If EBITDA = "BS earnings," is EV/EBITDA a great valuation measure?

[dabuff]

"BS earnings" of course being how Charlie Munger references EBITDA (and in similar terms, Klarman does the same in "Margin of Safety").

However, it seems EV/EBITDA outperforms other valuation multiples, including P/E and P/B in multiple books and studies (Acquirer's Multiple, What Works on Wall Street, and much more).

 

What do you all make of this? Is EV/EBITDA a good ratio across many stocks because the inputs are easily comparable across industries? Should it be reserved to basket/mechanical investing? Or is Munger simply encouraging us to be more judicious in our evaluation of companies beyond simply reducing it to a single number?

[TwoCitiesCapital]

"BS earnings" of course being how Charlie Munger references EBITDA (and in similar terms, Klarman does the same in "Margin of Safety").

However, it seems EV/EBITDA outperforms other valuation multiples, including P/E and P/B in multiple books and studies (Acquirer's Multiple, What Works on Wall Street, and much more).

 

What do you all make of this? Is EV/EBITDA a good ratio across many stocks because the inputs are easily comparable across industries? Should it be reserved to basket/mechanical investing? Or is Munger simply encouraging us to be more judicious in our evaluation of companies beyond simply reducing it to a single number?

 

I think they say that because it's abused by companies as opposed to judging its overall usefulness.

 

It's like when analysts say "excluding energy, everything looks good." No shit. It's always the case that everything looks good if you ignore all of the bad. EBITDA allows people to ignore all of the bad and companies abuse that option.

 

GAAP has it's own problems, but at least it generally tries to not mislead people. EBITDA has the potential to ignore the replacement costs of running a business. Pro-forma earnings ignores everything that went bad in the quarter. Adjusted EBITDA ignores both. Etc. etc. etc.

[TheAiGuy]

I don't value with EV/EBITDA.

 

None of these measures will tell you want you want to know, which is typically price per value created over unit of time. Sometimes GAAP E or EBITDA will get you close enough to "value created over unit of time" but more often than not, in any type of value situation, that won't be true. Some times, measures like like "adjusted X" (where X is EBITDA, FCF, whatever) that management provide are pretty good estimates of Value per unit time. Most often, they're too rosy.

 

If you want something better, you'll have to dig through the accounting notes, adjust earnings and/or cash flow in a way that makes sense to you based on your understanding of the business and how you see it progressing, and value of that.

 

Edit:

In some really interesting situations, you won't be able to estimate value per unit time very well at all, and price can become unhinged (in both directions).

[augustabound]

I think they say that because it's abused by companies as opposed to judging its overall usefulness.

 

That's how I read it also. It's one thing for investors to compare companies using certain multiples but something entirely different when management uses it.

[Liberty]

These are all just tools that can be misused or used properly. The idea is always to learn something about the underlying economics of an asset. Different tools can answer different questions. GAAP EPS can be very misleading in some cases, just like EBITDA or P/FCF can be...  Same for ROIC, ROE, ROA, etc.

 

If there was a perfect metric out there, we'd all know by now and everyone would be using it for everything.

 

For example, if you take two almost identical businesses, one with high FCF and another with no FCF, most people will at first glance prefer the one with high FCF. But if you do some work to understand the underlying economics and realize that the first business is quite good, but has no reinvestment opportunities, so it is basically a cash cow, while the second business is just as good, but has huge reinvestment opportunities that can absorb all of its FCF to generate profitable growth, then maybe the best choice is the latter. You can't learn that just from looking at a metric.

 

Same if you find a business with super high ROIC. That by itself sounds attractive, but the number won't tell you if this is a sustainable position, if there are barriers to entry for competition, if management is good, if the business is cyclical and currently at the peak of profitability, etc.

 

 

[Guest Grey512]

Not to nitpick, but doesn't EV/EBIT outperform EV/EBITDA?

[scorpioncapital]

I like EV/"Warm & Fuzzy"

 

I.e. Is the business alive and thriving, is it a remarkable product or service with great future prospects. That will make a much bigger difference to success I think than just the numbers, even if they look decent.

 

 

[KCLarkin]

Maybe EV/EBITDA outperforms due to the EV not the EBITDA. If you buy companies with a very low PE, they will tend to have more debt and less cash.

[dabuff]

If I recall correctly from the mechanical investing studies EV/EBITDA > EV/EBIT > P/E.

 

Bottom line if you choose such a basket strategy you'll be well served by any of the above.

[CorpRaider]

Not to nitpick, but doesn't EV/EBIT outperform EV/EBITDA?

 

Yup.

[ni-co]

"BS earnings" of course being how Charlie Munger references EBITDA (and in similar terms, Klarman does the same in "Margin of Safety").

However, it seems EV/EBITDA outperforms other valuation multiples, including P/E and P/B in multiple books and studies (Acquirer's Multiple, What Works on Wall Street, and much more).

 

What do you all make of this? Is EV/EBITDA a good ratio across many stocks because the inputs are easily comparable across industries? Should it be reserved to basket/mechanical investing? Or is Munger simply encouraging us to be more judicious in our evaluation of companies beyond simply reducing it to a single number?

 

Like Liberty, I think the answer to the last question is clearly: yes. Think about what you're doing when using a measurement is the message here.

 

EV/EBITDA is meant to give you an idea about a company's earnings power without regard to the underlying financial structure and taxes. Why is this useful? Because, with enough money, you could take control of the company and change its financial structure, i.e. reducing or adding to its debt load – and, in theory, there wouldn't be much of an effect on the underlying business. Since debt payments are tax deductible, the financial structure of a company has huge influence on a company's tax bill. Further, with complete control, you may also be able to move the company into a lower/higher tax jurisdiction, you may decide to spend money on growth - which also reduces your tax bill - and so on. This is the reason why EV/EBITDA is very important for PE investors who take control of companies.

 

I think the reason some value investors despise it is because it doesn't take into account maintenance capex – in other words the money you need to reinvest to keep the flywheel going. You might be able to delay those investments for some time but in the long run you'd shrink/eventually kill the company.

 

Therefore, from a value investor perspective EV/(EBITDA - maintenance capex) is the way to go for a large variety of companies. There are industries that require only very little maintenance capex, e.g. most software, internet or service companies. Because of this, EV/EBITDA gives you a reasonable guess at their earnings power and (with regard to their long term EBITDA) cheapness.

 

Because EV/EBITDA is obviously "before interest", it doesn't work for companies earning their money in interest (like banks).