Rules of Thumb for Ratios Using EBITDA, EBIT and Net Income

[rukawa]

I normally think in terms of PE and my guideline would be <10 P/E cheap, around 15 p/e normal and >20 expensive. However a lot of people use ratios based on EBITDA or EBIT and the question is how to convert between these?

 

I'd like to start a discussion. I propose the following based on the table below:

EBIT/Net Income = 1.5 so for any EBIT ratio multiply by 1.5 to obtain the corresponding P/E ratio.

 

E.g. EV/EBIT of 6.6 means a PE of about 10.

 

Similarly EBITDA/Net Income = 2 and thus EV/EBITDA of 6 translates to a P/E of 12.

 

I would propose the following cheap, normal, expensive scale.

Cheap is < 5x EBITDA, < 10x Earnings, < 6.6x EBIT

Normal is 8x EBIDTA , 16x Earnings, 10x EBIT

Expensive > 10x EBIDA, > 20x PE, >13x EBIT

 

Now something to assist in this is the following table which provides the total earnings, revenue, market cap, EBITDA, EBIT, Net Income, Capital expenditures for all companies that have positive earnings. Based on this you can obtain the average EBIT/Net Income worldwide or in the USA. The attached spreadsheet contains all the source data I used to calculate this. The data was downloaded in 2014.

 

.	|	USA	|	World
Total Revenue	|	14,215,502	|	50,202,975
Total Debt	|	5,401,687	|	19,192,144
Market Capitalization	|	22,390,228	|	58,851,933
EBITDA	|	2,447,460	|	7,635,562
EBIT	|	1,809,426	|	5,338,728
Net Income	|	1,171,988	|	3,647,796
Capital Expenditures	|	805,240	|	3,193,362
P/S	|	1.58	|	1.17
P/E	|	19	|	16
Profit Margin	|	8%	|	7%
EBITDA/EBIT	|	1.35	|	1.43
EBITDA/Net Income	|	2.09	|	2.09
EBIT/Net Income	|	1.54	|	1.46

 

And yes I understand that I really should not be using multiples and also that different industries should have different multiples. I should instead be discounting future cashflows. But I like rough rules of thumb.

financial_data_for_all_companies_worldwide.xlsx

[Shooter MacGavin]

With all due respect, I don't think your rules would be useful to you.  I'm using examples at extremes to make the point:

 

Let's say you have a company that has EV/EBITDA of 6x but its levered and the company's interest charges and depreciation (or maintenance capex) is consuming 100% of EBITDA.  Then an EV/EBITDA of is 6 equal to P/E of infinity.  Or in other words, the company, with its current capital structure is not adding any value to shareholders.  If no changes can occur to the cost structure, the shares of this company are worth zero.

 

Now let's say you have a company that has a recurring annuity stream of a software license with an EV/EBITDA of 6x, fully funded by equity, and has no capital requirements.  In this case, an EV/EBITDA of 6x is more like a P/E of 12.5x with taxes being the only leakage from EBITDA.

 

This is why you can't compare EV/EBITDA of a capex heavy company to an EV/EBITDA of a software company. But you certainly can compare their multiple on economic earnings.

 

At the end of the day, the EV/Rev, EV/EBITDA, EV/eyeballs multiples are a shortcut to a multiple on "economic" or "run rate" earnings. 

 

Now let's say you had two companies trading at 10x.  The first was able to reinvest 100% of the economic earnings at 20% incremental returns on capital.  The second reinvested 100% of economic earnings at 5%.

 

Well, Is 10x cheap for both?  Not really.  The first company will make you 2.5x your money in 5 years (20% IRR), assuming the multiple is flat.  The second will make you 1.2x your money (5% IRR).  Therefore, you can't generalize and say 10x earnings is cheap.  10x is not cheap if companies are reinvesting at low rates of return.  20x earnings is extremely cheap if companies are investing at very high rates of return.

 

Let's complicate this further.  10x may be cheap if a company is reinvesting at 1%, but it has $100B of excess cash or real estate that isn't factored into the stock price.  But it may only be cheap if shareholders have a way to force management to either do something useful with those excess assets or distribute it, hopefully in a tax efficient manner. 

[Travis Wiedower]

Maybe it's my lack of financial/Wall Street background, but I'm amazed at how common the use of EBIT and EBITDA are. There are a lot of very real expenses below those lines!

 

Also, all those ratios mentioned can be fine ways to quickly gauge what a company is selling for, but don't forget that a company is worth its future cash flows, not past cash flows. A company selling for 30x trailing EBITDA/EBIT/earnings can be cheap and another selling for 8x can be expensive.

[KJP]

 

I would propose the following cheap, normal, expensive scale.

Cheap is < 5x EBITDA, < 10x Earnings, < 6.6x EBIT

Normal is 8x EBIDTA , 16x Earnings, 10x EBIT

Expensive > 10x EBITDA, > 20x PE, >13x EBIT

 

 

Before discussing whether they make sense, it would be helpful to know what you want to use these "rules of thumb" for.

[tooskinneejs]

The best advice when you see companies touting adjusted earnings results: Run the other way.

[rukawa]

Before discussing whether they make sense, it would be helpful to know what you want to use these "rules of thumb" for.

 

A quick determination of whether something is cheap or expensive.

 

Well, Is 10x cheap for both?  Not really.  The first company will make you 2.5x your money in 5 years (20% IRR), assuming the multiple is flat.  The second will make you 1.2x your money (5% IRR).  Therefore, you can't generalize and say 10x earnings is cheap.  10x is not cheap if companies are reinvesting at low rates of return.  20x earnings is extremely cheap if companies are investing at very high rates of return.

 

You say 20x is extremely cheap but only if companies are investing at high rates of return. But that is the point! When I see a 20x I am going to expect that the company earnings should have very high quality and growth to justify the price I am paying. But I would never expect that if the ratio was 4x. If it was 4x I would ask the opposite question, why is it so cheap?

 

These ratios are informative because they structure how you analyze the company and they enable you to make quick decisions.

 

 

[arcube]

Good thread and question Rukawa. Great answer Travis!

 

M&A market touted the frequent use of EBITDA till some slick public companies' CFOs realized that they can use it as well to CYA!...and then Micheal Pearson started using it a lot...

 

Maybe it's my lack of financial/Wall Street background, but I'm amazed at how common the use of EBIT and EBITDA are. There are a lot of very real expenses below those lines!

 

Also, all those ratios mentioned can be fine ways to quickly gauge what a company is selling for, but don't forget that a company is worth its future cash flows, not past cash flows. A company selling for 30x trailing EBITDA/EBIT/earnings can be cheap and another selling for 8x can be expensive.

[rukawa]

M&A market touted the frequent use of EBITDA till some slick public companies' CFOs realized that they can use it as well to CYA!...and then Micheal Pearson started using it a lot...

 

I blame John Malone. I have never been a fan of EBITDA. But as much as its criticized its still used extensively on this forum and on VIC. I accept that its going to be used...given that I would like to have some idea of what if anything it indicates.

[KinAlberta]

A few times a year I just google EV/EBITDA to see what shows up. Today's search yielded the article below. The fact that its about SocGen research caught my attention.:

 

 

EV/EBITDA ratio for the MSCI AC universe is at the highest level in more than two decades

By Rupert Hargreaves on April 10, 2017

 

 

 

...

EV/EBITDA is very high – SocGen

 

Equity investors may be celebrating such a strong first half equity market performance but SocGen’s analysts urge caution. The team notes that such a strong equity market performance coupled with a continuing buildup in corporate debt and sluggish profits growth has pushed stock valuations up to historical highs. The current median EV/EBITDA ratio for the MSCI AC universe is close to 12, the highest level in more than two decades. The median reported P/E is around 18.5 below the two-decade high of around 21.

 

Still, while markets may look cheap on a P/E basis, according to SocGen’s figures, based on historic performance the average one-year ahead excess returns over US 10 year Treasuries for the market coming from a starting EV/EBITDA ratio of 12 or more is less than 2%. As the note concludes, “if you think equity deserve a risk premium then you have a problem.”

 

“The question then is does valuation matter? The answer depends on..."

 

http://www.valuewalk.com/2017/04/equities-evebitda/

 

[aceskc]

The only metric that matters to me is EBITDA- normalizend mCapX- working capital changes - Taxes- Interest ...I ignore all other multiples and all other types of earnings. Even i leveraged stub type situations this is a good starting point, and you would ideally want it to grow from here .

 

BTW Rukawa, Where can you get data in this type of downloadable format?

 

[rukawa]

The only metric that matters to me is EBITDA- normalizend mCapX- working capital changes - Taxes- Interest ...I ignore all other multiples and all other types of earnings. Even i leveraged stub type situations this is a good starting point, and you would ideally want it to grow from here .

 

BTW Rukawa, Where can you get data in this type of downloadable format?

 

Cap IQ