Rank: Advanced Member
Posts: 105
|
In the "Risk" section, I believe "Forecas EPS" (sic) should be TTM EPS. It appears, however, that latest FY end EPS is being used.
Also, in the "Price Analysis" section, there appears to be a minor calculation error for the "Price for 14.9% Total Return". For a company with no dividend, that price should be exactly 1/2 of the Potential High Price (from the "Reward" section) but the value shown is a bit less than that. I suspect the calculation is being made for an actual 14.9% return (rather than 14.87%).
Also there, the calculatioin of "Price for 25% Risk Index" isn't quite right. The first part (3 x Potential low) is what looks wrong. Using UNH as an example, the Potential low is 28.77. Multiplying by 3 gives 86.31 but the form shows 86.53 (that's off by too much to be rounding).
Other than that, it looks very much like my TSSW spreadsheet!
-Jim Thomas
|
|
|
|
|
|
Rank: Advanced Member
Posts: 126
|
Sweet, thanks for letting us know Jim. I have filed the bugs, fixes will be in next week and i'll update this thread.
|
|
|
|
|
|
|
|
I put FAST in it and it seems to have a problem. It shows the
growth rating as 3.2 or unacceptable. Then it says sales
predictability is 9.7 with 15.1% historical growth and 19.8% recent
growth. EPS predictability is 7.3 with historical growth of 16.6% and
recent growth of 22%. I don't see how that could result in a
3.2 rating. Is this a bug or am I not thinking clearly? (I'm tired so anything is possible.)
|
Nancy C. |
|
|
|
|
|
Rank: Advanced Member
Posts: 561
|
Nancy,
Take Stock is a fairly unforgiving program.
For comparison, I plugged FAST into the BetterInvesting version of Take Stock. The results pretty much are the same: Quality is 3.2 and the mood indicator is chilly.
At least the SSG and TSSW versions of the program are consistent ...
|
Joe |
|
|
|
|
|
|
|
Jim,
To answer your questions, here are the procedures of our calculation:
question 1: --- forecast EPS in risk section is determined by Min(fcstQuarterlyEPS,
TTMEPS) , and fcstQuarterlyEPS is calculated internally on the base of
fcstEPSGrowthRate, lastEPS and quarter number. An example in take$tock
4 is IBI: TTMEPS is 1.27, lastEPS is 1.11, the risk section EPS is
1.17, which is fcstQuarterlyEPS.
Question 2: --- the calculation of totalReturnPrice is based on potentialHighPrice, currentYield and period at the target return of 14.9%. For example, HUN has no dividened, it's potential high is 30.07, it is totalReturnPrice is 15.02, which is not the exact half of potential high price.
We highly value your oppion and comment. So if you could give us your thoughts of which alternative is a better way, we'd appreciate it.
Question 3: this is fixed.
Thank you for your input and we look forward to hearing more from you.
|
|
|
|
|
|
|
|
Nancy,
If you look at it , you'd see EPS predictability is not stable enough to make it a quality stock. That is why the growth index result in 3.2, which is lower than the required 3.4.
Thanks for your input and we look forward to hearing more from you.
|
|
|
|
|
|
Rank: Advanced Member
Posts: 105
|
> Question 2: --- the calculation of totalReturnPrice is based on potentialHighPrice, currentYield and period at the target return of 14.9%. For example, HUN has no dividened, it's potential high is 30.07, it is totalReturnPrice is 15.02, which is not the exact half of potential high price.
Yes, and that's wrong. I suspect the problem is that the software is actually using 14.9% in the calculation, which is not the correct target return.
"Period" should always be exactly 5 (years). The "target return" should be that which will produce a double over 5 years (the period). The return which does that is approximately 14.87%. "15%" and "14.9%" are sometimes used in text as shorthand for that return. If your software is actually using 14.9% as the target return, it should be changed to use the correct value. 14.87% is probably close enough. The exact value is 2 ^ (1/5) - 1.
|
|
|
|
|
|
Rank: Advanced Member
Posts: 105
|
> question 1: --- forecast EPS in risk section is determined by Min(fcstQuarterlyEPS, TTMEPS) , and fcstQuarterlyEPS is calculated internally on the base of fcstEPSGrowthRate, lastEPS and quarter number. An example in take$tock 4 is IBI: TTMEPS is 1.27, lastEPS is 1.11, the risk section EPS is 1.17, which is fcstQuarterlyEPS.
Just to be clear ... we're talking about the web-based TakeStock at StockCentral.com. For IBI, that shows very different numbers than the example you quoted from the Take$tock4 software.
I wasn't aware of that use of fcstQuarterlyEPS by TakeStock. Thanks for filling me in!
|
|
|
|
|
|
|
|
Thanks Jim, that is fixed according to your suggestion.
|
|
|
|
|
|
|
|
Yes, I was refering to the desktop version when I gave the example. I meant the calculation is consistent with take$tock 4.
|
|
|
|
|
|
Rank: Advanced Member
Posts: 105
|
Yes, I see that the fixes have been made already. Great!
Don't forget to fix the spelling of "Forecast EPS" in the Risk section (the "t" is missing).
|
|
|
|
|
|
|
|
>Yes, I see that the fixes have been made already. Great!
>Don't forget to fix the spelling of "Forecast EPS" in the Risk section (the "t" is missing).
it is done, thanks.
|
|
|
|
|
|
Rank: Advanced Member
Posts: 105
|
Looking at FAST, I notice that for "Latest Dividend" (in the Reward section), Take Stock is using the 2005 annual dividend ($0.31) rather than the current annual dividend ($0.40). Is that intentional (using dividend information that's now 16 months out of date)? I understand that the FY 2006 annual data isn't available yet, but the dividend paid during 2006 has been a fact for months (since Sept 1st 2006).
Also, in the Forecast Earnings section, the Forecast Growth Mode (shouldn't that be "Model"?) shows $1.95 EPS while the Business Model shows $1.87 EPS. So, in that section, Forecast Earnings per Share is $1.87, the lower of the two. However, in the Reward section, the higher value ($1.95) is used as "Forecast High EPs" (that should be "EPS"). Why isn't $1.87 used in the Reward section?
|
|
|
|
|
|
|
|
Jim,
Looking
at FAST, I notice that for "Latest Dividend" (in the Reward section),
Take Stock is using the 2005 annual dividend ($0.31) rather than the
current annual dividend ($0.40). Is that intentional (using dividend
information that's now 16 months out of date)? I understand that the FY
2006 annual data isn't available yet, but the dividend paid during 2006
has been a fact for months (since Sept 1st 2006).
-- This is because we haven't sync up takestock online with latest ssg file yet, will do this soon
Also,
in the Forecast Earnings section, the Forecast Growth Mode (shouldn't
that be "Model"?) shows $1.95 EPS while the Business Model shows $1.87
EPS. So, in that section, Forecast Earnings per Share is $1.87, the
lower of the two. However, in the Reward section, the higher value
($1.95) is used as "Forecast High EPs" (that should be "EPS"). Why
isn't $1.87 used in the Reward section?
-- This is in conform to take$tock desktop version, I will pass this question to Ellis Traub for explaination.
thanks.
|
|
|
|
|
|
|
|
Hi Jim,
here is the point from Ellis:
The thinking was that in the novice mode, no one was familiar
with the "Business Model." Just as, in NAIC, the "Preferred Procedure"
was an "advanced concept," the business model showed nowhere in
the calculation. So, when people checked the figures and found that
the low earnings was the result of estimating earnings growth, they
would not have a question about how a different figure was used.
It was merely to avoid confusion for those who had not been introduced
to the business model or preferred procedure.
Hope this answers your question.
|
|
|
|
|
|
Rank: Advanced Member
Posts: 126
|
Jim:
In regard to:
"in the "Price Analysis" section, there appears to be a minor calculation error for the "Price for 14.9% Total Return". For a company with no dividend, that price should be exactly 1/2 of the Potential High Price (from the "Reward" section) but the value shown is a bit less than that. I suspect the calculation is being made for an actual 14.9% return (rather than 14.87%).
Also there, the calculatioin of "Price for 25% Risk Index" isn't quite right. The first part (3 x Potential low) is what looks wrong. Using UNH as an example, the Potential low is 28.77. Multiplying by 3 gives 86.31 but the form shows 86.53 (that's off by too much to be rounding)."
Both items should be calculating properly now. Please let us know if you note any other discrepancies, thanks!
Dave
|
|
|
|
|
|
|
|
In working with AMGN using web-based Take Stock I cannot figure out where average high PE of 36.0 comes from. 5 high PE's are eliminated and the remaining 5 are all under 30 so how can the average high PE be 36.0? I got out my calculator and tried eliminating 1 then 2 then 3 of the highest PE's. I never could get to 36.0. Any explanation?
Sue
|
|
|
|
|
|
Rank: Advanced Member
Posts: 561
|
The explanation is that Take Stock isn't exactly calculating the average the way that you and I would do it. And, the result is for the "Average High P/E" and the "Average Low P/E."
I figured out what it must be doing by playing with some of the options in Toolkit.
Instead of averaging the numbers, Take Stock seems to calculating MEDIAN values this way. Remember that there are 10 values for the 10 years. So, it throws out the four highest and four lowest of the High P/E numbers, leaving two. Then it calculates the average (the point between the two) of the remaining numbers. So, it's calculating the average of the two middle High P/E or average of the two middle Low P/E numbers.
|
Joe |
|
|
|
|
|
Rank: Advanced Member
Posts: 105
|
Calculating the median of all 10 PEs (rather than the average of the 5 lowest) is what the desktop version of Take Stock does when you have not marked any outliers manually. Since the web-based TakeStock does not allow you to mark outliers manually, the median-of-all-10 method is (presumably) being used in all cases (and the fact that 1/2 of the PEs have been crossed out automatically is irrelevent).
Presumably this will change once the web-based TakeStock supports "Advanced mode" so you can mark your own PE outliers.
-Jim Thomas
|
|
|
|
|
|
|
|
Yes, Jim's right. TS online use the median for avg PE.
|
|
|
|
|
|
Forum Jump
You cannot post new topics in this forum.
You cannot reply to topics in this forum.
You cannot delete your posts in this forum.
You cannot edit your posts in this forum.
You cannot create polls in this forum.
You cannot vote in polls in this forum.