJ, and ZMH (312 words, 2 pages) and nothing about my
conclusions (163 words, 3 paras).