Yesterday, a visitor brought up a couple of points regarding the analysis of Johnson & Johnson. It looks as though growth might be slowing based on the 2001-2006 owner earnings growth rate and he brought up a good point-is it a cause for concern?
Looking at individual timeframes
You used the median FCF growth rate for 5 year time-periods which gave you 16.1%…is there a reason you consider this to be more accurate than say, the growth rate for the past 5 year period?
Buffett tells us:
Do not take yearly results too seriously. Instead, focus on four- or five-year averages.
If we focus on the single 5-year timeframe from 2001 to 2006, we only get a snapshot of owner earnings in 2001 and again in 2006. Doing so ignores the actual performance of all the other years in our ten years of data and makes us focus solely on two yearly results.
Was 2006 a relatively slow year? Was 2001 abnormally high? A single 5-year snapshot won’t tell us that which is why I like to look at multiple timeframes. That also brings us to the next part of the question:
Using median values for more accurate results
The past two rolling 5 year periods (ending in 05 and 06 respectively), are considerably lower than the 16.1% median used in your calculations.
From 2000-2005, the growth rate was 13.5%; from 2001-2006, it was 10.2%. Then again, from 1998-2005 it was 15.2% and from 1999-2006 it was 16.6%. Which one is the best? In truth, we don’t know. That is exactly why we use a median value of all of the data.
Looking at one period, we don’t know if a single year in the two we analyze is “normal” for the business. Looking at multiple timeframes, we see normal, above normal, and below normal. Taking a median value (instead of an average), we get the “middle of the road” number without any unusually high or low occurrences to skew our data.
(For non-math people, median gives us the number in the middle as opposed to average which is sensitive to a single ultra-high or ultra-low point of data.)
The importance of the margin of safety
Indeed, if you change that 16.1% to 10.2% in excel (the 01-06 rolling period), you would get a per-share value of $61.75- not far off from where JNJ currently trades.
Absolutely, which is why the margin of safety is so critical. According to the full ten years’ worth of data, JNJ grows owner earnings at 16.1%. But if it does not continue to do so, and growth slows to 10.0%, the company is still fairly priced today (at $61.20 with a value of $61.22).
Here’s the thing-if JNJ’s owner earnings grow at 10.0%, and if we pay $61.20 today, we should still earn 15%. Why? We are purchasing that cash using a 15% discount rate-our expected return. If the company grows owner earnings at 16.1% as projected, we should earn far more than 15%. At that rate, investors purchasing JNJ at $83.20 this year would earn 15%.
The margin of safety does two things-it enhances our returns if things go according to plan, and it protects us if things don’t. Mohnish Pabrai, value investor and Buffett fan, explains the margin of safety this way:
Heads I win, tails I don’t lose too much.
Not a bad way to make money, right?
So, how do we know if growth is slowing?
If all of the consecutive timeframes indicate that growth is slowing, then it may very well be slowing. But one or two abnormal periods out of seven is normal as businesses go through their regular cycles.
Of course, you need to look at the business as well. No matter how good the numbers look, a VCR maker with no other business line will likely run into problems in the upcoming years.
Hope that helps!
Filed under: How to Value a Business
Hi Michael,
For those not clear on the difference between GAAP earnings and Owner earnings, check out The Importance of Earnings.
In my opinion, there are a few fundamental flaws with Hagstrom’s approach. Here is my “short” response, but this is worthy of a stand-alone post next week.
Hagstrom’s Method Relies On Earnings
Earnings are for the government; cash is for the owners. Can you try to predict future GAAP earnings and use that as a basis? Perhaps. But, to what end? Owners don’t get earnings and businesses can’t use them for growth. Over the long-term, businesses will grow at roughly the same pace as their cash does, regardless of what earnings do. The opposite is not always true (in fact, it rarely is).
The Business Sense Of Hagstrom’s Valuation
From a business standpoint, Hagstrom’s valuation doesn’t make sense. A business can’t use GAAP earnings to pay down long-term debt; so, why would the calculation subtract the debt? Admittedly, I haven’t looked at the Quicken site for JNJ, but where do the rest of JNJ’s assets/net worth come into play? When buying a business, that is a critical portion of the valuation.
Munger’s Take On The Warren Buffett Way
This Quicken/Hagstrom valuation is based on Hagstrom’s teachings in The Warren Buffett Way. Charlie Munger, Warren Buffett’s partner, criticised the book as
not much of a contribution to human knowledge
(See the end of the linked article)
The Audited Results of Hagstrom’s Methods
If Hagstrom is using these methods to invest like Warren Buffett, they aren’t working for Hagstrom as they are for Buffett. According to Morningstar (as of the date below), Hagstrom’s Legg Mason Growth Trust Primary mutual fund has returned 9.04% on average for the past ten years.
As I said, that’s the “short” response. After the current series on business moats, I’ll play with the Quicken site and post a more detailed comparison.
In the meantime, I hope that helps!
( REPLY )
I start with a Morningstar (Premium) screen of all businesses with 10 years of positive free cash flow. I ran it just now and came up with a list of 504 businesses. That shouldn’t take me more than three or four hours (I only look at companies that jump out at me).
After that, I’ll head over to the NASDAQ or NYSE sites to see which stocks are posting the biggest drops. Wall Street tends to overreact to “bad” news that doesn’t necessarily affect the business but pushes the stock price around.
I also run a screen on MSN Money to see which stocks have a P/E less than 10, $500 million market cap, Book Value greater than zero, and profit margins above 20. (I’ll play with these numbers to get more results.)
Finding the companies is easy – look for beat up stocks and see if the news that beat them up will affect their business. Then, look for the moat.
( REPLY )
Is there any chance of you uploading your MSN search criteria? I believe they can be exported from the msn search app…..Otherwise could you elaborate on which Book Value you mean…there is price/book and bookvalue/share, also maybe explain the difference. Also with the profit margins, which do you use, and why? Maybe you have other criteria that could help narrow the searches.
Many thanks your site is top notch for the learning value investor, I too am really looking forward to the book. Keep up the valuable work.
Many thanks
Stuart
( REPLY )
Hello Joe:
Quicken.com provides a intrinsic value calculator based on Robert Hagstrom’s interpretation of Buffett’s writings on the topic. Here are the comments provided for JNJ:
“If we assume initial earnings of $10.6 billion grow at a rate of 9.31%, and we discount those future earnings at a rate of 15.00%, we arrive at a net present value for the company’s next 10 years of earnings of $80.9 billion. To account for potential earnings beyond the 10th year, we estimate a growth rate of 6.00%, a discount rate of 12.00%, and we arrive at a continuing value of $113 billion. To complete the calculation we add these two figures together, subtract the long-term debt for JNJ ($2.01 billion), and divide by the outstanding shares (2.90 billion) to get a per share intrinsic value of $66.09.”
One major difference in your approach and the one detailed above is using FCF vs. GAP earnings. The other difference observed is adding the book equity to the DCF versus the approach above which only subtracts the amount of long-term debt. Any comments on the difference in these two approaches and perhaps some logic behind them? No explaination is needed for the difference between FCF and GAP earnings.
Thanks, Michael
( REPLY )