Starting at the Beginning

Before we get our hands dirty with the valuation aspects of the investment decision, let us review a brief outline of the
qualitative and quantitative aspects of Buffett’s decision process as observed by Robert G. Hagstrom.[1] This map helps us
navigate the turbulent waters of Wall Street and is comprised of business, management, financial, and market tenets.

Investment Tenets
Business
 Is the business simple and understandable?

 Does the business have a consistent operating history?

 Does the business have favorable long-term prospects?

Management
 Is management rational?

 Is management candid with the shareholders?

 Does management resist the institutional imperative?

Financial
 What is the return on equity (ROE)? [Do not just focus on earnings per share (EPS).]

 How high are the firm’s profit margins?

 Does the firm create one dollar in market value for every dollar retained?

Market
 What is the value of the business?

 Can the business be purchased at a significant discount?

The valuation exercise explained in this article addresses the market tenets part of the process, which are only applicable if
the firm in question meets the other qualitative and quantitative requirements. It should be noted that Buffett stresses doing
your own homework and that entails more than just getting historical accounting data and dropping them into the model
being presented. Buffett has often said that a primary part of his investment homework is reading: including books,
newspapers, magazines, and the annual reports of the subject company and its competitors.[2][3] A thorough reading of
annual reports goes a long way in addressing the questions related to the business and management tenets.

So Many Numbers, So Little Time

Okay, so you have read the annual reports and found more numbers than you know what to do with. Which are important?
Which are essential? Which can be left out? In reality, all of the information presented in these disclosures is potentially
useful for several reasons.

For estimating the intrinsic value of a firm, Buffett attempts to determine the expected return on equity capital (ROE) and the
growth rate of book value (BV) per share, using the following accounting data: revenue, net income, book value of
shareholder equity, earnings per share (EPS), dividends per share, and total shares outstanding. Price-earnings (P/E) ratios
are also useful, but are not typically found in annual reports. Two of the best sources for this data are Value Line Investment
Survey[4] and Standard and Poor’s stock reports.[5] Both contain all the data necessary for the valuation exercise. While
they do require subscriptions, they can be accessed through most public and university libraries for free.
Estimating an Investment’s Expected Return
Ultimately, we will estimate the book value of shareholder equity 10 years into the future, although this process also works
for shorter investment horizons, and we will use that figure as the basis for calculating an expected rate of return on a
company’s stock. Buffett reportedly uses a threshold rate of return of 15 percent, anything less is considered unacceptable.
To get to that point, we will go through several steps that are explained in great detail in books by Buffett’s former daughter-
in-law, Mary Buffett.[6][7] These steps are summarized below, with a few of my own adjustments to make the analysis more
robust. We will use Eaton Corporation (ETN) as our sample firm because Buffett invested over $100 million in the diversified
power management company in 2008.

click image to view model

Step 1, Gather the raw data and create the basic spreadsheet.
In the accompanying valuation model, the historical accounting data for ETN are found in rows 11 through 19. Rows 11
through 16 contain the variables used by Mary Buffett in her Buffettologybooks. In columns N and O, the annual and
cumulative growth rates for each variable are calculated. The remaining data are used as a sanity check. For example, if
EPS growth exceeds that of the revenues and net income, it is likely that management is accelerating EPS growth by
repurchasing shares. While ETN’s average annual EPS growth (11.03 percent) outpaces its revenue growth (8.87 percent),
it does not exceed the growth rate of the net income (12.25 percent). This means that ETN has been issuing shares over the
past 10 years, which is confirmed by noting the cumulative growth rate in shares outstanding (15.06 percent) in cell O14.
Another relationship to keep an eye on is that between revenue growth and net income growth. For ETN, net income has
outpaced revenue by approximately 3.38 percent per year on average, or more than 85 percent for the 10-year period. It is
unlikely that this relationship can persist. Because there is a limit to cost cutting, the growth rates of these two variables
should converge over time.

Buffett also monitors the relationship between growth rate in shareholder equity and the growth rate in EPS. The growth rate
of EPS should outpace that of shareholder equity; otherwise, the incremental return on equity is decreasing, which suggests
that the firm is probably not a viable investment candidate. With ETN, these growth rates are almost equal.

Step 2, Calculate the book yield.

The “book yield” is simply EPS divided by book value per share (BV), as in row 24. Alternately, we could calculate traditional
ROE by dividing net income by total shareholder equity (row 25). The small differences found in some years are likely due to
rounding in the accounting data and per share values. We also calculate the dividend payout ratio annually (dividends per
share divided by EPS) and the 10-year median ratios for all three variables.

Step 3, Estimate the future BV.

Next, we estimate the growth rate in BV by multiplying book yield by the retention ratio from the most recent year. In our
example, the 2008 book yield is 17.8 percent (cell L24) and the retention ratio (calculated as one minus the dividend payout
ratio) is 70.7 percent (cell L27). Therefore, our estimated growth rate in book value equals 0.1261 or 12.61 percent (cell
C31).
Book value for the most recent year (2008) is equal to $38.30 (cell L11). This is the base value used to estimate BV every
year for the next 10 years (row 37). An alternative to this approach is to calculate the average growth rate in BV over the
preceding 10 years and to use that as the growth rate for the next 10 years. In this case, our estimated growth rate would be
10.32 percent (cell N11).

Step 4, Estimate EPS over the investment horizon.

Our annual EPS estimate is a straightforward calculation: multiply BV by book yield (row 38). For example, 2009 EPS is
estimated to be $7.69 per share. The method also produces a 2018 EPS estimate of $22.40, which is used to estimate the
future stock price.

For comparison and benchmarking purposes, we can check the 2009 EPS estimate against the current analysts’ consensus
estimate, which is $3.40 per share. While our near-term estimate is more than twice the current estimate, it is important to
note that we are smoothing out the earnings stream for estimation purposes. What we are attempting to do with this model is
to estimate with a high degree of accuracy the cumulative earnings and EPS 10 years out, not just next year’s earnings.

Step 5, Estimate future stock price.

To estimate the future stock price 10 years from now, we need to find the product of our 2018 EPS estimate and the
expected P/E ratio. It is suggested that the P/E estimate should be calculated by taking the average annual P/E for the last
10 years. However, this is a part of the process I would revise.

Rather than take the average historical P/E, I prefer the median historical P/E; if there are extreme values in our sample of
historical P/E ratios, they will influence the average P/E estimate more than the median P/E estimate. For example, if there
is an extremely high P/E in a recent year, it will cause the average P/E ratio to be higher than the median P/E ratio. And
because we are trying to estimate the P/E 10 years into the future, we must minimize the influence of extreme values and
compounding biases.

The historical high and low annual P/E ratios for ETN are found in rows 15 and 16. Cells P15 and P16 contain the medians
of these high and low P/E values, respectively. The median of the annual high P/Es is 16.5, and the median of the low P/Es
is 11. The median for the entire sample is 13.5, which is calculated in cell E44. Notice that the average P/E of 15.1,
calculated in cell D44, is reasonably close to the median in this case. In addition, note that we will also use the low and the
high values of the P/E sample in our valuation in step six. These values are 5 and 34, respectively, and are found in cells
C44 and F44.

The expected future value of ETN shares can be found in row 45. Four different prices are calculated for each of our P/E
ratio assumptions. For low, average, median, and high P/E ratio estimates, we obtain expected stock prices of $111.99,
$338.22, $302.39, and $761.56, respectively.

Step 6, Calculate the expected return.

The final step of our valuation process begins by calculating the expected annual dividends of the 10-year investment
horizon. The model must include the current price of the subject stock. I used $65.00 per share, which is an approximation
of the average price paid by Berkshire Hathaway when it was accumulating the stock during late 2008. This value is entered
in cell B52.

The model provides the results of our valuation exercise. Depending on our estimate for the P/E ratio 10 years from now, the
expected annual rate of return on investment ranges from 9.4 to 27.2 percent, with an expected return of approximately 17.9
percent (cells O56 through O59).

Step 7, Interpret the results.

Based on the analysis so far, we can see that Buffett’s purchase of ETN at an average of $65.00 meets the 15-percent-
expected-return threshold in our base case (cell 058), where we assumed average growth in BV and a median P/E ratio.
Our results also indicate that under an average P/E assumption of 15, the growth rate in book equity only needs to be 10
percent for the investment to provide a 15 percent return on investment (cells H66 to H74). Alternatively, if the growth rate
falls below 10 percent, it appears that earning a 15-percent return on investment becomes much less likely.

As a sanity check, we can compare our earnings estimates to those of the analysts following ETN. At the time of writing,
analysts’ consensus for 2009 EPS was $3.40, less than half of the projected 2009 EPS of $7.69. But before we throw the
methodology away, it is important to remember that we are creating a forecast that is supposed to be reasonably accurate
over a 10-year horizon, not just next year. That means that the forecasted cumulative earnings over the 10-year period
should be close to the actual cumulative earnings, and the EPS forecast in year 10 of the estimation period needs to be
relatively close to actual earnings in that period. It is impossible to say how big our estimation bias of these two numbers is
today.

Having noted the discrepancy in earnings projections, we can test the disparity by recalibrating the model to start with the
2009 EPS estimate at $3.40. Keeping the same growth assumptions and payout assumptions, our model tells us that there
is little chance of earning a 15-percent return on investment. The base assumptions provide for an 8-percent expected return
on investment. However, the outlook gets brighter once we factor in that ETN is trading well below Berkshire’s average
purchase price of $65 (around $40 per share) and revise the model accordingly.

Now, the model shows an expected return of 14 to 15 percent, right at the threshold of “acceptable.” Therefore, if one is
uncomfortable with the model and recalibrates the earnings stream based on analysts’ forecasts for 2009 EPS, the current
market price offers a return close to the 15 percent a Buffett-type investor would be seeking.

Conclusion
A relatively simple valuation approach is presented in this article. The valuation and rate of return calculations are based on
a firm’s estimated return on shareholder capital. Your best bet to find successful companies that outpace their competition is
to identify firms that generate high past returns on shareholder capital and maintain competitive advantages that will allow
them to do so in the future.

Once you identify such firms, this valuation tool will help you determine whether or not the market is offering the firm’s
shares at a sufficiently low price to enable you to earn your expected rate of return. Note that the value and rates of return
obtained from the model are only as good as the inputs. By doing some homework and analysis, you can reduce the
potential biases in your inputs and consequently, in your valuation estimates. This should allow you to have more confidence
in your investment decisions, an invaluable tool in minimizing the emotions that often arise when markets become volatile.

If you determine that this process is not for you, you can still invest with professionals who use a similar process. In addition
to Berkshire Hathaway, the Sequoia Fund,[8] Fairholme Fund,[9] the Longleaf Partners Funds,[10] and the Wasatch
Funds[11] (among others) are all deep-value investors, and they are often included in discussions of those who can
successfully implement this investment approach.