Investment Philosophy

Importance of ROIC Part 4: The Math of Compounding

I thought I’d circle back to discuss the topic of compounders and return on capital. I wrote a few posts about earlier this year, and there have been numerous comments and questions.

In this post, I want to discuss the actual math behind the compounders, to try and show why return on capital is so important to long term business owners (which is what we are as stockholders).

To recap what I mentioned earlier, I usually put investments in two broad categories, but ideally, I’m looking for:

  • A business that can produce high returns on capital
  • A business that can reinvest a large portion of earnings at similar high rates
  • A business run by good management, who will allocate the excess cash in a value creating way (preferably management owns a large stake in the business themselves, aligning interests)

These three factors combine to create the rate that the business compounds intrinsic value over time.

The Math Behind Compounding Intrinsic Value

The math is simple.

To briefly review, I find it helpful to use a back of the envelope formula as a way to think about the rate at which a business is compounding its intrinsic value. Basically, a business will grow its intrinsic value at a rate that equals the product of two factors: the incremental return on invested capital (ROIC) and the reinvestment rate.

A simple example: a business that can reinvest 50% of its earnings back into the business at a 12% return on investment will compound the intrinsic value of the enterprise at 6% annually (50% x 12%). See this post for more discussion on this.

Let’s look at a hypothetical example of 2 businesses (Company A and Company B):

  • Company A produces 20% ROIC and can reinvest 100% of its earnings
  • Company B produces 20% ROIC and can reinvest 50% of its earnings

In both examples, we’ll assume the investment is a long term investment over 15 years. Many investments don’t last this long, but in this exercise we are imagining ourselves as a long term partner in the business, and this is how I happen to think about stocks anyhow. And business owners don’t trade in and out of businesses every year or two. We’ll look at various starting and ending valuations, and we’ll see how both the valuation (the P/E ratios) as well as the quality (ROIC) impact our investment returns over the 15 years.

Example 1—Company A

In this example, we’ll assume earnings are cash earnings to make things simple. Let’s assume the business produces $1.00 of earnings per share, and let’s say Company A has a stock price of $25.00 (a P/E of 25). Let’s assume the business is a growing enterprise and it can reinvest 100% of its earnings back into the business at a rate of 20% after tax (20% Return on Capital). The business sells a niche product to a growing market of customers that rely on Company A exclusively, thus the business has a nice competitive advantage over potential competitors. This “moat” allows them to achieve 20% returns on capital for the next 15 years.

Let’s take a look at the value of the business in 15 years:

  • Year 1 EPS = $1.00
  • Year 15 EPS = $15.40

Pretty simple… if the business has an ROIC of 20% and can reinvest 100% of their earnings, then earnings will grow at 20% over time, and the growth of the intrinsic value of the business will also approximate this 20% annual growth rate.

Now, let’s say we paid 25 times earnings for this business 15 years ago ($25 per share). Let’s take a look at what the value of our stock (and the 15 year CAGR) will be at various P/E multiples (remember we paid $25 per share 15 years ago for Company A):

  • 10 P/E: $154.00 per share (12.9% CAGR)
  • 15 P/E: $231.00 per share (16.0% CAGR)
  • 20 P/E: $308.00 per share (18.2% CAGR)

So for this wonderful business, even paying 25 times earnings worked out to a stellar return for shareholders of around 13% annually for 15 years even as the P/E multiple contracted from 25 all the way down to 10, which would be a very low multiple for a great business like this.

Example 2—Company B

Let’s assume that this business–Company B–reinvests half of its earnings at a rate of return (ROIC) of 20% and pays out the other 50% as a dividend. In this case, the intrinsic value of the enterprise will compound at 10% annually (20% ROIC times the 50% reinvestment rate equals a 10% growth rate). Let’s assume we paid the same price as the first business (25 times earnings). Let’s assume the same $1 starting EPS.

In 15 years, EPS will equal $4.17 per share (assuming a constant 20% return on incremental capital and a 50% reinvestment rate). Let’s also assume a constant 2% dividend yield, since Company B can pay out a portion of earnings as dividends, unlike Company A which reinvested all of its earnings.

  • 10 P/E: $41.77 (5.5% CAGR)
  • 15 P/E: $62.66 (8.3% CAGR)
  • 20 P/E: $83.54 (10.4% CAGR)

So both of the above businesses produce 20% returns on invested capital, but Company A is clearly the superior investment if both are priced around the same level. This is simply because company A has twice the level of investment opportunity, as it can reinvest all of its earnings at 20%, whereas company B can only invest half of its earnings at 20%.

Logically, this makes perfect sense. If two businesses (Company A and Company B) have opportunities to make 20% returns on incremental investments, but Company A can invest twice as much as Company B at that 20% rate of return, then company A will create much more value over time for its owners than Company B.

Both companies above will show up in screeners as businesses that produce 20% ROIC, but one is clearly superior to the other because it can retain and reinvest a higher portion of its earnings, and thus will compound intrinsic value much faster.

Picking the “Right Business” is More Important than Picking the “Right Multiple”

In Part 3 of this series, I mentioned that ROIC is far and away the most important factor to consider, as the math (the back of the envelope formula) shows. A business that produces 6% ROIC will not compound intrinsic value regardless of how much they can invest back into the business (in fact, a business that produces low returns would be better off not reinvesting, as owners could likely reallocate those earnings at higher rates elsewhere).

So a business that can produce above average returns on capital is crucial when it comes to compounding value.

But I also wanted to demonstrate that picking the right business (i.e. the one that can invest large amounts of capital at attractive rates of return) is far more important than paying the lowest multiple to current earnings.

Notice in the simple example above, I assumed that we paid a P/E multiple of 25 for both businesses. Obviously, at the same price to earnings ratio, the business that can compound intrinsic value faster will create better returns for shareholders. But let’s compare investment results if we pay a lot less for Company B (the inferior compounder).

Company A vs. Company B at Various Valuations

The tables compare purchase price P/E’s for Co A and B (the top row) and compares that to various sale P/E’s 15 years later (the left hand column). As you can see, at a certain valuation, Company B is a better investment, but I noticed that you could roughly pay twice the valuation for Company A and still come out ahead over time. And I used 15 years, but the longer you own Company A, the wider the gap gets between Company A’s and Company B’s investment result.

Comp Math 1

So at a certain gap between the valuation levels, Company B becomes a better investment. But you’d have to pay 2.5 times the valuation of Company B in order to get an inferior result from Company A.

Below, you can see that at twice the valuation (P/E of 20 vs. P/E of 10), it is roughly even. But if you have to pay 15 times earnings for Company B, Company A is a vastly superior investment.

Comp Math 2

If the ending P/E is the same for both companies (unlikely given the better quality of Company A), Company A still outperforms Company B as an investment, even if Company A’s multiple contracts and Company B’s multiple expands and even though we paid 20 P/E for Company A and just 10 P/E for Company B.

Keep in mind, Company B is still a good company (it’s compounding value at 10%), it’s just not nearly as good as Company A, which will be the superior investment all the way up to twice the price relative to earnings.

If you’re right about the business, you’ll make a lot of money.” – Warren Buffett, University of Georgia, 2001

Some of you might be rolling your eyes at the simplicity (obviously, high ROIC and the corresponding growth will lead to good investment returns). Other readers are thinking—yeah, owning a great business is great, but the problem is predicting which businesses will be able to sustainably produce these high returns on capital.

I agree… it’s difficult. But the purpose here is to show the actual math behind the ROIC and the ability of a business to reinvest large portions of earnings and why those two factors are so important to long term owner returns.

This is why Buffett is always talking about great businesses. It’s not just because he wants to sound like a simple, wise, grandfatherly figure--it’s because of the math. If you pick the right business, the multiple you pay for the business is far less significant on your returns than the quality and sustainability of the returns on capital.

Now, I am not recommending paying 25 times earnings for quality businesses. As the skeptics pointed out earlier, it’s far too difficult to predict what the next 15 years will look like.

But that doesn’t mean I want to abandon the effort to locate great businesses. It’s just that ideally, we want to find them cheap enough to secure a large margin of safety in case we are wrong in our assessment of the quality of the business, as this gives us two things:

  • A margin of safety if we were wrong about the quality or sustainability of the business’ return on capital
  • The benefit of much higher returns if we were in fact right about the business

This is part art, part science, but over time, sticking to businesses that create high returns on the capital it invests will significantly improve the probability of achieving above average investment returns.

In other words, as some investors like to say—and a phrase I often repeat—Heads, we win. Tails, we don’t lose much…

In the next post, we’ll take a look at a real life example (one among many examples), to show that a business that produces consistent returns on capital over time does in fact create real value for its owners.


John Huber is the portfolio manager of Saber Capital Management, LLC, an investment firm that manages separate accounts for clients. Saber employs a value investing strategy with a primary goal of patiently compounding capital for the long-term.

John also writes about investing at the blog Base Hit Investing, and can be reached at

44 thoughts on “Importance of ROIC Part 4: The Math of Compounding

  1. This was a great post. One of the best in a long time, I have always read how Buffett covets high returns on capital, but never understood the why behind it.

    This post clearly explained it better than many other books I have read. Even Greenblatt’s book mentioned return on capital halfheartedly, the math wasn’t there, and so did Marry Buffett’s books, and Timothy Vicks, Hagstrom (Warren Buffett Way) and Pabrais’s book (Dhando) but I was always left wondering …ok …….why is it so good…. ? I mean we all say “yeah buy compounders”, what does that mean? Now I know.

    One question how does debt play into all this? I remember back in the old days when Buffett said return on equity capital he referred to ROE, but ROIC is a bit different. Would the above calculations and math still hold for ROE?

    For example if a company had a ROE 20% and can reinvest 100% of their earnings, and earnings will grow at 20% over time, then will intrinsic value of the business also approximate this 20% annual growth rate?

    Again Great Piece,


    P.S. I do remember Munger back in the day also mentioned

    “Over the long term, it’s hard for a stock to earn a much better return that the business which underlies it earns. If the business earns six percent on capital over forty years and you hold it that for forty years, you’re not going to make much different than a six percent return – even if you originally buy it at a huge discount. Conversely, if a business earns eighteen percent on capital over twenty or thirty years, even if you pay an expensive looking price, you’ll end up with one hell of a result.”

    Again, I was unsure if he meant ROE or ROIC
    (Buffett would refer to ROE as return on equity capital)

    1. Hi Judson, yes the Munger quote is a relevant one here. And I wouldn’t try to think of ROIC vs ROE. You could use either one… as an equity holder you’re buying the equity, so ROE is relevant, but you don’t want to look at ROE without looking at the balance sheet. Basically, the return on capital is what you want to know (debt and equity capital)… the concept is what investment opportunities does the company have and what kind of returns can they generate on the capital they allocate to those opportunities? Buffett talks about ROE sometimes interchangeably with ROIC, but he qualifies it by saying he likes businesses that produce high returns on equity without using a lot of debt. So in essence, Buffett is trying to figure out the return on capital a business produces. If a company has a lot of debt, they might have a sliver of equity and its ROE might be high, but the actual return on the invested capital might be mediocre. These aren’t the situations that Buffett was interested in for the most part. So a high ROE is only a good thing if it’s accompanied by a modest level of debt. Of course, these are just rules of thumb. Each business is different. You could just use ROA, which will allow you to compare businesses with various levels of leverage. But each situation is unique. For financials, I look at ROE but I consider the leverage, so the shortcut is just to look at the ROA, as ROE = ROA times leverage.

      The key is to understand the concept… which I could simply describe as: A company has a certain amount of capital (dollars). How good is the company at generating earnings from those dollars? (i.e. what is the rate of return the business gets on the dollars it has invested in the business?) Obviously, the higher the return, the better the business.

  2. John,
    I appreciate the simplicity of the math and the explanation. I will reread you article re: MKL, but it would help if you could answer the question regarding MKL’s amount of reinvested earnings/cash flow. I understand and appreciate MKL has the advantage of leverage that increases the ability to invest. Even so your using MKL as an example with specifics would be greatly appreciated. If all the information is in your prior article on MKL a compounding machine just say so and I will reread with a keener eye. Thank you.

    1. Hi Tony. Markel basically reinvests nearly all of their earnings (and a large portion of their comprehensive income comes from unrealized investment gains, which matter greatly in terms of value creation over time, but don’t flow through the income statement). I find it helpful to think about Markel as a holding company with a sizable investment portfolio alongside a profitable insurance operation. A simple way to look at it: Markel has around $1100 of investments per share, net of debt. They have around $500 per share of equity. If they can produce 5% returns on their investment portfolio, they’ll get $55 of earnings from the investments, which is around 11% return on equity just from the portfolio. They’ve historically been profitable in their insurance operation, and so any profits from insurance would add to that base.

      Since they retain the vast majority of their earnings, their book value has compounded at around the rate of return they’ve generated on their equity capital (which has historically been better than 15%).

      As long as they can continue to achieve profitability in their underwriting and can find ways to invest their portfolio at the same return over time, they’ll continue to create the same returns on equity and the same growth in intrinsic value.

      You can read the article for more details. Book value (and intrinsic value) have compounded at close to 20% since IPO in 1986. I assume lower compounding, but I see no reason Markel can’t achieve similar results to what it did in the past for a long time to come.

      1. Hi John,

        I greatly appreciate your response and the way you put it together. It is much clear now how they arrive at the numbers. Thank you so much. Now I see how the puzzle pieces fit together to arrive at the highly desired consistent result.

  3. John,
    I am curious about the math in your example for Company B where you pay 10x earnings. Company B earns a ROIC of 20% and retains 50%, while paying out the other 50% as a dividend. In the first example of Company B, you assumed a constant 2% dividend yield ($1 eps x 50% payout = $0.50 dividend…divided by $25 stock gets you the 2% dividend yield). Yet in the second example for Company B, you used the same 2% dividend yield. Wouldn’t the dividend yield be 5% (same $0.50 dividend, yet now divided by a $10 stock…gets you 5% dividend yield)? Thus changing the return profile for paying 10x earnings at the beginning and end of the 15 year period from 12% to 15%?

    1. Mark. You’re right. I used the same dividend rate for Company B, and the lower purchase price (at least at 10 times earnings) does in fact tip the scale in favor of Company B (15% CAGR vs 13% CAGR roughly). I was taking the roughly $15 of dividends and was not reinvesting them. If we reinvest them, we’ll see a return of roughly 15% for Company B at 10 times earnings. I discovered that the “breakeven” point in this example is 12 times earnings. So basically, if you had to pay 25 times earnings for Company A, and you knew that the multiple would contract to Company B’s multiple after 15 years, Company A would be the better investment all the way down to 12 times earnings. At under 12, Company B (if you reinvest dividends) would have a slight edge. However, this assumes that A and B would trade at the same multiple, which is unlikely.

      But thanks for correcting me on that error. I need to update those numbers, which I’ll do now.

  4. John,

    I love your ROIC series and, in general, strongly agree with your conclusions.

    I believe, however, that there are a couple of problems with your examples.

    The first is the math. For Company B, if they pay out 50% of their earnings as dividends, and you initially pay a 10X multiple, then they must have a 5% dividend yield, not a 2% yield. Your annual return for Company B would be 15%, not 12%.

    The bigger issue, I believe, is that as an investor, I am going to take the dividends I receive from Company B and I am going to reinvest them. If I can reinvest all of the dividends and get a 20% return on them, then, absent taxes and if the P/E multiples were the same, I would be indifferent between Company A and Company B.

    I suggest instead that there are 2 reasons why Company A is far superior to Company B. The first, of course, is taxes. With Company B, I have to pay taxes on the dividends before I can reinvest them. Over 15 years, this would make an enormous difference. Second, finding companies that can generate a 20% ROIC for 15 years is extraordinarily difficult. I would much prefer to let Company A reinvest my share of the earnings back into the business at a 20% ROIC rather than having to find a new 20% ROIC opportunity (that was available at a reasonable price) every year into which to reinvest the dividends.

    1. Greg,

      I corrected the error. It was a bit confusing the way I presented it anyhow, so I decided to make a table. I did not assume reinvestment of the dividends for Company B in the second example, and corrected that in these tables.

      I haphazardly put the examples down after just picking some example valuations to compare, but the idea I had was to demonstrate that a business that compounds faster will create much more value over time. And in fact, both of these businesses are decent businesses (one compounds at 10%, the other at 20%). My idea was to show you could actually pay a decent premium (as it turns out, roughly twice the price, which makes sense since it’s compounding twice as fast) and still get the same result.

      But the key (at least for me), is that the longer you own the better business, the wider the gap (or the better the relative returns get) between the good and the “less good” business.

      Your point about dividends is a good one. I thought of that also, although I thought the post was getting too cluttered as it was to try and demonstrate that. I think depending on reinvestment opportunities, the dividends could work in the investor’s favor, but it’s complicated because you have to assume a variety of things such as interest rates, tax rates, and reinvestment opportunities.

      And generally speaking, I think a business that can reinvest the earnings at 20% (such as the hypothetical Company A) will be a very high hurdle because unless you are in a tax advantaged account, you’re paying capital gains on those dividends as they come in, thus lowering your after tax results and widening the gap between Company A and B.

      Anyhow, you bring up good points… thanks for the comment (and pointing out the dividend error).

      I plan to post something about a real life example in the next post.

  5. One question how does debt play into all this? I remember back in the old days when Buffett said return on equity capital he referred to ROE, but ROIC is a bit different. Would the above calculations and math still hold for ROE?

    For example if a company had a ROE 20% and can reinvest 100% of their earnings, and earnings will grow at 20% over time, then will intrinsic value of the business also approximate this 20% annual growth rate?

    Again Great Piece,


    P.S. I do remember Munger back in the day also mentioned

    “Over the long term, it’s hard for a stock to earn a much better return that the business which underlies it earns. If the business earns six percent on capital over forty years and you hold it that for forty years, you’re not going to make much different than a six percent return – even if you originally buy it at a huge discount. Conversely, if a business earns eighteen percent on capital over twenty or thirty years, even if you pay an expensive looking price, you’ll end up with one hell of a result.”

    Again, I was unsure if he meant ROE or ROIC
    (Buffett would refer to ROE as return on equity capital)

  6. John,
    Great article. I learn a great deal every time I read one of your posts. This maybe a little off this exact topic but I believe it was Jeremy Seigel who mentioned that Altria (MO) was one of if not the highest returning stock of the 20th century. This is a company ( as well as its international brother PM) that returns a great deal of cash to shareholders because they have no way to keep reinvesting in the business other then maybe acquisitions. Now I don’t want to go through the exact math of taxes, div reinvestment, div % etc but do you consider this a superior or inferior method for shareholders to earn a return?

    It seems these companies are able to return cash to shareholders (via dividend raises) on average in the 8-12% range without share buybacks and in 11-15% range with (total shareholder yield) outside of any additional increase in the actual price per share. It seems in this case since they cannot reinvest in the mature business itself other then general maintenance etc they are increasing the “intrinsic value” of the business synthetically via the dividend/share buybacks. Granted the dividend in this example decreases the total return in a taxable account but this strategy seems to have worked very well as Siegel explained. I guess my question really comes down to…is this a valid comparison? and if not how do you explain such high share holder returns for tobacco companies such as MO and PM over the last ~50 years?

    Also this maybe a bit of a basic accounting question but in PM’s case where this is negative equity due to buybacks does one just add back the treasury stock to figure out ROE?

    The TTM ROC Greenblatt % is 146.96 for PM and 377.78 for MO, both ridiculous numbers with the ROE’s being very high also.

    Thanks in advance as I have thought about this for a while.

    1. Hi Matt, I haven’t spent much time looking at the tobacco companies, but yes, the reason the return on capital the way Greenblatt measures it is simply because their balance sheet has very little tangible capital (Greenblatt uses working capital less cash plus net PP&E for the denominator). They have a lot of depreciated property, a lot of goodwill, and it appears they’ve levered up the balance sheet by taking on debt and buying back shares (which has reduced the equity into negative territory). Greenblatt’s formula is a decent rule of thumb, but it’s not always the best way to analyze each company. Some companies have a business model that relies on acquisitions, and in those cases I’d factor the cost of those acquisitions (and the goodwill associated with them) into the return on capital calculation.

      For PM, I haven’t spent any time looking at it (other than a quick glance at the financials), but they appear to be a cash cow without much reinvestment opportunity. When trying to evaluate how good the business is at producing earnings from its capital base, I’d probably look at the total assets, or maybe the total assets less non-interest bearing liabilities (which act as a form of 0% short term loan, thus reducing the amount of capital the business needs).

      I probably wouldn’t pay much attention to a tangible capital number like Greenblatt’s formula, as the denominator is so low (or even negative in the case of ROE) that it is somewhat non-meaningful. But certainly, the business produces high returns (20%+ ROA).

      I think it’s always good when a business is producing these types of high returns. However, I also think that a business that needs very little capital can produce really high returns on the capital that has already been invested into the business, but that doesn’t necessarily mean that it has lots of opportunity to reinvest capital at those rates going forward… and if they can’t retain earnings and reinvest at high rates, then they may not be able to grow earning power.

      One exception to this general rule of thumb: some businesses can grow earning power without using more capital. But these businesses are somewhat rare. I’d always be on the look out for increases that are accompanied by higher debt and higher payout ratios (which might mean the situation won’t last), but in the case of the tobacco companies, they’ve had a long history of producing steady, if not fantastic, shareholder results.

      For businesses that pay out all of their cash and have little need to retain earnings, I typically think in terms of earnings yield. If the stock is trading at a P/E of 12 (or a cash earnings yield around 8%), then this is about the best return you can expect if the business is paying out all the cash in the form of dividends and buybacks and can’t find places to reinvest them within the business.

      Thanks for the comment… those are just some thoughts. Not sure if I answered your question, but those are just some ramblings that come to mind.

  7. John,
    I know we have talked about this before, but I thought I would put it out there to get some comments from the peanut gallery. In a discussion of return on invested capital, I think there is a lot of focus on the new dollars that are coming in and where they go. However, if the dollars that are already invested in the company start to earn lower returns than historical averages, that is technically a larger portion of capital and its return should matter…

    For example, think about McDonalds (mainly because it is simple). If MCD invests a new $1m and earn $150k in the first year, then you have 15% ROIC, and MCD keeps up this process by investing another new $1m each year. However, after 5 years, the first $4m of invested dollars start to earn lower returns, say 10%, then the 15% return in the first few years was really just pulling forward a lot of the return and perhaps the investment will not even earn back the capital invested in the project. That’s cumbersome to point out, but lots of investments by companies show high initial returns on invested capital only to slowly peter out.

    I think implicit in the math of ROIC is that the “old” invested capital continues to earn the high returns and I am not sure this is always the case.

    1. Matt, good points. I think you need to look at the invested capital (the capital already on the balance sheet) to determine the rate of return on the “old” capital. If MCD is investing in new locations that produce high returns initially, but low returns eventually, then those lower returns will show themselves in the return on the capital the MCD has already invested in the past. But yes, it’s true that you want to know how much capital a business is using, and just as importantly (in terms of growing earning power), what can they retain and reinvest, and at what rate of return going forward.

      A business that is producing high returns on individual projects initially, but mediocre returns overall will have a lower ROIC number. Also, a business with a high ROIC number with very little ability to reinvest earnings at that same rate of return will simply be producing stable cash flow on a stable capital base, but won’t grow earnings.

    2. Matt,

      You make a very good point. IMO the shale companies are experiencing this lower return on ROIC. Every month/year they must reinvest new capital at ever increasing rates to increase production. Yet, each new well drop production by 63% on average in the first year and near 90% after 3 years.

  8. John,

    You really hit the nail on the head. IMO this is really one of the most important concepts in investing yet the least understood or appreciated. It amazes me sometimes when people get excited over a company trading at deceptively low p/e multiple yet has horrendous ROIC.

    If your buying business with excellent returns on capital for reasonable prices even if your not right about reinvestment opportunities you have still probably made a decent investment. However, if you are right it could very well be a home run.

    What comes to my mind is Buffet’s investment in Coca Cola. High return business with a large opportunity to put a lot of cash to work. This is forgetting the ridiculous multiple that followed.

  9. Thanks John for great article!
    One question I hope you could help with. It is about deal with goodwill when assess ROIC or ROCE. Some companies constantly, repeatedly do M&A during a long period time. M&A strategy is a part of their business model. (They need M&A to help maintain their competitive advantage and growth momentum). As a result, they accumulated a huge amount of goodwill in BS. This is where I get confused. You have stated before, what’s matter, theoretically, is how much return we can get from every new/incremental capital invested. So goodwill belongs to the past, you don’t need reinvest, in theory. Then we don’t include goodwill in calculation of ROCE. Also Greenblatt also doesn’t take account goodwill according to your previous post. But I really doubt it for the type of company I mentioned. What I see is they, on a recurring basis, do M&A, and will continue. So from incremental capital perspective, I do think part of goodwill need to be considered. But among this huge goodwill accumulated/sit in Balance sheet, I also don’t think, all of them should be considered for the reason, you don’t really need to reinvest that much goodwill to maintain current competitive advantage. But then I’m confused, how should I do my adjustment in assessing this important return. It makes so much difference ( completely excluding goodwill, you can get 100% ROCE, full included, you get only 10%….).
    Anyway, I hope you could shed some lights for me. Thanks!

    1. Nina, the real concept to keep in mind is to think about how a business invests its cash, and what return are they getting when they put out that cash. I think Greenblatt excludes goodwill because as a new investor, he’s not paying for previous acquisitions (you don’t have to replace goodwill). However, if a company’s business plan is to consistently acquire other businesses, then I think you need to consider those acquisitions as normal investments, and you’ll need to consider the goodwill and intangibles into your capital calculations. Thanks for the comment Nina.

  10. Thanks for the fantastic article John. I just have a quick question regarding your reinvestment rate assumptions. I know your examples were simple, back of the envelope type calculations, but if we wanted to use this methodology for actual investment analysis, then how would you go about projecting a companies reinvestment rate for the next 10-15 years? I know there is no single correct answer for this, but can you clarify how you decide if a company is going to reinvest at a 50%, or 100% rate? This likely varies depending on what industry or specific company you are looking at, but could you shed some more light on this generally?

  11. John,

    Fantastic post as always. Many thanks for sharing your insights.

    Dont you think the post should be titled “the importance of growth”? High returns are a necessary precondition for high growth but the differential in CAGR in your examples is mostly driven by the growth differential.

    Warm Rgds, Yogi

    1. Hi Yogi, thanks for the comment. Yes, basically, if we can locate a business that can produce high returns on its invested capital, and can retain those earnings and reinvest them back into the business, we get growth (in earning power and intrinsic value). Growth that creates value is great… obviously, as you know, not all growth is valuable, but if a business can reinvest earnings into the business at high rates, then growth will create enormous value. Like a savings account that pays 20% interest that compounds and allows you to reinvest interest at the same rate.

  12. Excellent post as always and looking forward to the next one!

    Simple question, what about the effect of share buybacks?
    If share base keeps getting smaller and smaller, company will be able to keep 20% roic without growing earnings. Isn’t this a trick to keep roic high?

    1. Reading back article helped me understand 🙂
      Yes Roic will stay high but low reinvestment rate will diminish intrinsic value. Basically your early example with dividends applies to share buybacks as well. Assuming that shares are bought back at fair value.

    2. Hi Alex, I wouldn’t say it’s a trick… some businesses just don’t have the option of reinvesting earnings at high rates. Some produce oodles of cash that can’t be reinvested and thus should be used for dividends or buybacks. And yes, this reduces the capital base, but it’s not necessarily a bad thing. It’s just that the business won’t compound earning power as fast as one that can use those earnings to reinvest at the same rate of return (all other things being equal).

  13. I agree with the idea behind this article although I’m not sure about the presented math. I found which returns include vs exclude dividends confusing. The ROIC idea intuitively explains why great business models include the scalable consumer monopolies, the premium brands, the low-cost providers with scale advantages, and hedge fund general partners.

    The way I look at it is, when you buy you have an immediate “buyer’s return on capital” due to the margin of safety over the number of years (typically 3-5) to converge to intrinsic value, minus the taxes you will incur when selling. Thereafter you have a compounding return, due to factors such as ROIC, ROE, moats, willingness of management to screw things up, optionality in different business sector growth, etc. So you can just draw these as a graph of return on capital versus year. You take the geometric average of your initial return on capital with the compounding return on capital for however many years you held to get your after-tax return rate.

    One thing I don’t understand is if ROIC is so good then why don’t backtests like they do at Greenbackd show it as more important? They generally show quality factors such as ROIC as adding some return, but it is far less significant than the raw value factors such as “book/market.” Now most of these tests assume 1 year holding periods, but still 10 years is just the multiplication of 10 one-year holding periods (but tax advantaged due to the long holding period of course).

    My belief from working in tech and also academia is that most managements do not think at all in terms of rate of returns. Managements are nearly entirely devoted to squabbling over spending money, political fiefdoms, getting the most power or resources, maximizing their options which typically reduce return on capital, buying back stock at high levels (when rationally they should be doing a dilution arbitrage, so that investors who bought at rational levels would receive a positive return of cash provided by those who irrationally buy into bubbles), not buying back stock at low levels (when rationally they should be buying, to arbitrage the other direction), etc. So I think the natural compounding potential of a business is very frequently diluted by this “human nature” failure of everyone rowing in different directions and trying to screw each other over.

    I’m not sure how to identify the best managements who have carefully set up the right incentives and have the proper ownership structures to prevent the typical “squander and waste everything” human behavior. So I generally rely more on a margin of safety.

    But interestingly even in personal finance most people err on the side of “squander and waste” — such misbehavior is by no means confined to corporate misbehavior.

    For example, given the prevalence of car ads I would guess that people are allocating far too much money to new car ownership. Generally cars can be bought used, driven for a couple hundred thousand miles, and parts that break can be promptly repaired. You just do the math on expected capital expenditures in a spreadsheet. But to really optimize the capital structure of a car properly, there are whole subcultures such as “car longevity,” where some people have driven Volvos and Mercedes cars up to 3 million miles:

    What admirable rationality!

    Logically I think one should be able to find similar telltale “fingerprints” of rational corporate managers. Including frugality, careful books, and relentless use of capital structure optimization. I just haven’t learned to identify these fingerprints generally. I’ve seen a few faint indicators such as insider buying at a fixed low price point with high insider ownership percentage. And obsession with restructuring to avoid taxes, as John Malone and Warren Buffett do. But I would like to have a better understanding of the psychology and “tells” behind such (infrequent) rational behavior.

    1. Thanks for the comment Connelly. Thought provoking, as many of your comments are. One quick thing I’d point out about the backtests (Greenblatt’s specifically). Those use ROIC as a number that provides you with the following general info: how good the company is at producing earnings on the capital it has invested in the business previously. In other words, most ROIC numbers (earnings divided by debt+equity) tells you about the company’s earning power relative to the capital it has invested up this point in time. But it doesn’t tell you about whether the company can retain those earnings and reinvest them at the same (or similar) rates of return. Some businesses use very little capital, and thus might have very high ROIC’s, but don’t have any place to reinvest that capital (think See’s Candies), and thus the business is still a good one, but it lacks the ability to compound earning power by retaining and reinvesting back into the business at high rates of return. These cash cow businesses are still good, and they still might be able to grow over time without using much capital, but their earnings will have to be used for other things such as buybacks, dividends, or acquisitions. Otherwise, capital would pile up in the business and ROIC would gradually get reduced.

      So the best businesses are the ones that produce lots of earnings on its invested capital, AND, can use each year’s earnings to invest back into the business at similar rates of return. This is the true compounding machine, because capital is growing, but ROIC is not decreasing, thus earning power is compounding…

      So ROIC numbers in a screener don’t always separate the businesses that return 100% of their free cash flow to shareholders due to lack of reinvestment opportunities. Again, these businesses can still be great businesses and investments, but all things equal, they likely won’t compound as fast as the one that can reinvest.

      1. That’s a good point.

        Nevertheless I should think if high-ROIC companies returned their capital in a tax-efficient manner to shareholders when it could not be reinvested, we would see much more compounding going on in the marketplace than we currently see. Also the amount of AUM and returns for activist investing would be much lower.

        I think this is also related to the iterated prisoner’s dilemma and super-rationality, as discussed on Wikipedia:'s_dilemma#The_iterated_prisoners.27_dilemma

        If you can trust everyone in a corporation to act in a manner of cooperation and enlightened self-interest, then it always makes sense to retain earnings, re-invest, invest your own savings in the company, and cooperatively compound. But it only takes a few bad apples to spoil the cooperative strategy, whereupon everyone reverts to ordinary (non-iterated) game theory rationality. Here the dominant strategy is for management to maximize their own bonuses and cash out before everything blows up.

  14. Great article, but could you provide some additional clarity on how you estimate a companies reinvestment rate going out 10-15 years? Do you use a dynamic rate going forward or just a flat rate for each year? It just seems very difficult to be confident in ROIC or reinvestment rate adjustments going out over such a time period, where small changes in assumptions would produce outsized effects on your projected earnings power. Thanks

    1. Hudson, yes those are good points, and all true. It’s difficult to project out that far. In fact, it’s one of the reasons I don’t use DCF’s. I don’t like getting a false sense of precision. This post was more about the “why”… in other words, the reason that ROIC is important, and the math to demonstrate the power of a business that can reinvest earnings at a high rate. As far as going forward, it will depend a lot on individual analysis of each situation. That is the art part of the equation, and tough to explain because each business is different. The general idea is to look for businesses that are durable. That is one of the most important things I think about. To help me, I like to look and see how the business has done in recent economic cycles (2008-09 is a good “stress test”). But of course, the track record doesn’t guarantee the company will continue producing good returns going forward. But it helps. I spend a lot of time thinking about how the company will compete and how their returns might be diminished, and in the ideal situations, I find it hard to imagine a scenario where the subject company’s durability will deteriorate.

      So there is no real formula to that aspect of investing, at least not for me… it’s really about checking the historical track record, identifying a company producing good returns, and thinking about the probability of those returns continuing.

      Hope that helps a bit…

  15. Hi,

    Great post. A question- ‘How do you calculate return on incremental invested capital’. Also, the link to ROIC and Greenblatt part is not working and returns ‘404’ error.

    Always look forward to your posts,

    Thanks again,


    1. Hi Pankaj,

      The returns that a company gets on the capital it invests incrementally is not always easy to determine. It takes an understanding of the business to know how it uses retained earnings (where are they investing capital each year and what returns are they getting on those investments?). There are relatively simple ways to calculate the yearly returns that a business is getting on the capital it has already invested (some investors use a formula that Greenblatt has espoused–EBIT divided by (Net working capital less excess cash plus net PPE)… I tend to use net earnings and compare it to equity and debt capital–depending on the business, you may want to adjust for things like goodwill, intangibles, etc… to determine an accurate picture on the capital that a business will need to use going forward). But the concept is that a business value will grow at the rate of its ROIC times the reinvestment rate. So the amount it can earn on invested capital is important, but just as important is how much of those earnings can it invest going forward at similar rates? If it produces high earnings on invested capital but can’t reinvest anything, it will produce a lot of cash flow but the business won’t grow.

  16. The light bulb has been flickering for a while, but thanks to this post it is now fully on.

    Fantastic post. Appreciate the clarity.

  17. Hi John,

    just stumbled upon your blog recently and great content! In regards to ROIC, do you have any opinions of a company such as IBM whose ROIC is constantly above average?


  18. John,

    What just revisiting this post after a couple of months and have a question on the following: How can you reconcile Buffett’s comments regarding (1) the benefits of businesses that are able to reinvest a large portion of your earnings at high ROIC versus (2) the benefits of a business which require very little reinvestment to keep growing. The two quotes by Buffett are:

    1. “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. A business is also wonderful if it takes money, but where the rate at which you re-invest the money is very satisfactory.”
    2. “Lord Thompson, once he bought the paper in Council Bluffs, never put another dime in. They just mailed money every year. And as they got more money, he bought more newspapers. And, in fact, he said it was going to say on his tombstone that he bought newspapers in order to make more money in order to buy more newspapers [and so on].”

    It seems that in the case of the second business, if you are not reinvesting anything back into it, given your formulas above, then the incremental return on invested capital and therefore earnings and thus intrinsic value growth will be low no??

  19. John,

    What just revisiting this post after a couple of months and have a question on the following: How can you reconcile Buffett’s comments regarding (1) the benefits of businesses that are able to reinvest a large portion of your earnings at high ROIC versus (2) the benefits of a business which require very little reinvestment to keep growing. The two quotes by Buffett are:

    1. “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. A business is also wonderful if it takes money, but where the rate at which you re-invest the money is very satisfactory.”
    2. “Lord Thompson, once he bought the paper in Council Bluffs, never put another dime in. They just mailed money every year. And as they got more money, he bought more newspapers. And, in fact, he said it was going to say on his tombstone that he bought newspapers in order to make more money in order to buy more newspapers [and so on].”

    It seems that in the case of the second business, if you are not reinvesting anything back into it, given your formulas above, then the incremental return on invested capital and therefore earnings and thus intrinsic value growth will be low no??

    1. Yeah Juan, some businesses have no use for additional capital. That doesn’t mean that they can’t grow. Some businesses can grow without any additional capital. Moody’s is an example of a business that pays out virtually all cash flow (it doesn’t need to retain anything) but can still grow. This is because of the pricing power of the business. See’s Candies is another example of a cash cow that spit out far more cash than could be consumed and reinvested. Charlie Munger once was asked about turning Sees into more of a franchise model and he responded by saying that Sees already produced far more cash than it could retain. Same with Lord Thompson’s newspapers. They produce significant cash flow, but each individual paper had very little reinvestment opportunities. They were just milked for cash. I wouldn’t say intrinsic value is low… there is a certain intrinsic value of each of those papers, and that value would be approximated by totaling the present value of all future cash flows, but there is very little growth and little opportunity to retain and invest cash at high returns. Some businesses can grow without capital (essentially the return on capital is infinite in this case), but these are rare birds.

  20. Hello John,

    Thank you so much for writing this awesome article explaining the power of compounding companies. I do have a question regarding the CAGR you calculated for company B. I cannot come up with the same number and I would appreciate a lot if you can show the math on how you got those numbers. For example, in the first table, column: Co. B CAGR, Buy P/E = 10 and row Ending P/E after 15 years = 12, your answer is 15.8%. How did you come up with that? When I do the math 1 eps x 50% payout = 0.5 eps. Buying at P/E = 10 would result in a dividend yield of 5% ($0.5/$10). The cagr from paying 10 x 1 EPS and turning into 12 x 4.17 EPS would result in CAGR of 11.33%. So the total return would be 11.33% + 5% = 16.33% but you have 15.8%?? Can you please explain this John. Thank you so much for your help!

Leave a Reply

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