Bruce Greenwald on Valuing a Franchise: Pt. 5

So, what was missing from our equation? If you guessed margin of safety, you guessed correctly. Value investors want to buy $1 bills for 50 cents, not for 89 cents.

To incorporate a margin of safety we want to include a term that is:

Probability of Something Going Wrong x Impact If It Does

Now, before Greenwald would have even begun forming the expected return equation he would have not only read the 10-K, he would have computed the firm’s asset reproduction value (ARV) and earnings power value (EPV). Part of the verification that we’re working with a franchise is to look for a company earning far more than just cost of capital on the assets they’re employing.

If a firm that previously had a competitive advantage loses it, what happens? Theoretically, EPV collapses to ARV. I haven’t computed ARV and EPV and it would probably take another 5 posts or so to properly attempt that. So again I’m going to take a shortcut Greenwald probably never would and say, if IBM has a competitive advantage now, and loses it later, I’d expect its P/B to drop from the current 10 to 2.5 (a 75% drop).

And I’m going to say there’s a chance of that happening sometime in the next 15 years. There are concerns that management might not be re-investing enough back into the company to set it up for long-term success. I’ve even heard their Roadmap 2015 plan referred to as Roadkill 2015! So I’m going to assign an MOS factor of:

(1 in 15) x (-75%) = -5%

Updating our equation…


Our expected return is now 12% and includes an allowance for something possibly going wrong. Greenwald says that Glenn Greenberg’s hurdle here is 15%, while Buffett’s is 13%. Magically (or not! if this is indeed Buffett’s approach) we’re estimating a 12% return with the stock trading near the upper range of where Buffett started buying it in 2011.

But the true magic of this whole process to me is in how the equation guides our research. How many times have you read 10-Ks of companies, not really knowing what you’re looking for, but knowing that great investors supposedly spend a lot of time reading 10-Ks?! Guilty as charged.

But if you’re tempted by the 12%, think about how you’d investigate IBM now. Dividends and buybacks are hugely important to what return we’re estimating. We’d begin by looking for any and all guidance the company has issued about what they plan to do here going forward. Have they ever discontinued or cut a dividend in the past? Have they been funding the buybacks unsustainably with debt?

And how about organic growth? Does IBM have a competitive advantage we can put our finger on to insure we get that 5% GDP tailwind? Is there evidence of that recently? Incidentally this where I personally ran into issues with the company that I couldn’t give it a pass on. Shrinking revenue in constant currency the last few quarters despite acquisitions.

The bad news about the process is that it’s very company-specific in terms of what you have to research to fill out your equation with accurate numbers. You have to gather evidence and use your own introspection to draw conclusions. It’s not just pulling ROC numbers from Morningstar and making assumptions of P/B falling from 10 to 2.5 as I’ve done. The good news is that because of this, you can’t be replaced by a computer – and you have an edge over someone else who’s just lazily crunching numbers.

20 thoughts on “Bruce Greenwald on Valuing a Franchise: Pt. 5”

  1. Hi man, let me first say congratulations on compiling this article with such clarity.

    Could you please help me by providing the source of this? Was it an interview or is it in Prof. Greenwald’s book?

    Thanks in advance for your help,

  2. Thanks Brian. Unfortunately I don’t have a particular source I can refer you to. You can google “Bruce Greenwald video” and find hours worth of lectures & interviews from various sites – many of which I’ve studied. Every one is another small piece of the puzzle. He does discuss this approach at a high level in a recent interview when he’s discussing Nestle:

    I do own a few of his books – you’re probably referring to the one on value investing – but I don’t think it’s in there. I’ve heard him say something about that book not being the greatest and that a new version will be coming out soon. Amazon already has a placeholder for it for July 2014, so something to watch for.

    Greenwald is the Socrates of value investing: so skilled at the art of rhetoric that you could set up a sporting debate between us, tell him he is to argue for the legalization of murder and me against, and I would probably lose. I believe and attempt to implement most of what he says to do despite not knowing his actual investing performance record. Caveat emptor.

  3. Return on Capital is Return on Equity in this case. Return on Equity = FLEV x Asset Turnover x Margin. So why do we need an organic growth part if it´s already included in the Margin (=Earnings/SALES)?

  4. I think I understand what you’re saying. I would do it in whatever way makes sense to you but remember that the E[R] you calculate is not what you actually expect to make, in perpetuity. Rather it’s to help you make a binary decision – is the stock obviously cheap or not. If it falls very close to your hurdle rate – you may want to replace your storyboarding with something more precise.

    I myself actually work with per share numbers from Value Line and more with ROE instead of ROC. For a first swag, I just add up EPS retained over the last 4-5 years or so and see how much higher EPS is now than it was 4-5 years prior. That growth is the combined result of organic growth, active growth, and buybacks – all in one number. Debt may have grown since then too.

    But for a swag, I don’t worry about trying to separate it all out. If their return on retained earnings (due to all those factors) is 12%, they pay out 25% in dividends, the P/E is 15, and the Cost of Equity is 10% then:

    E[R] = 1 / 15 x [25% + 75% x 12% / 10%] = 7.5% before MOS, which is about what most safe large caps I look at today work out to: not obviously expensive or cheap.

    A previous reader asked about sources. Here’s an old interview where Greenwald is looking at some companies during November of 2008. See the part on American Express.

  5. Hi, first off congratulations on the website, very informative!

    I am unsure on how to value a company that is funding its cash return to shareholders with debt, resulting in a distribution return of over 100%.

    In such cases, how do we value the company in return space? We account only for the cash return and organic growth, not assigning any value to its reinvestment return?



  6. hi joao,

    it’s a good question and one i’ve pondered myself. i don’t know what greenwald would advise here but what i personally do in that situation is net out the buybacks & dividends with the new debt. so if a company gave us 100M in buybacks and dividends but took on 20M in new debt, i’d call it an 80M cash return since that looks more like what would be sustainable in the long run.

  7. Thanks for the clarification, it makes sense. Does this mean that on your IBM example you too should subtract net borrowings to your dividend+buybacks calculations?

    If so,

    IBM net income (3y) = $48B
    Dividends+Buybacks-Debt (3y) = $39B

    Cash Payout = 39/48 = 81% (instead of 94%)

  8. That’s a good point Joao. I usually net out the new debt only for cases where buybacks + dividends exceed net income. My thinking is that many companies increase debt (sustainably) as they grow – so I don’t want to do the netting out then – but a payout greater than what is being generated is a situation that can’t persist. That being said, there are situations where it can persist (companies where FCF is usually larger than NI).

    I think the important thing to remember is that the formula can be used with very rough numbers just to first get an idea as to whether the company is obviously cheap, obviously expensive, or in the gray zone. If it’s in the gray zone (I’d put IBM there) you’re going to have to familiarize yourself with the company: what marginal return on capital have they been generating lately? can they continue increasing debt like they have been? is the dividend in jeopardy of getting cut? is their real cost of capital something far different than 10%?

    As an engineer I have to be on constant guard not to replace reading & thinking with math.

  9. What if the payout after discounting for new debt is still above 100%? Like TUP in the last 5y…

    Net income = 1086
    Dividends+BB = 1545
    Debt issued = 392
    Payout ratio = 106%

  10. hi joao – it looks like that’s been possible with TUP because FCF has been almost 100M higher than NI over that period. that might be sustainable if depreciation is higher than maintenance capex. some companies could also look this way due to spending down their cash, so for completeness you’d want to look at change in NET debt, not just debt, in case i didn’t mention that. (that doesn’t look to be the case with TUP). also – not sure where you’re pulling the data from – i see a small error in Morningstar’s 5-yr CF/S for TUP (missing one year’s small new debt issuance).

  11. Indeed, I was looking at net debt, not debt issued – for the 5y period debt issued was 406, change in cash was 15; net debt issued = 392. Sorry.

    I understand what you are saying – companies with higher FCF than NI could return more than 100% to shareholders, since they are returning cash, not earnings. Wouldn’t it be right to assume that the payout ratio should more accurately be calculated as (dividends+buybacks-net debt)/FCF ?

  12. That certainly makes sense to me. You might run into the issue of FCF being much lumpier than earnings so it might be better to sum up all FCF and all NI over the last 10 yrs to get a sense of how much higher (or lower) FCF tends to be on average vs. NI, and just apply that scale factor to the current P/E. (This is my approach by the way, haven’t heard Greenwald talk about this).

    And speaking of P/E, if yo use that you obviously want to make sure you’re using “normal” earnings since the TTM P/E might not be typical. Greenwald says to compute the average margin over the last 5-10 years and apply that to the TTM Sales to get “typical” earnings right now.

    I personally wonder why he doesn’t take an EV/NOPAT approach instead of P/E since he seems to be big on EV in other parts of his lectures. I can only guess that he’d say, “look – forget about the exactness of the equation – you’re just trying to see if the company is OBVIOUSLY cheap or not”.

  13. Hey, it’s me again! I was thinking about growth rates and their role on Bruce’s valuation method. If a company is able to consistently generate higher returns than its cost of capital, shouldn’t its expected return be higher as growth rates get higher? However Bruce’s formula doesn’t seem to account for a company’s historic (and expected) growth rate. Note that this growth rate relates to he company’s business, not GDP. It should be incorporated in the third part of the equation, maybe subtracting growth to WACC (although this wouldn’t show the value destruction arising from reinvestment growth at lower rates of return than cost of capital).
    Any thoughts?

  14. Hi, I was thinking that when calculating “Cash” and “Growth” returns, Bruce doesn’t seem to account for the company’s growth rate.
    Considering 2 companies (companies A and B):
    Both have the following metrics:
    – Earnings Yield = 10%
    – Payout Ratio = 80%
    – ROIC = 15%
    – WACC = 10%
    – Organic Growth = 2% (they both compete in the same industry)
    – Company A has an average 20% growth in sales and bottom line YoY for the last 10 years. Company B has no growth.

    When computing the franchise value for these 2 companies, they both have an expected return of 8% + 3% + 2% = 13%.

    Shouldn’t company A have a higher valuation because it is able to reinvest a higher amount of capital at those attractive returns? As Bruce says: Growth only matters within the franchise. When a company is growing and maintaining higher returns than cost of capital, growth creates value.


  15. i think maybe the missing piece is that the ROC you should use is their marginal ROC (not accounting). if company B is retaining and reinvesting capital but getting nothing to show for it, their marginal ROC is 0%, and therefore organic growth is 0.

    here’s a simple example for 2 companies, both with a durable competitive advantage
    – Earnings Yield = 10%
    – Payout Ratio = 80%
    – WACC = 10%
    – Same store sales (passive growth) = 2% per year
    – Company A recently earning 15% active return on reinvested capital
    – Company B recently earning 0% active return on reinvested capital

    Expected Return for Company A:
    10% x [.8 + .2 x 15% / 10%] + 2% = 13%

    Expected Return for Company B:
    10% x [.8 + .2 x 0% / 10%] + 2% = 10%

    To get the marginal ROC you might examine how much in earnings they’ve retained in the trailing 5 yrs or so, plus factor in new net debt, and see how much higher earnings are today vs. 5 yrs ago.

    One other thing – the WACC you should use is also the marginal WACC. So if a company has historically generated their earnings in a safe developed market but all their new active growth is coming from expanding into a riskier emerging market country, the WACC used should reflect the riskier nature of those new cash flows.

  16. Thanks for the input.
    I am sorry once again for my imprecision. When I said ROIC I should have said ROIIC.
    However, the question remains: What value should be ascribed to Company A’s 20% 10Y average sales (and bottom line) growth?
    All things being equal, company A should be worth more because it is able to compound retained earnings over a growing capital base…

  17. hi joao – apologies but there’s something i’m just not understanding in your question!

    i assume you’re using the 10Y past growth as an estimate for future growth, right? – since for current value we only care about future cash flows

    company A’s growth is worth more, hence its expected return is higher (13% vs. 10%) despite being at the same E/P. the value of the growth is the ROC/COC.

  18. Very nice piece! In my opinion the key problem with Bruce’s formula is the assumption on the ROIC and organic growth rate in the long run. In the current environment both input variables are extremely important as earnings yields are low (high valuation).

    However, I believe very few companies have a ROIC that is structurally higher than the COC. I believe this is the case for Unilever and Nestle. See this article on Nestle ( and Unilever this article on Unilever (

Leave a Reply

Your email address will not be published.