A Deep Look at CROIC, ROIC and ROE

March 2, 2010 | Comments (20)

Forget ROE and use ROIC

What is CROIC

CROIC

Return on your investment (ROI). No, it is CROIC | pbarnhart_cedarpark

Most people use ROE (Return on Equity) as a measurement of performance but ROE has a big drawback.

ROE = Net Income / Book Value

As you can see, book value is the denominator which means that if book value was to be reduced, the ROE would in fact increase. How would this happen? If a company writes down any of its assets, the book value would immediately decrease which results in a higher ROE.

The same would happen if the company increased its debt since book value is calculated as assets – liabilities.

The company didn’t perform any better, yet you are given a false picture. So ROE isn’t that great to use. Not even in a screen because you have a high chance of getting value traps.

Instead, ROIC (Return on Invested Capital) is a much better alternative performance metric to find quality investments as it measures the return on all invested capital , including debt-financed capital. It is the effectiveness of the company’s employment of capital.

ROIC is a lot of math but luckily it’s simple stuff without any of the Greek symbols that make your head spin.

ROIC = NOPAT/Invested Capital

For the numerator (the top part of the fraction)

NOPAT = Net Operating Profit After Tax = Operating Income x (1 – Tax Rate)

The denominator (a variation used by F Wall Street)
Invested Capital = Total Equity + Total Liabilities – Current Liabilities  – Excess Cash
where
Excess Cash = Total Cash – MAX(0,Current Liabilities-Current Assets)

Update: I’ve updated the formula because I was coming across a handful of companies where the calculation for invested capital was coming out incorrectly. Mainly due to the role excess cash was playing.

The value was coming in unsustainably high and so it has been modified back to a standard definition so that you can apply it across a range of companies and industries without seeing huge differences.

Invested Capital = Total Equity +Short Term Debt + Capital Lease Obligations + Long Term Debt

CROIC Cash Return On Invested Capital

An even better metric that you can use is CROIC which I use to great effect in my stock value spreadsheets. CROIC is a.k.a CROCI but they both stand for Cash Return On Invested Capital.

You can also check out the CROIC ROIC screen backtest to see how it performs.

CROIC shows how much free cash flow per dollar the business generates from invested capital. I find this to be the ultimate performance metric as it shows so clearly how effective management is and the strength of the business.

CROIC = FCF/Invested Capital

The higher the CROIC, the more cash the company is generating and it also indicates that the business is a profitable one. Rarely do you see a high CROIC but low or negative FCF.

Of course you can use owner earnings instead of FCF.

How to Use CROIC

Until recently I’ve been looking for companies with high CROIC, usually above 10%, but with my latest screen and backtests, I’ve concluded that it isn’t the level of CROIC that is important but whether CROIC is increasing.

This makes sense because if a business has a CROIC of 15% in year 1 but then in year 2 it drops to 12% followed by a drop to 10% in year 3, the average is 12.3% but the picture is different to why I first liked the company to begin with.

Instead, if CROIC is low or even negative, assuming CROIC was to increase, it indicates that management is getting things back on track which will make for a much better investment.

This is something I’ll be going over I’ll be going over in the posts to come. I’ll detail the strategy and performance of stock screens based on ROIC and CROIC.

Update: CROIC strategy backtest has been performed.

About Jae Jun


Jae Jun is the founder of Old School Value. He is on a mission to provide practical and actionable value investing tools, tutorials and educational material to help empower the individual investor. Keep in touch with Jae via any of the methods linked below.

  • Ozz

    Hi Jae,

    Great site, I really enjoy you insight on value investing! I have a question: I was looking at the AAPL analysis, and I was wondering if you could explain your process in determining DCF intrinsic value. Specifically, it seems you initially used CROIC for the DCF growth rate, but opted for FCF because you mentioned CROIC is a “performance” (resets every year) measurement rahter than “growth”.
    So do you use the CROIC in the initial screens when looking for ideas, and then if the company has a good CROIC performance, move into more in depth analysis and calculated DCF using growth rates?

    Hope that made sense.

    Thanks!

  • http://www.oldschoolvalue.com Jae Jun

    Yup made plenty of sense.

    Yes I did initially use CROIC and for the reason you stated. I use CROIC as well as many other variables in the initial phase. I’ve gotten to the point where I don’t look at metrics all the time. I learnt to read, evaluate and interpret financial statements to get a picture and good understanding of the business.

    The metrics are supplementary data. Just because a company has high CROIC, doesn’t mean I’ll invest in it, but it certainly would make me more interested.

    My in depth analysis also relies on financial statement analysis at which point, i’ll have a good idea of what growth rate should be used.

  • http://stockpup.com Sergei

    It is important to use a combination of metrics.

    One problem with CROIC is that a great company in a growing market may quite reasonably choose to reinvest all of its cash flow from operations into expanding its business. That would lead to a low FCF, and a low CROIC.

    The aberrations in ROE can often be mitigated by inspecting changes in shareholder equity. Write downs of intangibles, for example, clearly show up as drops in shareholder equity.

  • http://www.fusioninvesting.com/ Dean

    Great article, but are your explanations for a rising ROE accurate?
    If a company writes down any of its assets, the book value would immediately decrease which results in a higher ROE.
    The same would happen if the company increased its debt since book value is calculated as assets – liabilities.

    My understanding is:
    If a company writes down it’s assets it’s earnings fall so while the E falls the R falls faster.

    If a company increases their debt they generally buy assets or their own share; either way that is a wash.

    Financials are interconnected and considering any one in isolation leads to incorrect assumptions.

    I agree that ROIC is a better measure than ROE and looking at your CROIC screen that appears to be worth looking at more as well. Thanks for sharing.
    .-= Dean´s last blog ..Fusing Business Momentum and Value =-.

  • Robert Crawford

    Jae Jun, nicely presented — especially the adjustment for cash with short-term liabilities.

    One concern with increased expenditures toward future growth is the question of whether such investments will promote growth for its own sake (with a detrimental diluting of the stockholder’s claim on the business). To avoid this issue, it is necessary to determine whether the return exceeds the company’s cost of capital.

    Unfortunately, I’m aware of no easy method for assessing this. Bruce Greenwald performs an assortment of calculations to identify the company’s replication value and its earnings power, with the difference serving as a measure of the franchise value. An easier set of calculations is weighted average cost of capital, but to say this is easier is like comparing the discomfort levels accompanying recovery from a finger amputation versus the same procedure performed on the leg. Neither should be undertaken without alcohol, preferably served by a sexy and alluring creature who is supremely motivated toward alleviating your anguish.

    While it is poorly done, I’ve written about WACC: http://rcrawford.wordpress.com/2008/06/23/cash-return-on-invested-capital-and-weighted-average-cost-of-capital-calculations

    In my defense, it should be noted that this was based on a reader request and was hastily put together.

    Also, kudos to you for suggesting the use of Owner’s Earnings in place of FCF with CROIC. I compare the two over time and find that one or the other may lead or lag directionally on a year-over-year basis. Joe Ponzio indicates a preference for CROIC results exceeding 13%, because this serves as insurance against posting losses during market downturns (both industry and broad economy declines). That, for me was an “a ha” moment when first reading it, as was your indication that trends in CROIC are more predictive of market results (makes sense, I just hadn’t thought of it).

    If memory serves, Ben Graham indicated that healthy companies should post ROIC results in the 6% to 8% range consistently (80% of the time, perhaps). Those producing greater results were, of course, more desirable.

    The problem with such stellar performance is that it lures competition, and competition promotes pricing pressure that drives returns toward WACC. This is why Buffett has written and spoken so often about moats — sustainable competitive advantages (i.e., franchise value). This appears in the 1983 shareholder’s letter, especially (see the addendum to that letter, as well). This is where he addresses the subject of amortizing goodwill and describes the differences between accounting goodwill and economic goodwill.

    In any event, I hope all is well with you, and congrats on producing an excellent blog.

    Robert

  • http://www.oldschoolvalue.com Jae Jun

    @ Sergei,
    Can you elaborate on what you mean by “growing market”? Do you mean the industry is expanding? An example would be like the e-book reader market that was created by the kindle?

    @ Dean,
    Yes thank you. You are right since net income already accounts for expenses whether or note they are cash items.

    @ Robert,
    I’ve gone over Greenwald’s book about 3-4 times and analyzed every useful page and made a spreadsheet out of it but the part that I just couldn’t come to grips with was the maintenance capex calculation. It is probably the best method that I’ve come across to date but I see some shortfalls that make me go hmmm. Not a big issue but always at the back of my mind.

    I don’t recall Graham talking much about ROIC. Was this from security analysis?

    I’ve applied WACC into a test spreadsheet version but it never worked out. I just try to keep it safe and set the discount rate to 15% most of the time. But what is your view on using WACC? Maybe I haven’t thought about other aspects of it except that variables such as beta immediately produce an incorrect value.
    My opinion is that it is just GIGO. Garbage in, garbage out.

  • Robert Crawford

    Thought I had responded to this earlier, but it appears my memory is flawed.

    The CapEx breakdown in Greenwald is an estimation, by his admission. CapEx, of course, includes actual outlays to sustain existing operations AND capital expenditures supporting new initiatives. Absent the ability to dissect the two, we are left with depreciation as the only means of differentiating the two. So, Greenwald’s estimated disaggregation seeks to determine whether the company has kept pace with sustainment investment — taking the view that, if maintenance CapEx (estimated) is consistently below depreciation, the company is under-investing in this vital area of operations. While imprecise, as an estimation, maintenance CapEx is not a steady expenditure requirement. The company may depreciate a corporate car yearly, but it does not replace it in yearly chunks, and the same holds for more expensive PPE outlays.

    As you note, this is not entirely satisfactory, but the flaw is with FASB, rather than Greenwald. While he does not say this explicitly, I gather this is a greater concern for firms posting larger PVs than EPVs — since these are firms that would most strongly benefit from growth CapEx expenditures.

    The ROIC standards I wrongly attributed to Graham were actually written by Martin Zweig in his commentary to the Intelligent Investor.

    With WACC, there are an assortment of different calculation approaches. I use the simplest of the described approaches (there are several) from Copeland, et al. “Valuation, 3rd,” pgs. 134-136. It segregates debt and equity financing. The opportunity cost (discount rate) is different for the two forms of financing in my model. For debt, I use the actual interest-paid yield (interest divided by total debt), since this is the risk adjusted premium demanded by the market/lenders for this specific firm. I do check to determine that short-term debt is appreciably less than short-term assets, and I do check on total debt levels in order to insure lenders haven’t gotten delusional. The tax benefits of debt follow, as does calculating the after tax cost, leading to weighted average cost of debt. For the equity side, I use the 30-year bond as the risk free cost of capital and add a market risk buffer of 8.6%. With RFCC at just over 4%, the total discount level falls marginally below the 15% you use.

    Calculating WACC is beneficial for a number of reasons. First, it identifies the long-term sustainable return on capital rate, since, theoretically, any return greater than the cost of capital will promote competition and a subsequent price war ending at a WACC equilibrium. Second, considering this over the span of a decade provides a strong feel for the preferred financial structure of the company. Fourth, the normalized rate can then be inserted into Greenwald’s growth multiplier calculations (M) as the cost of capital (R). Fifth, disaggregating the sources of capital financing allows the investor to make choices concerning the degree to which lowering the calculated rate for the equity contribution is possible, due to the one-time nature of equity financing — as long as the company requires minimal debt capital to sustain on-going operations and finance growth initiatives. And, lastly, comparing CROIC to WACC serves as a second check to Greenwald’s PV/EPV approach to determining when growth CapEx spending is justified and benefits the stockholder.

    I should mention that the approach I use to disaggregate share prices into shareholder’s equity and the investor-paid premium (with appropriate bond yields used to adjust owner’s earnings yields) allows me to consider discounting the equity portion of WACC (in my twisted mind, at least). With the equity portion of WACC, this represents a one-time source of capital financing for the company, which is unlikely to support short-term operating-capital needs. If the company has no history of secondary offerings, is not prone to dipping into treasury stock, requires little to no debt, and has a strong cash position, it seems to me that according 10% (or more) to equity secured more than a decade earlier serves to inflate the actual required cost of capital – which is why WACC is calculated.

    This idea struck me as I was considering the popular assertion that stockholders are risk-takers that finance the entrepreneurial sector. While those directly purchasing IPO and secondary-offering shares serve this purpose, tertiary purchases through the exchanges represent hand-off of proportional ownership between investors, but the company receives no fiscal benefit — beyond recognition that IPOs and secondary offerings would be undesirable if shares were subsequently traded in a suspect or illiquid market. As value investors, if we truly believe Buffett’s assertion that he does not care if the stock market closes for several years, share-sale liquidity becomes all but a non-consideration, and the only portion of WACC that reduces returns once the stock is purchased is recurring, non-equity debt.

    Of course, there is the “little” problem of preferred shares to consider with this view — with some firms (many) it is more than a little problem. So, ideally, any discounting of the equity portion of WACC would factor this, but, under FASB, interest on preferred is included under interest paid as a general category, within the debt side of WACC – serving to overstate the tax benefit of creditor debt if not adjusted. A further adjustment for diluted and non-diluted shares would be necessary, if seeking an accurate accounting. I’ve been working on this issue of breaking down returns (to common, creditors, and preferred), and, while it is included in one B-School text, it is given cursory and unclear treatment. It certainly makes this level of analysis all but inaccessible to the average investor. If you have broken he code on this, I’d love to get your ideas.

  • http://www.oldschoolvalue.com Jae Jun

    What is your background btw? I believe I read somewhere that you teach.
    I have to admit that I don’t understand the full theory of WACC and the components so I can’t offer any of my thoughts…. lol

    But, as a value investor, I am taught that since beta is useless, WACC is unnecessary as well. Is this just a myth or is there more to this?

    If WACC increases, so does beta and ROE, but I don’t see how beta has any place in valuation to begin with?

  • http://valueinvestortoday.com Jim

    I believe BETA is useless but the WACC is a viable tool. Using a BETA of 1.00 in the calculation seems correct to me. According to Bruce Greenwald, however, its perfectly fine to use a discount rate that represents what you expect the investment candidate to produce for you such as 15%, however – I believe you should also calculate the WACC because in some cases it costs the company more than 15% of their capital to operate and in that case a standard 15% discount rate would not be sufficient. In any event, finding a company who’s costs exceed 15% of capital is probably not a business you want to invest in anyway.

  • http://valueinvestortoday.com Jim

    Robert,

    I think you’ve presented a realistic WACC the way I interrupt its use also. The only difference is your market risk buffer. I use the 7.5% the S&P has historically averaged. Since I haven’t checked that number since the recession, it very well could be 8.6% now. But essentially, I use the current t-bond rate plus a 3.5% buffer bringing the cost of equity to 7.5%. The cost of debt calculation I use is the interest expense / total interest bearing debt which is essentially the same as what you’re saying or what I believe you tried to say. If there is $100 Million it total debt but only $50 Million that is collecting interest, you wouldn’t account for interest of the total $100 Million; just the portion that is exposed to interest. Also, the 7.5% cost of equity would only affect the equity portion of the business and I wouldn’t discount the entire market capitalization figure I come to by 7.5%; only the equity portion would be discounted. Therefore separating the capital structure of the business in two. In any event, doing it this way, for me, produces a very realistic discount rate to use in my PV calculation without the need of a BETA.

    Doing it long hand for Company XYZ:

    Total Assets: $190 M
    Interest Bearing Debt: $101 M
    Interest Expense: $8 M
    Risk Free Rate: 4%
    Risk Premium: 3.5%

    In this example, $7.92 Million of Interest Expenses would be taken out of the value of the firm (whatever value you’ve arrived at) regarding the cost of its debt and $14.25 Million would be taken out of the firm regarding the cost of its assets. Therefore, it will cost this company $22.17 Million in order to just maintain their current capital structure. So, for example, if this company produced $35 M in Owner Earnings and you’ve placed a fair value of this business at a 7.41 P/E, you’ve valued the business at $259.26 Million. You’d account for the costs of that business by deducting the above mentioned amounts of $22.17 Million arriving at $237.09 M, then you’d add back excess cash, say, $2.5 and deduct 1% of revenue in order to continue operations (700K) arriving at $238.89 M. Last, remove all interest bearing debt of $101 which would arrive at a $137.89 Million business.

    This is my understand and use of the discount rate. Although, I do believe using a 15% rate is of logical means but it would be important to know the actual costs of doing business for that company because if it would end up costing a company more than 15% for its capital structure, using a one size fits all 15% will end up overvaluing that business. With that said, I believe if you find that the capital structure of a business costs more than 15% – it is probably a very risky proposition especially if that business is highly leveraged.

  • http://valueinvestortoday.com Jim

    In continuation to what I previously wrote:

    I’m currently testing the relationship between SG&A and/or COGS to Total Assets as being a cost of equity. For obvious reasons this may work and it may not. Drawbacks are that some businesses spend the same in SG&A regardless of of how much or how little they bring in each quarter. However, COGS often change as a reflection of this. Therefore, for not, its just an untested idea for me but it could possibly work in the same way the cost of debt calculation is conceived by taking interest expenses divided by interest bearing debt to find the cost of debt. In any event, I’ve read through his annual shareholder letter than Warren Buffett bases his cost of equity from the T-Bond interest rate and not added to it a risk premium. I think he’s in a position to be less conservative by doing so. Depending on the business, my cost of equity factor is done that way also and then adding the actual interest expense of the business. As we all know, all of this is to get an idea rather than to be precise. As Graham said, you often don’t need a weigh scale in order to come to the conclusion that someone is obese or underweight.
    .-= Jim´s last blog ..Asta Funding – Postion Exited & Analysis =-.

  • Harley

    What is the company is making lost and tax rate is -ve? Would the calculation for NOPAT & ROIC make no sense? TQ

  • http://www.oldschoolvalue.com Jae Jun

    If the company received a tax benefit, you would just use a tax % of zero. No such thing as a negative tax rate.

  • Fab

    Hi Jae,

    in Europe, ROIC is:

    NOPAT/(Operating Working Capital + Fixed Assets)

    What do you think about? It sounds better?

    I’ve always used this version and I don’t understand a piece of your version.

    What do you mean for:

    Excess Cash = Total Cash – MAX(0,Current Liabilities-Current Assets)

    The first item is ok but I can’t understand the second.

    Could you explain better?

    Motley Fool defines “Excess Cash” as a percentage of sales (10%-20%).

    Let me know if you feel like!!

    All the best.

    Fab. greetings from Italy.

  • tuckandturn

    Jae,

    With regards to the formula: Excess Cash = Total Cash – MAX(0,Current Liabilities-Current Assets)

    For “Total Cash” are you simply using “Cash and Cash Equivalents” from the Balance Sheet or are you including “Short Term Investments” as well? Or is it a reference to some other combination?

    Many thanks!
    Doug

  • http://www.oldschoolvalue.com Jae Jun

    Total cash should include everything in the form of cash. This includes short term investments.

  • Tarun

    Hi Jae,
    It may be a silly question, but I was trying to use numbers from balance sheet for Operating Income and such. Should I use number from most recent quarter or use the yearly figure.
    Which is a better approximation and Why?
    Also we are in 2012 and Google Finance still does not show “12 months ending 2011-12-31″, do you know why?

    You work on Finance is highly appreciated.
    Regards,
    Tarun

  • Richard Gordon

    Hi Jae,

    I’m in awe of your writing abilities. I have come accross very few writers who can communicate as effectively as you can. Well done mate!

    Richard Gordon

  • Richard Gordon

    Fab,

    I think the second term “MAX(0,Current Liabilities-Current Assets” in the context of “Excess Cash = Total Cash – MAX(0,Current Liabilities-Current Assets”

    means: Total Cash – What ever is the greater zero or Current liablities – Current Assets.

    In other words, Current Liablities -current assets cannot be a negative number.

  • http://www.oldschoolvalue.com Jae Jun

    @ Tarun,
    Use the TTM or the previous year numbers if possible.
    G Finance may not show 2011-12-31 because companies do not always have their fiscal year set up like a calendar year. To some, the end of the company year could be in June.

    @ Richard,
    Thanks!

x
Receive Email Tips and Strategies on Achieving Higher Stock Returns

Bonus: FREE Set of 9 Investing Spreadsheets