PDA

View Full Version : EVA and Buffett/Munger


FatOtt
11-26-2005, 02:52 PM
This was brought up in another thread, but I thought it deserved its own discussion.

I want to talk about the Buffett/Munger viewpoint on EVA, but first Iíll talk about economic value in general. The creation of economic value (as opposed to accounting profits) occurs when the return on a firmís capital is greater than the cost of that capital. This is a pretty fundamental concept and there really shouldnít be that much argument about it. Itís simplest to think about for a bank from a purely debt-financed perspective:

- If a bank can borrow money (by encouraging people to deposit money in checking accounts, savings accounts, and CDs) at 3%, but lend that money (in the form of mortgages, car loans, and personal loans) at 7%, they will make an economic profit.

It gets a little more complicated when the firm is financed by other means, namely equity. Unlike debt financing, which generates an explicit Interest Expense on the Income Statement, equity financing does not generate any cost on the Income Statement. The effect of this is that the cost of equity is not incorporated into the firmís reported results on their Income Statement. But the use of capital has a cost, even if that cost isnít recognized on the Income Statement. Investors do not give the firm money (i.e., buy shares of stock) with the expectation of no return. Rather, they buy that stock with the expectation of making some return, even though they donít have the contractual right to that return (as debtholders do).

Because of this, itís important to differentiate accounting profit (as shown on the Income Statement) from economic profit. Accounting profit recognizes a cost for debt financing, but not for equity financing. Economic profit is the more appropriate measure, but how is it calculated? Well, what you need to do is look at the firmís total capital: all of the firmís assets, regardless of whether they were financed with debt or equity. (Remember that Assets = Liabilities + Equity, so you can think of Assets as definitely being financed by either debt or equity.)

You then need to recognize that there is a cost associated with that capital: debtholders require an explicit return, while equity holders require some positive return in expectation, even though that required return is not observable.

So at the end of the day, the economic profit for the firm is equal to the Net Operating Profit After Taxes (NOPAT) Ė (Capital * cost of capital). If the firm was financed 100% via debt, this would be the number on the Income Statement. However, if the firm is financed by equity (which all publicly traded firms are), the cost of equity needs to be accounted for.

Iíll stop here and say that Buffett and Munger will agree with everything to this point. In fact, I believe they would say this is obvious Ė a firm must earn more than its cost of capital (however financed) in order to earn an economic profit. This is evident in their speeches, in their actions (see the imposition of a hurdle rate on reinvestment for their subsidiaries), and their comments in Annual Meetings. Iíll also point out that Berkshire and Coke (as mentioned by Evan) are not nearly the only firms who think in this manner. For example, I was reading the Constellation Brands annual report yesterday, where the CEO spent a lot of time talking about this very topic. I think itís obvious that many (if not most) firm managers understand intuitively the need to earn more on assets than those assets cost.

The problem comes down to specifics. There are two components in EVA calculations that can cause controversy:

NOPAT: Net Operating Profits After Taxes is not an accounting measure (meaning that GAAP does not specify this calculation). Stern Stewart takes the GAAP Income Statement and makes a lot of adjustments to the Net Income number:

1. Back out Interest Expense. This isnít at all controversial, just a recognition that if youíre looking at the return on assets, you canít deduct interest expense. That interest expense is a return on assets that is being paid back to debtholders (not necessarily paid back in cash, just economically). The tax effect of this is removed as well.

2. Other items that Stern Stewart views as being incorrectly treated by GAAP. Now, I donít work for Stern Stewart, so some of these may be factually off. But hereís the type of thing they adjust for: GAAP requires advertising expenditures to be expensed on the Income Statement, reducing Net Income. SS would say, ďWait a minute, those advertising expenditures are creating value Ė thatís how really valuable brand names like Coke and Starbucks are created. Weíre going to treat those items as an asset rather than expense them. Same thing for Research & Development costs.Ē So there are some very specific adjustments made to the Income Statement. Some of these adjustments are probably very acceptable to Buffett and Munger, others might be considered twaddle.

Cost of Capital: Hereís the big issue. Buffett and Munger would both agree that there is some cost of equity. That is, they would never say that issuing equity has no cost or that there is no such concept as cost of equity. What they go ballistic about is attempts to actually calculate a firmís cost of equity. This is where measures like beta come in to play. Bear in mind, a cost of equity calculation does not require beta, itís just something thatís historically been used a lot. So a calculation of EVA might include the fact that the firmís cost of equity is 11%. Thatís more or less the hurdle rate on equity. If the firm had no interest-bearing debt, youíd say that EVA is equal to the firmís earnings minus 11% times book value. That 11% times book value is the required rate of return on the firmís capital base. If the firm isnít earning at least 11% on its equity, they are actually destroying value. The cost of equity equal to 11% is similar to saying that investorsí opportunity cost is 11% - if they donít invest in this particular firm, they could invest their money elsewhere and earn 11%. So the failure to earn at least 11% would be a destruction of value.

Overall, this is a major quibble that I have with Buffett and Munger. They take a concept like EVA that is at least 80% correct and argue with the remaining portion that they disagree with. Iím sure that they would say, ďWell, the part that is correct is obvious, so we wonít talk about it. Weíll just talk about the part thatís incorrect.Ē

The problem is with people (and Iíve met a few) who donít have the Buffett/Munger depth of understanding and who actually form opinions based on the Buffett/Munger soundbites. They hear the criticisms of EVA and infer that EVA as an entire concept must be fatally flawed and simply twaddle. They donít realize that the bulk of whatís included in EVA (economic profit vs. accounting profit) is a very valuable way of thinking.

Sorry for the long post.

Evan
11-26-2005, 04:54 PM
Good post. Just a couple things I'd like to point out.

[ QUOTE ]
Iíll also point out that Berkshire and Coke (as mentioned by Evan) are not nearly the only firms who think in this manner.

[/ QUOTE ]
Berkshire is not an EVA firm. I'm not totally sure if you meant to say they were, but I just wanted to clear it up.

Second, pretty much everyone who doesn't like the EVA method has the same reason; it maximizes firm value, not equity value. I've never heard either of the Berkshire managers talk about their specific reasons, but I would be shocked if this wasn't it.


Other than that I think I agree with everything you wrote. Hopefully someone here willl turn out to be an EVA supporter (from a management perspective). That should turn out to be an interesting debate.

Oh, one more thing. I just wanted to make it clear that the net cost of capital here is an after tax weighted average of debt and equity (only saying this because your notation was a little unclear imo). So it would be something like this:
Cost of capital (WACC) = Cost of equity * (equity/equity+debt) + After tax cost of debt * (debt/debt+equity)

After tax cost of debt = Cost of debt * (1 - effective tax rate)


I think Stern Stuart uses book weights for debt and equity, but someone please correct me if I'mm wrong. Most Wall St. firms use market weights; that is another source of debate with the EVA approach because it will yield a different optimal capital structure and a different cost of equity assuming you use a built up beta.

edtost
11-26-2005, 07:13 PM
[ QUOTE ]
Second, pretty much everyone who doesn't like the EVA method has the same reason; it maximizes firm value, not equity value. I've never heard either of the Berkshire managers talk about their specific reasons, but I would be shocked if this wasn't it.

[/ QUOTE ]

Not having heard of EVA before this thread, this immediately jumped out at me as a big problem. One scenario in particular where attempting to maximize the value of the firm as opposed to the value of equity is really really bad for stockholders is when the company is near default/bankruptcy (value of firm ~ debt). In this scenario, equity holders, who see nothing no matter how badly the firm does if value of firm < value of debt, would rather increase variance of the value of the company in the future, even if, in doing so, they lower the expected value of the firm. It would suprise me greatly if there were not smaller diffferences in optimal strategy for maximizing firm value/equity value for different debt/value ratios.

DesertCat
11-27-2005, 01:33 AM
[ QUOTE ]
This is where measures like beta come in to play. Bear in mind, a cost of equity calculation does not require beta, itís just something thatís historically been used a lot.

[/ QUOTE ]

Hmm, I'm pretty sure that their criticism is that the promoters of EVA use Beta (http://www.investopedia.com/articles/fundamental/03/031203.asp) to calculate cost of equity. It's "Beta" they've long derided as twaddle, as they regard it as meaningless and with zero correlation to capital costs or equity risks.

Of course they've been imposing hurdle rates on their subsidaries since the sixties. So they don't disagree with estimating a cost of capital, they just don't think there is a formula for it.

Evan
11-27-2005, 04:31 AM
[ QUOTE ]
[ QUOTE ]
This is where measures like beta come in to play. Bear in mind, a cost of equity calculation does not require beta, itís just something thatís historically been used a lot.

[/ QUOTE ]

Hmm, I'm pretty sure that their criticism is that the promoters of EVA use Beta (http://www.investopedia.com/articles/fundamental/03/031203.asp) to calculate cost of equity. It's "Beta" they've long derided as twaddle, as they regard it as meaningless and with zero correlation to capital costs or equity risks.

Of course they've been imposing hurdle rates on their subsidaries since the sixties. So they don't disagree with estimating a cost of capital, they just don't think there is a formula for it.

[/ QUOTE ]
I think you guys are both saying the same thing. The OP was talking about calculating the cost of equity in general (in the passage you quoted), not specifically the EVA method.

DesertCat
11-27-2005, 01:44 PM
[ QUOTE ]

I think you guys are both saying the same thing. The OP was talking about calculating the cost of equity in general (in the passage you quoted), not specifically the EVA method.

[/ QUOTE ]

Let me be more clear then. Stern Stewart has registered a trademark for EVA, so when you say "EVA" you aren't talking about the general concept, you are talking about Stern Stewarts EVA(R), whose underlying process is kept proprietary BTW. But if you read their materials, it appears clear that EVA(R) always uses Beta to estimate cost of equity. This is why EVA(R) is criticized by Munger.

Other than that, trying to "own" a concept so simple and obvious is the only other thing I'd criticize Stern Stewart for. Of course you never borrow money to invest at a lower return that you pay in interest. Just like you never invest any capital at lower returns than are available elsewhere.

DesertCat
11-27-2005, 01:50 PM
[ QUOTE ]


Overall, this is a major quibble that I have with Buffett and Munger. They take a concept like EVA that is at least 80% correct and argue with the remaining portion that they disagree with. Iím sure that they would say, ďWell, the part that is correct is obvious, so we wonít talk about it. Weíll just talk about the part thatís incorrect.Ē

The problem is with people (and Iíve met a few) who donít have the Buffett/Munger depth of understanding and who actually form opinions based on the Buffett/Munger soundbites. They hear the criticisms of EVA and infer that EVA as an entire concept must be fatally flawed and simply twaddle. They donít realize that the bulk of whatís included in EVA (economic profit vs. accounting profit) is a very valuable way of thinking.


[/ QUOTE ]

Let me add some color on why Munger and Buffett criticize EVA(R). Stern Stewart is essentially a consulting organisation that's trying to get lucrative contracts from companies where Munger and Buffett are on the board. The fact that Stern Stewart gets 20% of the process wrong renders their version of EVA useless. You can't build a skyscraper on an inadequate foundation. So if Munger and Buffett can convince a few more company managers to view Stern Stewarts version of EVA as twaddle, they'll probably save those managers millions in wasted consulting fees and poor resulting decisions.

Evan
11-27-2005, 02:34 PM
[ QUOTE ]
[ QUOTE ]

I think you guys are both saying the same thing. The OP was talking about calculating the cost of equity in general (in the passage you quoted), not specifically the EVA method.

[/ QUOTE ]

Let me be more clear then. Stern Stewart has registered a trademark for EVA, so when you say "EVA" you aren't talking about the general concept, you are talking about Stern Stewarts EVA(R), whose underlying process is kept proprietary BTW. But if you read their materials, it appears clear that EVA(R) always uses Beta to estimate cost of equity. This is why EVA(R) is criticized by Munger.

Other than that, trying to "own" a concept so simple and obvious is the only other thing I'd criticize Stern Stewart for. Of course you never borrow money to invest at a lower return that you pay in interest. Just like you never invest any capital at lower returns than are available elsewhere.

[/ QUOTE ]
It still sounds like everyone's saying the same thing to me. Maybe I'm misunderstanding the OP, but so far I've agreed with everything you've said.

FatOtt
11-28-2005, 02:25 PM
Ok, back in town.

I don't think there's any argument here. My point was that Buffett/Munger agree with the bulk of what goes on with EVA (specifically talking about the idea of economic, rather than accounting, profits), just not some of the specifics put into place by Stern Stewart. So I agree with Evan if he means that Berkshire doesn't use EVA as implemented by Stern Stewart, but I disagree if he means that Berkshire doesn't use the idea of economic profit and a cost of equity/hurdle rate that makes up 90% of what goes into EVA.

The problem that I have is that many people (and, again, I've met a few) will listen to the Buffett/Munger comments and throw out the baby with the bathwater. They hear, "EVA is twaddle" and interpret that as "Calculating economic profits after deducting a cost of equity is twaddle". They hear, "Beta is stupid" and interpret that as "Attempting to calculate a cost of equity or even recognizing the cost of equity as a concept is stupid."

I agree with Evan - it seems we're all in agreement here. Having read prior posts from both Evan and DesertCat, I'm pretty sure we think alike on this.

FatOtt
11-28-2005, 02:29 PM
I'm also not sure about this:

[ QUOTE ]
Second, pretty much everyone who doesn't like the EVA method has the same reason; it maximizes firm value, not equity value. I've never heard either of the Berkshire managers talk about their specific reasons, but I would be shocked if this wasn't it.

[/ QUOTE ]

Economic Value Added (not in the trademarked sense, but in the more general conceptual sense) is often calculated as:

Adjusted Net Earnings of the firm (including interest expense and adjusting for things like R&D and advertising) - [Book Value (as adjusted) * Cost of Equity Capital].

Other than situations where the equity is out of the money (as described by edtoast), I don't see this as a big problem. At the end, considering that any incremental firm values accrues to the equity holders and not the debt holders, I don't see a big deal with maximizing firm value vs. equity value (again, when the equity is in the money).

Evan
11-28-2005, 04:57 PM
[ QUOTE ]
any incremental firm values accrues to the equity holders and not the debt holders

[/ QUOTE ]
You can increase the value of your debt without increasing the value of your equity (or at least increasing it less). Without spending a ton of time coming up with examples, I think it would be pretty easy to envision a senario where optimal capital structure for maximing equity value is different than optimal capital structure for maximizing firm value.