Guest post: Investment Theory: RoCE – some thoughts on a helpful, but sometimes misleading concept

I am happy to present one of the infrequent guest posts. This time a very interesting general post on RoCE (Return on Capital Employed) and Brand value by contributor Knud Hinkel.

Executive Summary:

The RoCE is an important ratio for value investors. However, as it regularly relies on balance sheet data, the concept is susceptible for at least two inconsistencies: (1) Items on balance sheet are not correctly reflected in the EBIT, and (2) items that contributed to EBIT are not reflected in the capital employed. My hypothesis is that the RoCEs of industries with significant self-created intangibles like consumer and software companies are subject to a systematic upward bias and this might light lead to a wrong judgment of (1) the underlying company performance, (2) acquisitions, and (3) the capital intensity of the business model.

Full post embedded below:


  • Assuming you get somehow to an estimate of the intrinsice value of the operating activities (multiples, DCF, replacement cost, …), would you add the brand value to your valuation? or would that be double counting, as the brand value is already reflected in higher margins/better ROIC? Maybe use it just as an additional margin of safety?

    • The latter it is, in my view. The brand of companies like Adidas is an indispensable part of the business model and you can’t really it separate from the rest of the operations. What would be Adidas worth without the brand ? Probably not a lot. As you said, the “invisible” brand asset is already reflected in higher returns on the “visible” assets on balance sheet. If your valuation already accounts for that, a brand does not provide an additional MoS.

  • Just out of interest I looked at Adidas, a company which clearly lives from its brand value. Accoring to the link, Adidas has a brand value of around 7 bn EUR. Over the last 10 years, Adidas had a ROCE of around 11% on average. A quick and dirty calculation resulted in around 10 bn “capital employed”.

    So including Knud’s brandvalue, the “true” ROCE would have been around 7% or so.

    Hard to understand why people are paying ~30 earnings for such a stock..

    • -Adidas has 2679 stores. Due to their retail activities they also have increased employees and capital expenditures. Those activities should lower the core ROCE. I believe the “Adidas core” has high ROCE. Maybe new management will show this.
      -Adidas increased sales and marketing “investments” (expenses) in 2015 compared to 2014 heavily.
      -EPS for the third quarters was already €3.36 and even higher excluding discontinued operations. –> This is not trading as high as 30x earnings.
      -return on bayern münchen under the line. Asset value is 80 mio. Is a minor point but should be adjusted for.

      • Actually I forgot to include the stores this would increase the employed capital by ~2 bn for the operating lease liabilities. Funnily enough I found a very interesting passage in the annual report, I am not sure if one would need to add this to the invested capital as well:

        Die finanziellen Verpflichtungen aus Promotion- und Werbeverträgen erhöhten sich im Geschäftsjahr 2014 um 37% auf 5,193
        Mrd. € (2013: 3,791 Mrd.€). Hauptgrund hierfür war der mit Manchester United F.C. geschlossene langfristige Promotion-Vertrag

        • I think I would add it to capital employed if it’s a discounted number, and therefore qualifies as an interest bearing liability. Otherwise, it should be fully netted against the corresponding asset, i.e. ManU contractual obligations (to play football in a German jersey, the conditions must have been very generous).

  • A stable company with a great consumer brand typically spends on advertising every year. If one were to capitalize the expense this would not only increase capital employed but also increase the return because this lowers expenses. It is not clear that this would lower ROCE meaningfully. Those companies have indeed good ROCE.

    • True. On the other hand, we have also take into consideration some kind of “amortization” as the carry-over effect of advertising will not last forever. I would argue that younger companies’ advertising may very well exceed “amortization”, while in more mature industries, this relationship is probably more balanced, and the positive effect you described is less certain.

  • Another excellent post, thank you! A few observations (probably wrong):
    1) By capitalising R&D/Brand/Leasing you surely underestimate the compounding ability of a company? As surely it is the Return on Equity (RoE) that ultimately determines the equity owner’s final payoff. If you did this with Coke or Starbucks 30 years ago I am guessing you would have concluded these companies are fairly unexciting….
    2) By capitalising R&D/Brand you are obfuscating a DCF further and increase the likelihood of garbage in, garbage out
    3) Should you not lower your discount rate as these forms of capital are lower than debt with covenants that can’t be scaled?

    • Thank you for your comment!
      1) I would not say so. The equity on balance sheet also neglects the value of self-created intangibles. In fact, I would consider the “underlying” equity to be much higher for these kind of companies. This is in my view one reason these companies often trade at high P/B multiples and their RoEs look great (and probably are actually great in many cases). Therefore, the return on equity which is adjusted for brands etc. will not be as high as the RoE based on accounting data and can – in extreme scenarios – be lower than cost of equity, in my view.
      2) I did not recommend to capitalise R&D / brands for your DCF. This was just another angle to look at the often formally very attractive RoCEs of software and consumer companies.
      3) I would not do that. As I regard self-created intangibles as “hidden equity”, and the cost of equity is above cost of debt, it should increase, not lower your hurdle rate.

      Does that make sense?

      • Addendum: According to financial theory, the “correct” answer on 3) is “neither increase nor lower your hurdle rate” because as we all know, the only relevant weights in the WACC formula are market values of both equity and debt and they remain unchanged whatever you capitalize on the balance sheet / DCF.

      • Knud,
        Thanks for the thoughtful response, hard to disagree!

        I also presume by capitalising R&D/Brand you will increase the volatility of RoCE as I presume theoretically (?) there should be economies of scale from these ‘investments’. Unlike PPE & W.C where you tend to get a linear return on each additional unit of capital employed…..

        Best wishes,

  • Hm, interesting post. Some thoughts:

    “Though accounting regimes do allow the capitalization of expenditures on buildings, machines etc., they do not allow the capitalization of such a windy thing like marketing / promotion expenditures.”

    I guess there is a good reason for this. While it would make sense in some cases to allow for capitalization, there are cases where this is not true. My fear is, that if allowed officially, some windy managers would take advantage of this, to prop up their results – especially the managers that do not care about ROCE (or shareholders for that matter)…

    “As regards consumer companies, you can do two things in order to assess whether the management has actually earned its costs of capital: Firstly, you can add estimated brand values as proxies to the capital employed.”

    I would not want to do that. This is outsourcing of one of the most important activities of an investor – valuing assets.

    “In contrast to self-created intangibles, acquired brands and acquired technology are allowed to be capitalized according to IAS

    this discrepancy might simply result from the different treatment of self-built vs. acquired brands / software on balance sheet”

    The last sentence is definitely true. The hardest part with acquisitive companies is to find out, how much they really earn. One way, the most simple, is to “clean” the balance sheets of capitalized acquired assets and the P&L of the corresponding P&L-items.

    “make up your own back-of-an-envelope calculation

    With regard to self-created off balance sheet technology, the recommended procedure is analogous”

    This is theoretically sound, but it is a very tough and imprecise thing to do. Why not do the opposite – expensing?

    Expensing is maybe theoretically dirty, but it makes comparisons more consistent than capitalizing because capitalizing is so prone to estimation errors. You get the “raw data” instead of some heady amortization expenses, the money that flows (at least close to). After expensing, the value of these intangible assets can be derived/estimated from the excess of ROCE on a tangible basis (‘tangible ROIC’ a la Greenblatt) over cost of capital (whatever that is).

    The thing is: the valuation of a brand (or technology or whatever) should be based on the earnings you can generate from it, not from the costs/expenditures you threw at it to acquire/develop. At least that’s my opinion.

    These were just some thoughts that immediately came to my mind when reading the article, there are more, but I guess I have to stop now 😉
    Thank you for the post anyway. This is definitely one to think about.


    • Thanks! I think we can easily agree on most of your statements. Again, the intention of my post was not to “disimprove” accounting standards or DCF techniques but to provide a new perspective on the often very high returns in some industries. Furthermore, I tried to substantiate my view that in these industries, some returns are not as great as they seem to be and some intra-industry acquisitions might make more sense than RoCE suggests… The way I tried to clarify my point of view is definitely debatable.

      • I can agree on that 😉

        One more thought that came to my mind:

        “This gives you an estimate of the “net brand value” (= cumulated relevant marketing expenditures minus the cumulated fade effect”

        Isn’t there another problem? If you take the marketing expenses and capitalize these for estimating the brand value: you would overestimate brands of companies with “inefficient” marketing teams and underestimate the brands of “efficient” marketing teams, wouldn’t you? Companies that have to spend more on marketing on a per unit basis would appear more valuable because they have higher marketing expenses… obviously the “efficient” marketing team would be unfairly punished because their marketing is more efficient and so they have lower marketing expenses.

        Is there a way to adjust for such a (possible) effect?

        Thanks again,

        • Sure, the efficiency argument is valid, in my view. But it can be applied to other assets as well: A company can also over- / underpay for buildings, machinery, real estate etc., but on balance sheet you will only see the price it actually paid…

Leave a Reply to Knud Cancel reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.