This summer I have taken a Private Equity course at the University of Texas at Dallas as part of my MBA coursework. In this course, the instructor employs case based discussions along with industry speakers.
In the first few cases, we focussed on Levered Buy Outs (LBO). It was interesting to learn that private equity investors also focussed on cash flows just as a public equity investor should. The measure often used by private equity investors is EBITDA or Earnings Before Interest Taxes Depreciation and Amortization. Now, EBITDA is not the same as cash from operations however it is often used as a quick shortcut for cash flows.
In an LBO, a private equity investor takes a public company (or a division of a public company) private by taking on more debt or levering up the company. The rational is that the company has good consistent cash flows and those cash flows can service the debt taken on. The Private equity investor would typically contribute 20-30% of the transaction amount in equity and the rest would be debt. After the financial crisis of 2008, the equity investment needed in LBOs is in the 40-50% range and the transaction sizes have greatly reduced.
One of the reasons for using EBITDA rather than EBIT or Net Income is that when you compare companies across the world which have different capital structures as well as different tax rates and depreciation schedules, it makes sense to adjust for all these differences. EBITDA is an easy way to adjust for this and compare such companies. In the podcast by theValueGuys, Val also often refers to the EV/EBITDA as a form of valuation. Read about this show in this post.
I found that Tom Russo also uses EBITDA as a measure when looking at investment candidates. Here is an exercept from his talk at the Value Investor Congress
In 1989 (I started investigating foreign companies in ’87) this is where it starts. Weetabix’s EBITDA was 19 million pounds and it had 7 million pounds of cash. I determined the business was worth 8 times EBITDA plus cash, less debt, and I came up with an intrinsic value of 13 pounds. With 11.87 million shares outstanding it was trading at 5.95 pounds per share. Then each pound was worth US$1.61 So that’s the beginning of the story. Now, I would mention that I used the word EBITDA, which Buffett this weekend suggested was false signs and all sorts of horrors would ensue if you even mentioned those six letters in rough approximation to each other. But for me at least, EBITDA is useful because it helps me cross countries because of different depreciation schedules and amortization schedules.
It’s one thing to say in a U.S. context that you’ve looked at pre-tax earnings rather than EBITDA. But when you start to look at Heineken versus Cadbury Schweppes versus Budweiser, and if you don’t adjust for those non-cash charges — and one way to do it is using EBITDA — you have a more difficult time comparing across the market. I have used EBITDA, and I confess it in front of you.
Now, I also consider the EV/EBITDA multiple of a company I am trying to invest in or consider selling along with other measures of valuation. I also consider the EBITDA multiple of its public competitors to see what is being offered in general for similar companies.
Let us look at some fashion retailers.
American Eagle (AEO) : 4.5
Abercombie & Fitch (ANF) : 5.9
Aeropostale (ARO) : 5.1
Hot Topic : 3
J Crew 6
Gap 3.6
Retailers throw out good cash flow most of the times, however they can be considered cyclical and often fall out of favor with their fashion fickle customers.
Caveats:
1) Capex: EBITDA does not take into account the capex required by a business (as compared to Free cash flow which is Cash from operations – Capex). An adjustment that one can make is to look at
EV / EBITDA – Capex.
After accounting for capex if the valuation is not as attractive, then you can pass it over.
Example: Consider Tenet Healthcare (THC)
EV : $5.85 billion
EBITDA: $1 billion
You will see it trades at a somewhat attractive 5.84x EBITDA. However, if you subtract capex of $456 million from EBITDA, you will arrive at this adjusted measure of 10.74x. Clearly, not as attractive as originally thought.
2) Interest coverage/ high leverage: EBITDA also ignores the interest payment required and the existing level of leverage in the company. I would advocate looking at two additional measures.
(i) Interest coverage as indicated by EBIT/ Interest expense. A low interest coverage should be a red flag.
Coming back to our THC example, interest coverage = 1.46. A very low number.
(ii) LT Debt / EBITDA or Debt / Equity: Graham advised against investment in a company where Debt is 2x equity.
If you look at THC, LT Debt = $4.27 billion and equity = $646 million.
So, LT Debt / EBITDA = 4.27x and Debt / Equity = 6.6x.
Disclaimer: Long ARO
Great article Adib… I’m curious that since the issue, mostly, with EBITDA is not being able to account for non-cash charges, would it be simpler to use Buffett’s “owner earnings” instead of EBITDA in instances where its used? Would love your opinion. Thanks.
Hi Robert,
thanks for the great question.
My preferred valuation method is EV / FCF in most cases. I define FCF as Cash from Operations – Capex. I do not break out Capex into maintenance and growth capex unless its given in the SEC filings ( example: ESV). Owner earnings is very similar to that and is even more conservative I think. So, yes use owner earnings everywhere or FCF. Surely prefer it to EBITDA.
I think if you want to look at EBITDA, use EBITDA – Capex atleast. So, a company that looks cheap on EV / EBITDA may not look as cheap when you account for capex. I gave the example of THC. Check those numbers i think THC is still public.
I do look at EBITDA to compare companies in the same industry at times. Like retailers have less of capex. So, I find an ARO at 4x EBITDA v/s JCG / GYMB that are going private at 8x EBITDA.
Hope that helps.
Why are public EBITDA/EBIT multiples typically higher than private EBITDA/EBIT multiples?
Liquidity. Its much easier to exit a position in a public company. Hence, the higher multiple for public companies. Thats my best guess.