Considering Kohl’s (KSS) and Reconsidering NAIC’s Preferred Procedure

Gary Simms wrote a thorough and insightful article about Kohl’s Corp. (KSS), a large chain of retail department stores, in the July 2006 issue of NAIC’s BITS electronic magazine. See: http://www.betterinvesting.org/articles/bi/1481/12602
(you must belong to NAIC and subscribe to its S&P data service in order to access BITS articles).

Our SSGs were very similar with one major difference: we both used very different methods to estimate the company’s EPS growth for the next 5 years.

(A) Gary began by realizing that KSS’s growth has been slowing and started with 14.2% historical Sales growth for the last 3 years. He then calculated an estimated Sales growth for the next 4 years from Value Line data:

“In the ten-year data section they estimate sales will grow from $15,450 Million in ’06 to $25,950 Million in ’09 -’11. That’s about 13.8% over four years.”

Gary used the 5-12-06 Value Line report and calculated his rate by considering the Present Value as $15,450 from VL’s Sales estimate for 2006, Future Value as $25,950 from its Sales estimate for 2009-2011, Number of years between periods or N as 4, and got a Compound Annual Growth Rate (CAGR) of 13.84% or 13.8% rounded. Gary’s math is correct, but I take issue with his method.

– There are at least two other methods to project a Sales growth rate from VL data: (1) $13,402 actual Sales for 2005 and $25,950 estimated Sales leads to a CAGR of 14.13% Sales growth over the next five years and (2) Value Line’s method (PV = average of 2003-2005 or $11.795) leads to a CAGR of 14.04% Sales growth over the next six years. I think the five year “method” is superior because it relies on actual data (rather than estimated data) for the CAGR’s Present Value.

Gary did not explain why he calculated a 4-year rate from VL data when the SSG is based on a 5-year projection. Nor did he explain why he didn’t use VL’s method. Value Line’s method of calculating growth rates is explained in its manual, “How To Invest In Common Stocks: The Guide to Using The Value Line Investment Survey”, grey shaded box on page 14 (PDF page 19). This manual is available at no charge from: http://www.valueline.com/pdf/common.pdf

– All three methods resulted in small differences: 13.84% (Gary), 14.04% (VL) and 14.13% (Armin). Although the differences were not large, they could be if one or more of the years was unusual or atypical. That’s when the method can be critical and that’s why VL always uses an average of three years, to smooth out any abnormalities. Moreover, small differences often have a big impact on the SSG.

– Gary then lowered his 13.84% to 13.00%, to be conservative I think.

– He then used the SSG’s Preferred Procedure, which is NAIC terminology, to estimate future EPS growth for the next 5 years. The Preferred Procedure starts with estimated Sales growth, subtracts estimated expenses (from a projected Pre-Tax Profit Margin), then subtracts estimated Taxes, and then divides by estimated Shares Outstanding. Gary got 13.20% and chose to go with 13.00% for his SSG as his estimated EPS growth for the next 5 years.

– In summary, Gary calculated 13.84% projected Sales growth from VL data, lowered it to 13.00%, then used that rate to calculate 13.20% projected EPS by using the SSG’s Preferred Procedure (and using three other projected values), and again lowered that rate to 13.00% projected EPS for the next 5 years on his SSG.

(B) I took a different and more direct approach. I checked the long-term EPS estimates by First Call (18.00%), Zacks (17.12%), Reuters (18.14%), S&P (18.00%), and VL (17.50%). All five were between 17.12% and 18.14%, Reuters estimate less one Standard Deviation (which that website provides and which I often rely on) was 16.41%, and I chose to use 15.00% estimated EPS growth for the next 5 years on my SSG.

– Both our judgments were conservative, Gary’s more so. There is no right or wrong with judgments which, after all, are just personal opinions. However, I find it useful to look at the extreme high and low EPS estimates by all the analysts which I’m reluctant to be above or below, at least not without a good reason. The lowest estimate at First Call (13 analysts) and at Reuters (11 analysts) was 15.0%. Gary’s 13% estimate was even more conservative than the extreme low at First Call and Reuters. At Zacks, the lowest estimate (12 analysts) was 12.0%. I would never want to be below all of the analysts at all of these sites.

– I used to routinely use the Preferred Procedure, but no longer do so for several reasons: (1) it depends upon estimating future Sales growth and, apart from historic Sales, only Value Line provides the data to make such an estimate; (2) there are at least three methods for calculating growth rates from VL data which produce three different results; (3) the Preferred Procedure uses a total of four estimates which makes it more unreliable than using only one estimate; and (4) I can double check my one EPS estimate against five sets of analysts and against their high and low extremes to assess whether I’m in the ball park.

armin