CVS Caremark is the last session of this workshop and was the fifth Online Stock Study at the BI website.  It was led by Jim Thomas, VP of Education for the Puget Sound Chapter of BI.  Thanks Jim for volunteering to lead this study.

Materials available for downloading include the audio-visual presentation and related PDF slides, the Consensus SSG in PDF, 12-28-07 Value Line, a list of URLs prepared by Jim, and Q&A answered by CVS Investor Relations.  There is also a short summary on the “First Cut” page at the BI website.  Recently, BI added a typed transcript of the Webcast.

Company Background:

CVS Caremark (CVS) is a large, integrated drugstore chain and pharmacy benefits manager.  Value Line says that CVS is the country’s leading pharmacy chain and fills more than one billion prescriptions each year.  CVS acquired Caremark in 2007, Sav-On and Osco in 2006 and Eckerd in 2004.

CVS operates in two business segments: Retail Pharmacy and Pharmacy Services. As of the end of 2007, the Retail Pharmacy segment included 6,245 retail drugstores (6,164 with a pharmacy), CVS.com, the company's online retail website, and its 462 retail healthcare clinics under the Minute Clinic name.  CVS filled about 528 million retail prescriptions last year, or 17% of the United States retail pharmacy market.

The Pharmacy Services segment provides a wide range of prescription benefit management (PBM) services including mail order pharmacy, specialty pharmacy, formulary management, and claims processing.  This segment consisted of 56 retail specialty pharmacy stores, 20 specialty mail order pharmacies and nine mail service pharmacies, nine automated mail service pharmacies, and a network of 20 mail service specialty pharmacies.

These later facts came from Reuters.com which has one of the best and most thorough company descriptions on the web.

SSG Comparisons:

Here’s the BI Consensus SSG compared to two of mine (one with S&P data and the other with Hemscott data, but both with the same judgments) and to Take Stock.

** I updated the BI Consensus SSG on 4-28-08 to get current data and the current price, but did not change any of its judgments.

Discussion- BI Consensus SSG:

- Like other Online Stock Studies, one of the first things Jim did was to check the quality of CVS by looking for growing sales and EPS, stable or improving Pre-Tax Profit Margin, and stable or improving Return on Equity.  ROE quality, however, was problematic:

** The 2-15 SSG on the BI website without judgment shows “NMF” in the trend down column because 2007 ROE was not part of that data file; the 2-15 SSG with judgments does include 2007 data and does show a ROE downtrend.  Neither of these ROE trends was discussed during the Online Stock Study which was held on 3-5-08.

** The Stock Study used the BI Online SSG which doesn’t indicate the historical trend for ROE (and PTPM) unlike the Toolkit or Stock Analyst SSG.

** The downtrend in ROE was not apparent on the Online SSG (PDF slide 8) and Jim thought the trend looked stable.  It looks like the ROE downtrend on the 2-15 SSG was overlooked.

- To estimate future sales growth, participants were given three “blended” choices (6.6%. 9.0% and 9.6%) along with “other” as CVS’s recent merger made projecting sales growth difficult.  The Consensus choice was 9.0% which represented a blend of 13.0% growth for next year and 8.0% for the next 4 years.

- To estimate future EPS growth, Jim implied three growth rates from Value Line dollar data: 22.4% for the next 3 years; 16.3% for 4; and 12.9% for the next 5 years (PDF slide 35).  The group was then asked to choose from 13.0%, the low end of VL implied; 16.0%, a “typical” analyst estimate; 20.0%, the high end of VL implied; or “other.”  The Consensus chose 13.0%, the low end of VL implied.

** To be fair, Jim and I disagree about the merit of implying rates from VL dollar data and we have argued about this on the I-Club-List.  I think it’s more appropriate to point to VL’s explicit EPS estimated rate (17.0% when the Online Study was done, 15.0% now) rather than imply one or more rates from VL’s dollar data.

** That’s because I think Value Line projected dollar data is not comparable to SSG projected dollar data (even when VL dollars are converted to a rate) as VL uses a different method of projection then our SSGs: See VL’s free manual (grey shaded text box at page 14, PDF page 19), http://www.valueline.com/pdf/common.pdf

** Jim mentioned Morningstar’s estimated EPS rate of 16.1%, S&P’s 16.5%, Yahoo Finance’s 17.3%, and Reuters’ 16.1%, but strangely there was no mention of VL’s estimated 17.0% EPS estimated rate.

-  To project the Forecast High PE, four choices were provided: 21.7, the 5 year historical High average; 23.0, the 3 year High average; 25.0, the 5 year High average from Big Charts; and “other.”  The Consensus chose 21.7, the 5 year average High from the SSG.

- To project the Forecast Low PE, the choices were: 14.0, the 5 year average Low from Big Charts; 15.0, the 5 year average Low from the SSG; 17.0, the 4 year average Low from Big Charts; and “other.”  The Consensus chose 15.0, the 5 year average from the SSG.

- To decide the Forecast Low Price, the choices were: $29.55, the Low PE x Low EPS, or what I call the Growth Company Option; $21.30, the Average Low Price for the last 5 years; $26.10, the Recent Severe Low Price; and “other.”  The Consensus chose $29.55, the Growth Company Option.

- These judgment calls resulted in a SSG Buy for the Consensus SSG as it had a 3.9 Upside-Downside Ratio and a 15.2% Total Return when its price was $39.72 per share.   Two months later, when I updated the Consensus SSG, CVS’s price had risen $1.31 per share and was no longer a SSG Buy as its Total Return had fallen below 15.0%.

Discussion- Armin’s SSG and Take Stock’s:

- When I did my SSG, there had been only minor changes in the EPS estimates that Jim mentioned two months earlier.  I found that the analysts were estimating around 16.0-16.5% long-term EPS with VL now low at 15.00% and First Call as well as S&P high at 17.00%.  FactSet CallStreet was 16.00%, Zacks was 16.40% and Reuters was 16.56%.  Reuters less 1 Standard Deviation was 13.95% and that was the basis for my 14.00% SSG estimate.

- Another difference between the Consensus SSG and mine was that I considered 2003 an outlier and used the 4 year averages as my Forecast High and Low PEs.

- My slightly higher EPS estimate (14.0% vs 13.0%) and my slightly higher Forecast High and Low PEs (with 2003 eliminated as an outlier) explain why I got a SSG Buy with S&P data and with Hemscott data while the updated Consensus SSG did not.

- Take Stock did not get a SSG Buy as its “Buy Under” price ($35.48) was well under CVS’s Current Price ($41.03). Take Stock’s imputed 1.97 Upside-Downside Ratio was also well under the minimum 3.0 criteria.

Conclusions about Comparisons:

(1) Outliers eliminated in SSG Section 3 shape the Relative Value (RV) in SSG Section 4.  For RV comparisons to be meaningful, outliers should be identical, and here they are not so that’s why I didn’t compare RV.  Moreover, we usually don’t identify our SSG outliers which usually make RV comparisons questionable.

(2) Mixing data or drawing inferences from different historical data is not a good idea.  PE averages from Big Charts, for example, are not as pertinent as PE averages from our SSG data.  For example, I found the 5 year average PE for CVS to be different at MSN Money (17.88), Morningstar (19.40), Reuters (20.77), BI/S&P (19.40), and Hemscott (17.75).

(3) Because Value Line makes an explicit EPS growth rate estimate for the next 3-5 years, I see no reason to imply one or more EPS growth rates from Value Line dollar data as guidance for our SSGs.

(4) Maybe the Consensus SSG’s downtrend in ROE with ending equity is not a barbed wire barrier, a red-flag to stop and move on to another stock, because the trend in ROE with beginning equity is up, at least with the updated Consensus SSG.  However, ROE with beginning equity was not discussed by the Online Stock Study and the BI Online SSG also does not consider ROE with beginning equity.

(5) A major weakness of the BI Online SSG, which is separate from but used by the Online Stock Studies, is that the Pre-Tax Profit Margin and Return on Equity section does not report historical trends, unlike the Toolkit and Stock Analyst SSG.  As a result, the BI Stock Studies for both CVS and CL focused on peer group comparisons and the downtrends in ROE were not apparent.  Moreover, the Online SSG creates a vague and unlimited exception for a downtrend in ROE whenever any company is paying off debt. 

Workshop Summary and Final Thoughts:

Question:    What good is a comparison table, why read any of this stuff?

Answer:       Comparisons are a good way to evaluate judgments and to make

                   decisions!

- Armin Fields