It's not true that Wall St. is totally oblivious to CROIC. In fact some of the biggest brokers' use CROIC (or its variant) in prop strategies. cash measures, however, tend to penalise growing companies and favor stable/mature cos.

( REPLY | PERMALINK )

When measuring how "good" a CROIC is, shouldn't we be comparing it to how much that capital cost and not absolutes?
13% isn't very good if it cost the company 15%.

( REPLY | PERMALINK )

Hi Joe,
You mentioned that:
CROIC = Free Cash Flow divided by Invested Capital. Invested Capital is a combination of the company's net worth and any long-term debt it uses.

Why just long-term debt? why not also include short-term debt? So in your formula for CROIC why subtract Current Liabilities? I'm not understanding that part of it?

( REPLY | PERMALINK )

Current Liabilities are due within twelve months. A company can't really borrow money for, say, six months and use it to generate more cash. Instead, it will borrow for a few years and use those borrowings to acquire assets to generate more cash.

Invested Capital is the money that we have put into the business, the cash that the company has retained, and any long-term borrowings (or deferred liabilities) that it is not paying off because it is trying to generate more cash.

Hope that helps.

( REPLY | PERMALINK )

Graham, in "The Intelligent Investor," advocates ROIC as a valuable measure (along with a PE below 16 and PB below 1.62, among other measures) -- demanding greater than 6 percent ROIC for stable firms and greater than 10 percent for the rest. Other value investors advocate measuring the spread between ROIC and the Weighted Average Cost of Capital (WACC), which takes into account the cost of financing (both, debt and equity). Does CROIC serve as a viable substitute for this (perhaps, an improvement, given its comparative ease of calculation)? What are your thoughts?

Thanks,

Robert

( REPLY | PERMALINK )

There are various ways to value stocks and businesses. So long as your data and reasoning are right, you can approach it in multiple ways. If you know a company will grow earnings at 10% for ten years and will sell at 20 times earnings in 2017, you can figure out a purchase price today and make money. There's nothing wrong with that.

I value businesses as if I were acquiring the company in a private purchase. As such, P/E ratios and GAAP earnings (and numbers that rely on them, like ROIC) are not as important to me. I find that CROIC is a better measure for my valuation method.

One thing to remember about Graham is that, near the end of his life and after teaching (and then learning with and from) Buffett, long after writing The Intelligent Investor, Graham stated:

I am no longer an advocate of elaborate techniques of security analysis in order to find superior value opportunities.

Keep it simple and wait for home runs (unless you are with the Chicago Cubs - then just hit a pitch for crying out loud!)

( REPLY | PERMALINK )

Sorry, but your CROIC is flawed due to two fundamental reasons:
a) you are using accounting numbers
b) you are ignoring off-balance funding sources.

With respect to (a), that is self-explanatory since accounting numbers can be manipulated -- keeping in mind it is management that discerns the assumptions used.

With respect to (b), it can be shown that if you have two companies that are identical in every way except for their funding, the CROIC in theory should be the same but it isn't since the first company uses all debt financing while the second uses operating leases [note, there is a third way of using out-sourcing*] and the CROIC in the second company will dwarf the CROIC in the first company simply due to the fact the second company has a smaller denominator, respectively.

Cheers.

PS There are three ways a company can control assets: the company can own them outright sometimes using debt financing if the assets are expensive to acquire, the company can lease the assets using operating leases, or the company can sub-contract the work out to another company that already has the assets. In the long run, the company should be indifferent in which financing method it uses, i.e all three cases should be equally expensive to use. Your CROIC fails to capture this due to its bias to the balance sheet side of the equation, i.e. the denominator.

PPS You need to Damodaran's book that explains: equity cash flow to cost of equity, total firm cash flow to the cost of capital, etc.

( REPLY | PERMALINK )

FCF is much harder to manipulate than GAAP earnings.
Joe's valuation uses CROIC in conjunction with FCF growth for valuation.
You only have to be approximately right.

( REPLY | PERMALINK )

Exactly Peer : )

( REPLY | PERMALINK )

Also, leasing or contracting out the asset instead of debt financing reduces the numerator in CROIC thereby decreasing the relative return and balancing out the different methods over a 10 year time frame.

( REPLY | PERMALINK )

Exactly Peer : )

( REPLY | PERMALINK )

Chungst:

With respect to your comments:

1) Manipulation of free cash flow is very hard, albiet possible, but not nearly as much as the manipulation of the accrual basis income statements. Income Statements do take into account the "management estimates" you cite. Your cash balance either goes up or goes down, there is not management estimate to that (or as much of one should I say, we could into greater detail, but not needed here).

2) Off-balance sheet funding sources: true operating leases, especially with respect to large capital items used by a business, are not too common. I am not talking about copiers here. A manufacturing plant may have a huge lease on a piece of equipment. But based on accounting regulations, they most likely have to report this under a "capital lease" arrangement, which in a nut shell is like having a loan to buy the equipment. As for "out-sourcing", I am not so sure that is a usefull parallel either. I have not seen a company outsource a large production facility, and if they did, the cash outlay for such would be immediatly effective against the CROIC calc, the numerator mostly. Thereby driving down the ratio.

That is my 2cents anyway.

( REPLY | PERMALINK )

Here are my general responses:

1) When calculating free cash flow, how did people adjust for the fact the company may over-spend for Capex or under-spend for Capex, i.e. how do you know the correct level of Capex spend? The same is true for working Capex as well. I took a course with Professor J. Livnat because he co-wrote a book called "Cash Flow and Security Analysis" and got a better grasp of free cash flow in theory and in practice.

The same logic applies to earnings -- how do you know if the company underspent on marketing (or any other) expense to inflate earnings or worse yet under-invested R&D assuming these line items were broken out or disclosed? The reason I bring this up is GAAP earnings drive FCF and if GAAP earnings is suspect, then wouldn't wouldn't FCF be suspect as well? I have to thank Professor H. Fried (co-author of "The Analysis and Use of Financial Statements) for drive this idea into my head -- an analyst has to look at both the accounting numbers and the firm's economics and then compare the accounting against the economics. If earnings are suspect then (free) cash flow is suspect.

Also, as a cash-based analyst I had to normalize financial statements because the accounting numbers did not reflect economic reality. People need to understand that FCF is easy to understand in theory but difficult to execute in practice. For example if you have two identical companies but use different accounting assumptions and treatments allowed by US GAAP (i.e. the first company is conservative while the second company is aggressive) then the accounting numbers will be different but the ECONOMICS of both firms are the same. Under financial theory, investors could look through the accounting gimmicks. Even Warren Buffet comments about the need to use "look-through earnings" (source: Berkshire Hawathay's Owner's Manual). Seriously, how many people recast a firm's financial results and then all the firms' numbers in the same industry to get a true and fair view of what's occuring as part of the investment process?

In summary, while cash is hard to manipulate, people need to know that if a company didn't spend their required amounts on key line items (thereby saving on cash), the company's economics have been harmed and this won't be reflected in FCF.

Second: Operating leases, in the old days, was covered by FASB #13. Capital leases known in the leasing industry are called dirty leases because they are just loans documented under the guise of a lease. I have personally worked on operating leases in the $20 million range (i.e. co-generation plant) early in my career and have seen operating leases in the hundreds of million range. For example, FSC (aka Foreign Sales Corp) leasing of a Boeing 747 could ranging in the $140 million range in the early 1990's -- this isn't copiers folks! I never worked in real estate but based on reading financial footnotes, it's not uncommon to see synthetic leases in the hundreds of millions as well.

Today, as I look on the balance sheet and at the financial footnotes, I see operating leases and other off-balance sheet liabilities becoming more sophisticated and harder to unwind or understand. Then you have derivatives (i.e. FASB 133) which changes the economics of the original transaction, i.e. in the old days you looked at the coupon or interest rate to discern interest expense with great certainity but you couldn't do that anymore once the companies started to use derivatives.

Third. When you look at Flextronics or Rambus, these companies don't own billion dollar fab plants. Ford, for example, doesn't need to own it's assembly plants just like the numerous companys that sent the jobs abroad seeking cheaper labor costs. However the point is you need to have an apples to apples comparison so you need to recast a company that uses sub-contracting to the same as if the company owned plants or recast a company that owned plants into one that now uses subcontracting.

Lastly, with respect to the CROIC ratio, given the same dollar impact for either the numerator or the denominator, other things being equal, hits to the denominator have a greater impact than hits to the numerator.

Cheers.

( REPLY | PERMALINK )

In response:

I think you are trying to make things way too complicated.

In addtion, many of the "adjustments" you are referring too, if management continues to manipulate for the short term earnings benefit, the long term will suffer.

This is why ten years at a minimum, is a requirement.

I agree in general with your post, there are tons of shenanigans out there. But over the long term, FCF "normalizes" and you have evidence to base your analysis on. But as someone who has been auditing and putting together those financial statement for years, I have seen quite a bit of accounting "stuff".

Nice FASB reference by the way.

Comparision of two companies while it may be comforting, to me, is not relevant. I am looking at one company, is it good or bad? I am not looking to see if it is good or bad relative to others, that does me nothing.

Also, again with respect to assumptions made by management and GAAP. Those assumptions are reflected on the balance sheet and income statement, they can not be reflected on the statement of cash flows. All of your AR, Allowances, AP etc etc etc, the adjustments for those, year over year, are taken care of on the cash flow statement. Which is why Mr. Buffett focuses on "owner earnings".

Look through earning are relevant yes, more companies have off-balance sheet issues. I think Mr. Buffett is comenting with respect to GAAP requiring more inclusive reporting. Which I agree with.

But I think as Joe has posted on here, in a case like ENRON, you would have noticed FCF diving prior to the BIG EVENT.

I respect your very intelligent post.

But, again, I feel like you are trying to make it way too complicated. Like Mr. Pabrai has said, if he has to break out an excel spreadsheet it is a pass. And he is moving on to the more simple to understand business.

Would you invest in Flextronics or Rambus?

Cheers back!

( REPLY | PERMALINK )

I think it depends on who you are doing the valuation for.
Trying to nail down the nth digit after the decimal doesn't go far for the investor. There are plenty of other qualitative factors that are more important to the investors than quantitative factors.
May be it makes sense if you are the investment banker representing the buyer or seller in a sale or for IPO ,for tax purposes or divorce court
to use all the numbers available possible.
For investing it needs to be simple.
my 2 cents.

( REPLY | PERMALINK )

Joe,

"WHAT ABOUT BERKSHIRE'S CROIC?"

I have some concerns regarding your example. It appears to me you are comparing apples/oranges instead of apples/apples and oranges/oranges. If I can invest $10 in an investment and earn $5 (a 50% return) then Berkshire should be able to invest $1000 in the same investment and earn the same return (50%) which equates to $500. Berkshire can also invest $1,000,000 in the same investment and earn $500,000, etc. etc. A rather tidy investment for both of us to say the least.Conversely, if I invest $1000 and only get back $5 (CROIC = 0.5%) then a $1,000,000 Berkshire investment will only return $5000. Not so good for both of us.

Perhaps you do have a high level of confidence in the Berkshire management but if we are going to use CROIC as a comparative tool, then what is good for the goose must also be good for the gander. In that regard, with specific focus on CROIC, Berkshire would not get any gold stars for a 0.4% CROIC. As you have pointed out in the Berkshire post, we can not isolate our focus to only CROIC. Berkshire has great FCF growth, agressive asset growth and the management has a long history of success with respect to financial decisions and generating shareholder value. In my opinion, the same "bigger picture" needs to be taken into account when evaluation all of the non-Berkshire companies as well. In other words, after I have taken a look at the financial numbers (data) I need to sit back and ask myself "now what else do I need to consider that may not be reflected in the data". For example, Warren may be investing in BNI because he believes the future demand for food by highly populated countries (China, etc) suggest we will need to move an increasing amount of produce from central USA to export ports by rail. With the ever increasing price of fuels it makes rail more economical than truck and hence higher future demand for rail transport. Hopefully the higher future demand for rail will increase earnings, FCF, dividends and ultimately shareholder value.

Do I base my future cash flow model on FCF, CROIC, an average of the two, 75% CROIC which is deflated over time. Perhaps the answer is to just choose one and run with it. As long as we are somewhat consistent and use a similar model for multiple companies rather than changing the model for each company we look at we will be able to compare the performance of various companies and make our purchase decisions. For you info, I personally consider the Excel spreadsheet to represent the "similar model" and do adjust the FCF growth rate over time and also on a per company basis depending on the initial growth rate and how stable/consistent the financial history is. If I am starting at 30% (which is pretty agressive in the financial world) I may leave it for only year 1 or 2 before starting to decline. If I am starting out at 18% and the company is very stable then I may leave it for the first 5 years before starting to decline. Just an "educated guess" at best.

I would still like to get your thoughts on the Wikiwealth value model. I have a greater level of confidence in our Excel model if I can get some supporting agreement from Wikiwealth, iStock and GuruFocus.

Glenn

( REPLY | PERMALINK )

"The reason I bring this up is GAAP earnings drive FCF and if GAAP earnings is suspect, then wouldn't wouldn't FCF be suspect as well? I have to thank Professor H. Fried..."

Tell Professor Fried he is wrong. GAAP earnings do not drive FCF; FCF drives GAAP earnings. If there is no excess cash, GAAP earnings will suffer in a few months or years. If FCF is high, GAAP earnings will grow or remain fairly constant in the future.

Most people screw up when reading about Buffett and earnings. When he mentions "earnings", he is usually talking about FCF and owner earnings. More often than not, when he wants to refer to GAAP earnings, he qualifies the term by mentioning "GAAP".

There is nothing wrong with looking at the accounting numbers, but you should know that the accounting numbers ALWAYS follow the cash over the long term.

( REPLY | PERMALINK )

Jason Z.

Very well said.

( REPLY | PERMALINK )

The key is the cash that goes into the bank. Unlike the Medicare and Social Security Trust Funds, banks do not accept IOUs unless they are called loans.

Despite this, GAAP is sufficiently flexible to allow for overt and covert deception, even if it is the auditor who is most prominently deceived -- Enron. WB acknowledges this and, consequently, emphasizes the honesty and character of management as a key investment threshold. The problem is that Warren can get lunch with nearly any CEO in the world (he is heading over to China next week, I believe), while I have a hard time getting my teen-aged son to disengage with his girlfriend long enough for an in-house confab (Joe, your time will come).

Initially, I thought that reading the quarterly and yearly reports, scanning the news articles, and listening to the conference calls would provide a gut sense of leadership's character, but they all seem like such nice people for the hour they are on the phone each quarter. Even Ken Lay seemed most pleasant, and his wife came across as a more savvy Laura Bush -- must be a Texas thing.

So, if, regardless of the financials, an element of trust is necessary, how do you perform character due diligence on the management?

PS. Diversification of risk through a portfolio is another form of insurance against fraud and deception. Like doing sit-ups for years before taking a gut shot, it won't deflect the blow, but the pain shouldn't be quite as bad.

( REPLY | PERMALINK )

Jason Z. is wrong when he wrote: "Tell Professor Fried he is wrong. GAAP earnings do not drive FCF; FCF drives GAAP earnings."

I sat in Professor Fried's class and our class were used the drafts of 2nd edition of a book he co-authored, the same book that is the Bible for Financial Statement Analysis used for the CFA(r) program. Getting the CFA designation was difficult; it took me six attempts to pass all three exams!

Here's a very simple proof that Professor Fried is right on the money and Jason Z. needs to learn basic accounting. The proof is FASB #87 that GAAP earnings go drive FCF. Under FASB #87, it doesn't matter if a company chooses the DIRECT or INDIRECT approach to get to GAAP earnings, it is from this GAAP earnings that is the basis that arrives at the cash flow from operations (much easier to see in the INDIRECT approach since the accountant reconcile net income to non-cash items; under the DIRECT approach, you ignore non-cash items and thus the accountants get to the same cash flow from operations).

If you can't discern cash flow from operations (but you need GAAP earnings), you can't get to FCF -- unless of course, you also believe Jason Z.

As for Buffet, he looks at the economics and use a term call look-through earnings. Buffet, as I am aware, has never disclosed his definition of earnings.

Cheers.

PS For a a summary of FASB 87 go to this link: http://www.fasb.org/st/summary/stsum95.shtml

( REPLY | PERMALINK )

This will be my last post on this topic; please feel free to have the last word. I have enjoyed the discussion so far, and unfortunate I can no longer continue the conversation due to time constraints.

I will make two responses:

The first is a general response about the importance of looking at the economics by doing the exercise of normalizing for debt/equity financing, off-balance sheet financing (i.e. as much as $300 million for the new Airbus double decker) or subcontracting (and no I won't invest in Rambus or Flextronics unless these companies are undervalued along with a margin of safety). Here is a simple exercise, if you have access to bloomberg or any other financial database (i.e. Value Line), you can discern that the large publicly traded Home Builders are now trading at very low price to book ratio. Why is that you ask? Under my approach, the smart investors already discerned that these companies' book value are inflated and that they need to be written down. Why, because the economics do not support these book levels.

The second is specific response to Casey Mattson, who wrote: "But I think as Joe has posted on here, in a case like ENRON, you would have noticed FCF diving prior to the BIG EVENT."

Because of my training from Professor Fried, I now understand the importance of comparing the company's economics to the accounting numbers. In the following, I will give two real-life examples.

I will state that when I was a fixed income analyst at a pension firm (we had about $11 Billion in High Grade bonds at the time), I created the model that identified trouble credits at the time such as ENRON, Williams, El Paso, etc. In fact, I told the WorldCom analyst from May to July 2002 that WorldCom's cash flow numbers were suspect, it was in the billions, and that the analyst should look at Capex (note the fraud was disclose late July 2002).

What tipped me off about WorldCom was Jim Chanos shorting the stock because the Company had loan hundreds of millions of dollars to the CEO so the CEO could make his margin calls without selling WorldCom stock. At the time, the loans to the CEO was about 4% while WorldCom's yield to worst (a proxy for borrowing cost) was 15% meaning WorldCom was losing 11% (based on the economics) on these loans AND WorldCom was having a liquidity crisis at the time AND miraculously cut Capex by about $4B. Hmmm.

My story about WorldCom has a sad ending; I was written up for trying to save my firm hundreds of millions (I guesstimate the number at $300MM). Yes, I kept the documents. In my case, I was accurate about WorldCom's fraud as early as May 2002 so it wasn't a case of I told you so AFTER fact -- in my case, it was BEFORE the fact.

The case was on Ford Motor Company and I was one of the biggest bears on this company (I kept the transcripts from the S&P conference calls led by Scott Sprinzen) on Ford becoming a junk credit as early as December 2002. Back then, Ford did not disclose enough information for people to decipher Ford's true financial performance or financial condition. I got lucky when an UBS sell-side analyst (who had previously worked at Moodys) wrote a report about Ford Motor Credit Corp's (FMCC) leasing model but could not supply the correct numbers in the report. Since I had worked in leasing, I was able to work with USB and reconcile the numbers -- I then sent the numbers to Ford and Ford responded it was a fair representation of the leasing business since the numbers were derived from GAAP numbers that Ford had publicly disclosed. I discovered that FMCC was the largest buyer of Ford products and that was unsustainable; my investment thesis was that FMCC could no longer help support Ford's products. To me there's something wrong when the largest buyer is the company's own (indirect) subsidiary.

My point is that I was in the trenches when the stuff was happening LIVE. It's too convenient for people to talk about the subject after the fact, MANY years after the fact.

Cheers.

PS With respect to Buffet's look-through earnings, it came from the horse's mouth in a pamplet that Buffet himself wrote. I think it is safe to say Buffet knows what he is talking about and the rest of us are just commentators.

PPS If you are like Jason Z. who don't understand accounting, I strongly recommend a book called "How to read a Financial Report" by John Tracy; the book is excellent for understanding a concept called Financial Statement Logic.

Good luck.

( REPLY | PERMALINK )

Correction in my response to Jason Z. the correct FASB is #95 as noted by the link. Please replace #87 with #95; #87 was a typo.

Thank you.

( REPLY | PERMALINK )

Chungst,

I do understand a bit of accounting. Many of us non-CFA(r) business owners do. I don't really know the theory as well as you do, but I have some "in-the-trenches" experience with it.

Worldcom. I kind of saw trouble in 2000. I admit that I never looked at the leases. I remember that they reported $4.2 billion of GAAP earnings with some $3 billion of negative free cash flow. When they reported $1.5 billion of GAAP earnings in 2001, but only $100 million of free cash flow, I was supect again.

I can't speak to when you saw it in 2002 because I can't find the financials anymore, but fraud or no fraud, if you think like a business owner and buy the future cash, you would have quickly seen that Worldcom was overvalued long before your 2002 revelation.

It's all in the annual reports at the SEC (http://www.sec.gov/Archives/edgar/data/723527/000100547702001226/d02-36461.txt)

Ford is a cyclical, heavy-cap-ex business with thin profit margins. Hardly groundbreaking that it ran into problems. As a Buffett follower, you likely know that Ford is a terrible business and an even worse example. Regardless of net income, free cash flow, leases, or whatever, don't invest in bad businesses.

I'm not knocking you or your credentials. But it doesn't hurt to be a little more open minded than quoting FASB. After all, everything you are talking about relates to tax return numbers for the IRS. Talk to a business owner and you'll quickly find out that the IRS numbers and the survival and growth numbers are two totally different things.

Please don't call being a CFA bond analyst "in the trenches." It sounds as though you were as deep in the trenches as my old stockbroker. Build a business on GAAP earnings while disregarding free cash flow and then we'll talk. Though I suspect you'll be on the Enron side of the steel bars for that conversation.

Cheers.

PS: Buffett defined his "earnings" in the 1984 letter to shareholders. The fact that you think he should qualify his definition of earnings with the word "owner" so you don't get confused every time is insane. Why should he constantly repeat himself when there is so much free cash flow to buy?

PPS: I prefer to read Calvin and Hobbes. If I ever need a job at the IRS, I'll check out Tracy's work. Until then, I'll focus on how businesses grow, not how they are taxed.

If that really is your last post on the subject, I've enjoyed the banter. Thanks!

( REPLY | PERMALINK )

> But it doesn't hurt to be a little more open minded than quoting FASB. After all, everything you are talking about relates to tax return numbers for the IRS.

Jason Z. I strongly recommend you try to learn accounting because you clearly don't know what you are talking about.

FASB stands for Financial Accounting Standards Board and FASB #95 is short hand for FASB prouncement #95. If you had understood FASB #95, then you would not have made your comment that "GAAP earnings do not drive FCF; FCF drives GAAP earnings" in the first place because GAAP earnings drive Cash flow from Operations which then drives FCF.

Also, a company keeps more than one sets of books: (a) one set under US GAAP and (b) another set for taxes, etc. The fact we deal with the GAAP number doesn't mean the analyst has to forget about the tax books -- these two books are separate and distinct animals. For you to bring up the IRS angle means you are showing everyone that you are that clueless.

> they reported $1.5 billion of GAAP earnings in 2001, but only $100 million of free cash flow, I was supect again.

That doesn't provide anything especially is a company is building out or upgrading it's network or any company that is in the growth phase. People in the trenches like Jim Chanos can show the cause and effect. Chanos was shorting the stock before the fraud was announced and he disclosed his rationale.

> Ford is a cyclical, heavy-cap-ex business with thin profit margins. Hardly groundbreaking that it ran into problems.

Again, this is going over Jason Z's head -- Ford's indirect subsidiary was the largest buyer of Ford products and artificially supporting Ford as long as it could and this is INDEPENDENT of Ford being a heavy-capex-business with thin profit margins. What I stated was Ford was a shell game and I identifed how the game was unsustainable -- direct cause and effect by looking at the economics.

> should he constantly repeat himself when there is so much free cash flow to buy?

Because if Jason Z. understood accounting, and by looking at FASB #95, Jason Z. would see that not all the free cash flow goes to the (equity) owners if a company is levered.

( REPLY | PERMALINK )

I think there is some confusion here. It seems like Jason Z. is saying that today's free cash flow drives tomorrow's net income and Chugst is trying to explain that today's net income is the basis for today's net income.

I can't speak to the Ford situation. I agree with Jason Z. that Ford's capital intensive operations will eternally leave it in a wildly ciclical cycle - and it is a business not worthy of any further investigation.

( REPLY | PERMALINK )

I like Cash Return on Tangible Assets even better than CROIC, which is the same as CROIC but backing intangibles out of the denominator. If a company has a high cash return on tangible assets, it probably is a dominant player with a "wide moat" and can generate a ton of cash without significant tangible capital employed due (possibly) to a competitive advantage and/or strong brand name.

Thoughts?

( REPLY | PERMALINK )

BPal, I agree with your formula. It could be possible.

( REPLY | PERMALINK )

I think the cash return / funding sources is a great guide as to whether the company is giving a cash return above the cost of its funding (shareholder equity debt).

However, I have a number of questions -

(1) Because assets are not looked at, only funding sources are, and I understand depreication is added back to work out cash-return, can you confirm this WILL NOT catch inflated assets being written off against profits through large depreciations?

(eg; a series of overpayments for writing down overvalued assets usually seen in an overlarge intangible goodwill value) which are written off against profits continually though depreciation. Known as 'watered stock' in the 1950s I believe.

(2) The second main thing for me - in following the spreadsheet was trying to get my head around the multi-time frame analysis ; then a median taken to get the goal of the average/median growth rate of Equity and FreeCashflow.

(a) I believe this multiple timeframe approach maybe a big mistake and should not give more information, but give you less to project future values with.
It may lead to more errors, pretty significant errors, in sampling due to the selection of the timeframes and then taking a median of the values, not the mean or average.

I will explain my reasoning;

Medians are generally used where there is a set of items sold for different prices, and are known not to represent the entire spectrum well, as they merely take the middle-most item. This is used with multiple timeframes to represent the median growth over those timeframes.

The first thing is why multiple timeframes?

My understanding is that a multiple timeframe approach is valid where the sampling period of the data is different, for example 1 mins and 15 mins data periods will give markedly different information if the overall data period is different, say 1 days worth of 15min data, vs 1 hours worth of 1 mins data.

But if the 1 mins data is so much as to encompasses the full day, all the 15 min 'timeframe' data is redundant. You will have no more information to be gained by choosing the 15 min 'timeframe' data period over the 1 mins data periods, and will in fact loose information.

The real underlying information
---------------------------------------

As we have only the 1 year period data, spanning 10 years, that is the maximal information we have to work with. You cannot get more information out of the data, for the average rate of growth over the entire period, than the average rate of growth over the entire period without distorting the underlying trend I believe.

Selections of timeframes on top of this data will simply skew the data one way or another by the varying lengths and selections. In your spreadsheet a selection of 2 years spaced apart are taken from the 10 year selection, and various combinations of this are given 10 times.

This cannot be representative - picking 2 year period combinations, from 10 years there are 45, possible 2 year combinations that can be picked from a set of 10 numbers! (r Combinations from N = N!/(R!*(N-R)!) = 10!/(2!*(10-2)!) = 45)

You thus have only a part set of multi year data combinations.

Also, I believe you run into problems taking a median of the data rather than the mean. As the median is less representative than the average (being the middle value of a set of values low to high in such a set.

I think its important to use the best representation of the growth rate, accounting for the troughs as well as the bumps in the trend as you project the next 10 years from this trend. So

Another way of looking at it ;We want a general straightline trend through the bumpy free cash flow or equity data. A straightline is represented in plain math by y=mx c. c is the intercept of y (this is the start value of freecashflow).

The rest of the straight trendline is given by the gradient (average) that each x is multipled by to rise (or falls) to give y at that point. This multiplcation factor would have to be the average of the 10 year sample of data to fit in between all the data points and make sure the straightline fairly fits between them!

Obvoiusly this affects the calculation of the average rate of growth. But I think its important to get this as correctly worked out as possible if you are projecting the future from the past.

( REPLY | PERMALINK )

Well, just to update I was wrong on point 2 - well, mostly wrong anyway.
The very best thing is to fit a trendline to the trend of the growth of equity, FCF etc..
http://www.anderson.ucla....
This will smooth out the trend properly and give the underlying rate - and its not a straight line, but a log fit.

After reading up a hell of a lot on this - It is valid to fit 'timeframes' to the data over a selection of multiple years working out the CAGR between the years and averaging that. (But only as long as the time distance between the timeframe is the same or you do distort the data).

( REPLY | PERMALINK )

I just read the excellent report titled Free Cash Flow vs. Owner Earnings (http://www.fwallstreet.co...) and am wondering why CROIC uses Free Cash Flow in the numerator rather than using Owner Earnings. Any thoughts?

Thanks,

Rob

( REPLY | PERMALINK )

Okay, ROE can be wrong b/c debt can falsely balloon ROE. ROA can be wrong b/c any particular asset can be grossly under- or over-stated.

I like CROIC, b/c using the debt balance makes sense, but wouldn't the equity number in CROIC be off a little since it has Retained Earnings in it, a GAAP number? What if we added the DEP and AMORT back into Retained Earnings? Then, your CROIC formula would look like this:

CROIC= FCF/(Non-current liabilities SH Equity D&A).

It's a minor adjustment and it may not be worth the extra work, but wouldn't it be a little closer to reality? Thanks for any answers.

( REPLY | PERMALINK )

The formula didn't come out right above... it should be FCF/(Non-current liab plus SH Equity plus D&A)

( REPLY | PERMALINK )

BPAL, your definitely on to something. Its useful indeed

( REPLY | PERMALINK )

I think Joel Greenblatt uses the Tangible Assets in ROC. But, can't even tangible assets be over- or under-stated, especially if they've been carried on the books a while? How is this more accurate than DEBT plus EQUITY plus D&A?

( REPLY | PERMALINK )

Eliot: Sure that is a risk, especially if you try to apply that method to finance or insurance companies, which have many assets marked to market and therefore susceptible to big movements (similar to what's happening now). But otherwise, I think the risk of tangible assets be significantly over/under stated is low.

Accounting rules are pretty explicit on the need to write-down assets deemed impaired, helping reduce the risk of overvalued property. And since assets are added at cost, in an inflationary economy the risk of understated values is small. Most everything else is "current" (i.e used within one year) and so approximates fair value. Just keep an eye on other non current assets such as deferred taxes.

( REPLY | PERMALINK )

Nice, thanks for that explanation BPal! God I love these discussions lol =)

( REPLY | PERMALINK )

BPal, just to clarify, are you saying that the formula would look like this: Cash Return on Tangible Capital= Total Assets minus any Goodwill and Intangibles?

In other words, you are including all Current Assets and PPE? Thanks!!

( REPLY | PERMALINK )

Of course the numerator would be FCF. Sorry!

( REPLY | PERMALINK )

Any comments would be great! Thanks!

( REPLY | PERMALINK )

Eliot: Yes. Except that in addition to current assets and PPE, I would include other non-current assets as well, such as security deposits, long-term prepaid expenses, deferred tax assets (in some circumstances), etc.

( REPLY | PERMALINK )

BPal, thanks. Except I would have to disagree with putting anything current into the calculation. I tend to agree with Joe in that you can't expect to realistically earn a return off of anything current (after all, your current assets fund your day-to-day operations... they are not really "invested" per se, right?).

For example, Microsoft sat on an enormous pile of cash for years, but they didn't necessary earn a return off of that cash, it was the 'invested' capital that they built a moat and earned a return off of. The cash served merely as a reserve (if they called it that or not).

This is a fascinating discussion for me. I've been reading up a lot on this. The ROA, ROE, ROCE, ROIC, CROIC, etc are all great ratios, but each have weaknesses. Even after research, the CROIC is my favorite.

( REPLY | PERMALINK )

Current assets are considered even in CROIC. If the denominator in CROIC is Equity Liabilities - Current Liabilities, it's the same as saying Total Assets (which is Equity Liabilities) - Current Liabilities. I simply suggested taking the same denominator (Total Assets - Current Liabilities) and also backing out intangbiles and goodwill.

You should be including current assets (or more specifically net working capital) because that representative of how effectively management is employing its capital. If a company is sitting on a lot of cash earning 2% interest but could invest that cash and earn a higher rate of return, either CROIC or my hybrid version would "penalize" that company with a lower return on the basis that management is not effectively employing it's capital resources. Or if accounts receivable are high, one could argue that management is not effectively turning over it's receivables (i.e. quickly converting them to cash) thus resulting in a lower CROIC or return on tangible assets.

The key is understanding that the denominator in either version of the formula is TOTAL Assets - Current Liabilities.

( REPLY | PERMALINK )

You're absolutely right! Thanks. A lot to learn...

( REPLY | PERMALINK )

Question about Tonyhamm comment (Anyone welcome to answer)

He stated that ....

>(But only as long as the time distance between the timeframe is the same or you do distort the data).
>(a) I believe this multiple timeframe approach maybe a big mistake and should not give more information, but give you less to project future values with.
It may lead to more errors, pretty significant errors,

Could someone please explain the reasoning behind this statement?

In my mind (keep in mind I am still a student) having different time periods would help AVERAGE the growth rates more efficiently, giving you a broader look at how the company has grown FCF and CROIC over the years. Whereas if you were to stick with only one set of "timeframes" (say 5 years per time frame) it could possibly limit your view from a potentially broader one.

I just can't see why having multiple time frames could distort the data, while having a single time frame wouldn't. If someone could please explain this to me it would much appreciated.

Thanks to all for a fantastic discussion thus far !!

( REPLY | PERMALINK )

Joe, I'm new to your site and so far have really enjoyed what I've read. This is my first posting to you and I have a big question that has stumped me lately and maybe you could help shed some light on. My understanding of FCF is that it is defined as Net Income Amortization/Depreciation - Changes in Working Capital - Capital Expenditures. Invested Capital is Shareholder Equity Total Liabilities - Current Liabilities. CROIC is FCF / Invested Capital. I've seen some definitions of FCF as being Net Income Amortization/Depreciation Deferred Taxes Non-Cash Items - Changes in Working Capital - Capital Expenditures.

Which of the two is your definition of FCF?

My second question is this: Since value investors typically deduct intangibles & goodwill from total assets when valuing a business, wouldn't it make sense to also deduct them when coming up with the Invested Capital figure which would also, in many cases significantly change CROIC?

Thanks. Hope everything is going well for you and look forward to hearing from you.

( REPLY | PERMALINK )

Free Cash Flow = Net Cash From Operations - Capital Expenditures

Free Cash Flow can sometimes be used as a proxy for owner earnings -- a more precise figure for the cash a business can generate.

Your best bet (in my opinion) is to use Buffett's definition of owner earnings.

The decision to deduct intangibles/goodwill from CROIC would depend on how realistic those intangibles were. If Company A acquired Company B for $500 million more than the net assets of Company B, Company A would end up with $500 million in goodwill. To help determine if it was an intelligent acquisition, we would want to know what sort of cash flow it was generating on the acquisition.

At the end of the day, the $500 million of goodwill was shareholder money that was reinvested in the business, and should be considered as "Invested Capital" of the business.

( REPLY | PERMALINK )

Thanks Joe.

( REPLY | PERMALINK )

JOE SAID

"Current Liabilities are due within twelve months. A company can't really borrow money for, say, six months and use it to generate more cash. Instead, it will borrow for a few years and use those borrowings to acquire assets to generate more cash."

This year my stellar company has restructured it's short term borrowings into long term debt. This has had the effect of reducing its FCF to negative (it's worst FCF result in ten years). When I look at the balance sheet everything else seems to be in order and the business has returned to positive owner earnings for the first and second q.

Should I be concerned? Or is this a "one in ten bad year".

( REPLY | PERMALINK )

The question is: What effect will that debt have on its projected cash flow? If it was merely a shift from short- to long-term borrowing and the interest payments won't change much, it shouldn't have much of an effect.

The free cash flow change was due to the changes in the balance sheet (working capital) as short-term liabilities were reduced. Not a cause for concern if it was merely a refinancing of manageable debt.

( REPLY | PERMALINK )

Thank you Joe

( REPLY | PERMALINK )

Joe, is there anything wrong with using Working Capital plus Net Fixed Assets as the denominator in the CROIC equation?

Thanks

Great Site.

( REPLY | PERMALINK )

Makes perfect sense to me. The idea is to figure out what sort of capital outlay is needed to keep the business going, and looking directly at net fixed assets or making other adjustments is perfectly fine.

( REPLY | PERMALINK )

Thanks Joe, a question has been bugging me though.

What about intangibles like goodwill? If a brand is vital to maintaining business operations, shouldn't the price paid to acquire or to build that brand be factored into the required investment capital each year? I mean, after all, take certain brands away from company and you will effect its operations. And like you said we're trying to figure out how much capital any given company requires to maintain status quo.


Hoping you'll shed some more light on the subject for me and many others.

Thanks!

( REPLY | PERMALINK )

This is excellent! I am constantly looking for practical guides to investing jargon. You're quite right; I had never heard of CROIC (I'm a novice), but it seems like one of the better ways to calculate how well a company is being run. Of course, now I have to go back and read all your articles! Thanks!!

( REPLY | PERMALINK )

I have been looking at the CROIC of a few companies in Singapore and the CROIC seems to in some cases, have jumped from 12% to 30% over five years.

Have you looked at any singapore company in your analysis ?

Is it because the accounting standards are different ?

Regards

( REPLY | PERMALINK )

Hi everyone,

I’m still learning, so bear with me.

I’m having a couple of problems with the model and I would really like if someone could take 5 minutes and explain them to me. First of all - shareholder’s Equity.

From what I understand once a company issues shares for the first time, the amount of money they get (thus invested in the company) is the issuing price multiplied by the amount of shares issued. But the price changes over time (obviously), so when placing the current Shareholder’s Equity in the formula, am I not taking a number that doesn’t really represent the invested capital in the company?

I know that the SE figure is the relatively small figure in the denominator, but still.

Cheers everyone,

and happy Hanukkah-Christmas-New-year’s =)

( REPLY | PERMALINK )

Ziv,

Market price of the stock will not effect SE after the company initially goes public, unless they have follow-up equity offerings. SE is effected by earnings retained, dividends paid, treasury stock bought/sold, etc.

( REPLY | PERMALINK )

ROIC is not based on earnings. it's just EBIT * (1-t) / invested capital. The flaw with ROIC is that it used Book Value of assets in invested capital - which could be totally meaningless for say, a manufacturing company that doesn't mark to market its assets. Your CROIC does not fix this problem however.

With this definition of CROIC - the only difference vs ROIC is the numerator. The article's just going a step further with EBIT * (1-t) by taking out reinvestment needs and come up with FCF - which a good measure in its on rights . but still doesnt solve the problem that ROIC uses book value

( REPLY | PERMALINK )