The Mark of a Good Business: High Returns on Capital

“Leaving the question of price aside, the best business to own is one that over an extended period can employ large amounts of incremental capital at very high rates of return.” – Warren Buffett, 1992 Berkshire Hathaway Shareholder Letter

A good business is one that can earn very high returns on capital. Rarely can such a business invest all of its capital back into the business. One way to find companies that can is to look for companies that have grown book value at a high rate on a per share basis.

A business can still be a good investment if it can’t reinvest all of its earnings back into the business. An example is American Express. Prior to the 2008 economic crisis, Amex was earning over 30% on equity but was only reinvesting about a third of its earnings back into the business. The remaining two-thirds were paid out in the form of dividends and share repurchases.

There are numerous ways to measure return on invested capital. None of them is perfect. Any of the various metrics and ratios investors use to analyze a business are abstractions and, as such, typically tend to oversimplify the economic reality of the business. They are short-cuts we use to point us in the right direction so we can spend our precious time researching businesses that offer the most opportunity.

Return on Incremental Equity

I like to look at the total amount of equity that has been added to a business over the past decade and then calculate the return on that additional investment. This approach also allows me to calculate what percentage of the company’s earnings was reinvested, which in turn is useful in forecasting the future growth in earnings.

I typically use Value Line when I do this because the layout is very conducive to this type of analysis. It is one reason why investors like Buffett, Munger and Li Lu like Value Line.

It is useful here to remember Buffett’s reminder that it is not necessarily a cause for celebration if a business grows its earnings year after year. The same thing happens to a savings account if you add more capital each year, which does not make a savings account a good investment. It’s the return on this additional capital that determines whether something is a good investment or not.

To illustrate, let’s look at Johnson & Johnson (JNJ). In 2000, JNJ had shareholders’ equity of $18.8 billion. At the end of 2009, its shareholders’ equity had grown to $50.6 billion. We can calculate that, since 2000, JNJ invested $31.8 billion back into the business.

During that same time, earnings grew $8.1 billion, from $4.8 billion in 2000 to $12.9 billion in 2009.

By dividing the additional earnings of $8.1 billion by the additional $31.8 billion in capital, we can see that JNJ earned a return of 25.5% on its investment, which is very good.

It is also useful to look at what percentage of its total net earnings JNJ reinvested back into the business. The reason is that this is suggestive of how much of its future earnings JNJ is likely to reinvest. By multiplying the rate of reinvestment by the return on that investment, we can then calculate an expected growth rate for earnings.

Since 2000 through 2009, JNJ earned a total net profit of $89.7 billion. Since we already know that JNJ reinvested $31.8 billion over that same time period, we can calculate that JNJ’s rate of reinvestment is 35.5%.

If JNJ can continue to earn 25.5% on equity and reinvest 35.5% of its earnings, earnings should grow at about 9% (.255 x .355).

Keep in mind that this does not include dividends or share repurchases. The latter would cause earnings per share to grow at a faster rate. Also, it does not include an analysis of where JNJ is selling in relation to its intrinsic value which could have a material impact on the expected total return. Finally, this type of analysis works best with a stable business that enjoys durable competitive advantages, such as JNJ.

Another example is Southwest Airlines which is a successful airline that operates in the highly competitive and capital intensive airline industry. Between 2000 and 2009, Southwest’s shareholders’ equity increased by $2 billion. Earnings were $140 million in 2009 compared to $625 million in 2000 and have generally bobbed around over that time period. The return on that additional $2 billion has been relatively poor.

Calculating the return on incremental equity over a long-period of time should prove a useful tool in your analysis of prospective investments. Coupled with the rate of reinvestment, it can also allow you to get an idea of how fast a company can be expected to grow its earnings.

You can also use this approach to invert an expected rate of earnings growth to examine what combination of ROE and rate of reinvestment will be required to produce it.

In succeeding related posts, I’ll look at Buffett’s use of return on average tangible net worth and Greenblatt’s use of return on tangible capital employed to determine whether a business is good.


9 thoughts on “The Mark of a Good Business: High Returns on Capital

  1. TMurphy

    In the book, “It’s Earnings that Count” by Hewitt Heiserman (A more accurate title should be Its Cash Flow and Economic Value Added that Count) has a useful measure he calls Return on Greenest Dollar. The ‘green’ in greenest dollar refers to last years investment in the business, rather than total capital. Companies with authentic earning power tend to produce high returns on greenest dollar every year. A deterioration of return on greenest dollar is often the canary in the coal mine. I have found this to be a very useful book and I’m surprised it in not on more value investors reading list.

  2. adib


    Excellent post. thanks.

    How would you calculate the above for MSFT which has been buying back so much stock over the years? Book value in FY 2001 and FY 2010 are identical at $46 billion ( in fact 2009 is lower by $1 billion). Net Income grew from $7.7 billion to $18.7 billion ( $11 billion).

  3. Greg Speicher Post author

    Adib, I don’t think this type of analysis works well for a business that can grow its earnings without investing additional capital. This approach is best suited for businesses that require additional investment to grow and to determine if it is generating a satisfactory return. MSFT is one of the few businesses on the planet that can grow their earnings by $11 billion without additional capital. It does raise an important question in valuing MSFT and that is how much value to give their massive cash position. With $37 billion in cash the company is over-capitalized in my opinion.

  4. Serge

    Great post.
    @adib: In the case of a company that buys back its shares (like MSFT), I would add the amounts spent on buybacks to the retained earnings for the purpose of this calculation. So, you would calculate how much additional EPS the company produces for each $1 of earnings that it either retained or used for buybacks. In a sense, buying your own shares is no different from purchasing another business – it is a capital allocation decision that should produce returns in the form of EPS.

    I need to add this Buffett metric to 🙂 It is a good one!

  5. Pingback: D-Day for Dart, Anite, Games Workshop : Interactive Investor Blog

  6. Pingback: Tweets that mention The Mark of a Good Business: High Returns on Capital | GregSpeicher --

  7. Jim

    Excellent site! Just stumbled upon it today. This is a very logical calculation. I believe it could potentially be more useful than CROIC. In any event, there’s a couple minor errors that I wanted to point out regarding your example (JNJ).

    “During that same time, earnings grew $8.1 billion, from $4.8 billion in 2000 to $12.9 billion in 2009.”

    Your $12.9 B figure is from 2008. In 2009, their earnings were $12.26 B.

    “Since 2000 through 2009, JNJ earned a total net profit of $89.7 billion.”

    Very minor but it was actually $90.05 B during that time period.

    In any event, those numbers will change your percentiles a bit. Have you thought about using “adjusted earnings” in this equation (i.e. removing non-cash items such as impairments to intangibles which affect the bottom line for reporting purposes but don’t have an actual financial impact on the business)?

    All the best.

  8. value_123

    Thank you for a very interesting post Greg. I have heard Buffett talk about this issue many times in the past. The ideal business is one that not only earns high returns on capital but can incrementally re-invest the excess profit at high returns in the future. However, I have never seen Buffett quantify how to assess a company’s ability to re-invest their profits. Your article is the first attempt I have come across in putting a specific quantitative test to assess the re-investment of profits.

    My sense is your approach works best in a very stable, growing business. To put it to the test, I used Wal-Mart as an example.

    In fiscal 2002, WMT had shareholders’ equity of $35.1 billion. At the end of fiscal 2011, its shareholders’ equity had grown to $68.5 billion. We can calculate that over the past 10 years, WMT invested $33.4 billion back into the business. During that same time, earnings grew $9.8 billion, from $6.6 billion in 2002 to $16.4 billion in 2011. By dividing the additional earnings of $9.8 billion by the additional $33.4 billion in capital, we can see that WMT earned a return of 29.3% on its investment.

    The 29.3% return on incremental equity does not seem right in the context that the company’s 10 year avg. ROE is 20-21%. The major flaw with the above approach seems to be that Wal-Mart has been buying back a significant % of the outstanding shares over the past 10 years which most likely overstates the return on investment stated above (i.e. shareholders’ equity has been reduced while NI remains constant). In the case of WMT, they started the period with 4.481B shares and ended with 3.67B shares o/s (20% decrease in o/s).

    Any comments or suggestions on how I should adjust the figure for Wal-Mart given their significant share buybacks?

  9. John Huber

    Excellent post…. The comment above is a couple years old but just came across this post. Great examples Greg.

    In response to question above, what I would do is add the amount of additional treasury stock to the additional retained earnings to identify the true incremental capital. The best way is to actually look up the exact amount spent on buybacks during the last 10 years. But a quick way is to take the difference in shares and multiply by an average price. In this case, WMT roughly bought back 800 million shares. I didn’t look up the prices, but if they bought them at an averag of $50, they spent about $40 billion on buybacks. Add this to the additional $33.5 billion in equity capital, and you arrive at a more accurate additional capital of about $73.5 billion, making the return on incremental equity somewhere around 13.5%, which I think is probably a more accurate figure of WMT’s reinvestment opportunities.

    You can also break this down by per share earnings and per share book values to get a good measure for each year…


Leave a Reply

Your email address will not be published. Required fields are marked *