Last week's post on valuation brought a couple questions about EBITDA. The first of which is how do you pronounce that acronym? The answer to that question is e-bit-dah. The second of which is what does it mean? The answer to that is Earnings Before Interest Taxes Depreciation and Amortization.

The way I like to think about EBITDA is the pre-tax cash earning power of the business. It is not much different than the notion of Operating Income which is revenue minus cost of goods sold and operating expenses. But it takes out the two big non-cash items in an income statement, depreciation and amortization.

EBITDA originated in the buyout world where you use a significant amount of debt to buy a company. Since interest costs are tax deductible, you can load a company up with debt and not pay taxes. If you want to figure out how much you can borrow, you look at EBITDA and that is the amount of interest you can pay to wipe out all of your taxes.

Let's use an example. Let's say you have a business with no debt that does $100mm in revenue, has $20mm in cost of goods sold, and another $60mm in operating expenses. Let's say that you have no amortization and $5mm of depreciation. Then your Operating Income is 100-20-60 or $20mm. If your tax rate is 40%, then you will pay $8mm in taxes. Your Net Income is $12mm (20-8). And your EBITDA is 20 + 5 or $25mm. You can also get to EBITDA by taking the Net Income and adding back Interest, Taxes, Depreciation and Amortization. In that method EBITDA = 12 + 0 + 8 + 5 + 0 = $25mm.

If interest rates are 5% and you can borrow interest only, then you can borrow $500mm of debt, pay annual interest of $25mm per year, and no taxes because your interest costs wipe out your net income. After doing that your  Net Income goes to 100-20-60+5-25 = 0.

So that is where EBITDA comes from. But there are a bunch of problems with EBITDA. First is that even though depreciation is a non-cash expense, it is the accountants way of estimating how much capital investment you must make in your business every year. If you don't have any money left to continue to invest in your business infrastructure, you'll be in trouble. That's why a lot of people prefer EBIT to EBITDA.

It is also true that if you borrow so much that you have no margin for error, you are likely to run into a problem and go bankrupt. So buyout investors don't load up on so much debt that they use up all of EBITDA paying interest. What they do instead is value a business as a multiple of EBITDA. And that multiple is usually in the single digits (5, 6, 7, maybe 8). In our scenario, the business we talked about would be worth $150mm at 6xEBITDA. That's a very big difference from the $500mm you could borrow if you were willing to apply every penny of cash flow to paying interest. And a 5% interest only loan is not particularly common in a buyout scenario either.

But in any case, this is not really a post about buyouts. I'm not an expert in that topic. If the MBA Mondays audience is interest in a post or two about buyouts, I have some friends who can provide that. This was just an attempt to explain EBTIDA and give you all some context for where the measurement comes from and why. I hope it did that.

#MBA Mondays

Comments (Archived):

  1. Pascal-Emmanuel Gobry

    I think EBITDA isn’t just a measure of how much you can load up the business with debt, but also of how much the “core” business earns, taking out the cost of investment and debt. 

    1. fredwilson

      right, “the pre-tax cash earning power of the business”i saw your tweet about sequoia. that’s not anywhere near true. i like to think that USV is trying to create a sequoia like VC firm on the east cost (or benchmark, greylock, etc, etc, etc)

      1. negative beta

        I work on a sell-side trading desk as a research analyst covering the levered loan and credit space.  I liked your post on EBITDA and noticed that you touched on the topic of LBO’s.  While I understand this is not the focus of your blog, I see that MBA Monday’s sometimes touch on topic like this.  If you have any interest, I would love to do a quick write-up on the LBO topic if you care.

  2. Fernando Gutierrez

    I had a boss who was one of those you mention that prefer EBIT. She threatened to fire me if I ever went with a proposal that stopped at EBITDA level. Investments were important in the business (telecom) and she wanted us to always keep an eye on amortizations.

    1. Guest

      I have major issues with EBITDA myself. It is simple and therefore useful and such but many folks do not understand its limitations (what it is NOT telling the reader) and as such can get folks into trouble. 

  3. Rob Ganjon

    Large capital investments with advantageous depreciation schedules can produce big P&L losses, which can be really nice for the business owners’ tax situation. EBITDA helps to cut all the income statement tricks out to show you the “cash earning power” as you said. Thats the real importance of the EBITDA.

    1. fredwilson

      Yes exactly

    2. Daniel Nunes

      Imagine you invest 3 million dollars in servers to open a data center. In three years, it generates 1 million dollars in hosting fees. Suppose a 3-year amortization of the servers.Assuming there are no more costs 😛 and without regarding to present values, annual EBITDA of the company is 1 million dollars … 3 million after three years.EBIT is zero and in the end of the three years it is also zero. It is reasonable to a company whose value is zero, to pay taxes? I think not!Depreciation is a cost that measures the value needed to reconstruct the fixed assets of the company.I bet that in this imaginary company, managers will be rewarded because the annual bonuses are calculated by EBITDA … and not by EBIT.In a side note, there is some discussions to incorporate opportunity costs in the accounting systems … I think managers will fight hard against it: after all the annual bonus can not be at risk.

  4. Carl J. Mistlebauer

    I think you make a great point about EBIT/EBTIDA and the concept of capital investments.  So many times I have seen companies that were “cash cows” dry up because of the a laser beam focus on not paying taxes.But then again, when your finance vocabulary centers around terms like “cash cow” it just might be a reflection of my own stupidity.  

  5. Avi Deitcher

    I believe your operating income should be $15MM, not $20MM. Depreciation is considered for operating income, that is the whole point of depreciation, to have a major capital investment depreciate over time, the amount that impacts that year’s revenues. So if revenue is $100MM, COGS is $20MM (nice 80% gross margins), operating expenses are $60MM, and depreciation is $5MM, then you have $100-20-60-5 = $15MM. Depreciation is also recognized for tax purposes (although using different schedules, but we will assume that it is the same $5MM for tax purposes), so at 40%, your taxes are $6MM, for a net income of $9MM, instead of $12MM. Oh, yeah, and EBITDA should be $20. Of course, that makes sense, since it is “before depreciation”.Your cash flow, on the other hand, ignoring taxes for the moment, by the indirect method, is just your operating income of $15MM, and then add back non-cash-expenses of $5MM to get $20MM.

    1. fredwilson

      i wasn’t thinking of it like that. i was thinking the $5mm of depreciation was included in the $60mm of operating expenses.i guess i should have clarified that

      1. Avi Deitcher

        Makes sense.

  6. Guest

    Fred, I would think that the gap between an adjusted Op Inc figure (or an actual Op Inc figure) and EBITDA would be narrowing in your world. Isn’t there a lesser need to make CapEx given cloud-based resources for internet startups and therefore less depr. and amort.? I realize there might still be capped software costs but I would think in your world the gap between these two reporting metrics has narrowed since you first entered this investment space.{Updated}

    1. fredwilson

      yes, you are right for smaller companiesbut when the companies get very large (zynga, twitter, tumblr, etc), they need to own their datacenters

      1. markslater

        at what point have they made the switch fred? 

        1. fredwilson

          around 100mm monthly regular users

          1. markslater

            coulda guessed you were in the turntable.fm deal – congrats! ask them to add a proper MIX feature (beat matching minimum).

      2. PhilipSugar

        Just to be clear, when you say “own datacenter” you mean leasing space in a datacenter?I think once you get to one rack of servers the cost vs cloud is very easy to justify.

        1. fredwilson

          yes, that’s what i meantthanks for asking me to clarify

  7. Greg Kamstra

    The other nuance to this point is that from an investor’s standpoint, EBITDA (& EBIT as well) is capital structure agnostic.  If a business is currently loaded up with a ton of debt & making large interest payments, you don’t want to consider that when making your investment, because most private transactions are done “debt free, cash free,” so once you buy the company, the interest expenses will likely be totally different (or zero), & the amount of interest you’re paying will be based solely on how you structured the transaction.   

    1. fredwilson

      that is a great point. i should have stated that in my post.

      1. Jan Schultink

        At McKinsey we used NOPLAT: net operating profit less adjusted taxes as a basis for DCF models: build financial forecasts as if the company had no debt (focus purely on operations), then reflect the value of the tax shield created by debt in an adjusted discount rate. http://en.wikipedia.org/wik…

  8. Matt A. Myers

    Cleared up a few things for me (including the followup comments). Thanks all.

  9. Trish Burgess-Curran

    From my perspective the real value comes from using several different metrics – including EBITDA, EBIT and Operating Income – to meassure the revenue and profit generation capabilities of a company. They all have pros and cons and, although, in the end, they never paint the full picture, they all highlight relevant aspects. I am sure that I am preaching to the chorus…I can see where Deitcher’s confusion was coming from since, at least in my experience, there is no universally accepted definition of many of the terms used including Operating Expenses. Not a problem though, just needs a bit of additional specification as Fred provided!I also agree with Greg that it is important to take into account how the transaction has been structured but, at the same time, this is not relevant when analyzing the ‘core earning power of the business’ as a whole but rather the profit options for a specific investor. Not quite the same…

    1. Carl J. Mistlebauer

      I have seen lots of situations where investments were made on the logic that the “profit options for a specific investor” and the “core earning power of the business” did not match and in every situation the investor ends up getting burned.Unless an investor has no interest in the business then the core earning power is critical and at one time I might have agreed that there was a thing as old economy and new economy but the more I study Group On the more I realize there is not that big of a difference.Looking at Group On’s core earning power I have no idea where an investor would find profit options.

  10. celestus

    I was going to say before you got there yourself, no lender in the world is going to lend $500 million to a company that has $25 million in EBITDA.At that level, probably $50 million at 8% or so and another $25 million at 12%.  And I don’t think you’d even get that for a pure technology company.

  11. Jan Schultink

    EBITDA is a proxy for a business’ operating cash flow generation ability that can be pulled quickly from financial statements and databases.Capital-intensive data center operators are an example where investments in infrastructure become so huge that you start to need to take that into account almost as an operating expense. EBITDA of these companies looks great. EBIT is a different story.

    1. Kasi Viswanathan Agilandam

      data center??  some servers? we get dime a million!!! like ideas.

  12. Daniel Nunes

    All items of a profit and loss statement have direct or indirect influence on … cash.If a CEO (or CFO!) calculates the prices of their services and do not look for the annual depreciation of its assets, he may calculate prices that do not cover the reconstruction of its assets. That’s the way, amortizatizon call also be called reintegration. This will influence the company’s cash flow due to a poor estimate of the sales price.For me EBITDA is dangerous non-GAAP metric … I sleep better when I use EBIT. 

  13. Brad

    Fred,Excellent post. How do you determine the multiple to utilized during liquidation?Thanks in advance.

    1. fredwilson

      that was last week’s post :)http://www.avc.com/a_vc/201…

      1. Brad

        that is right. sorry for the redundant question.

  14. Thekid

    you sure did! Thanx!

  15. Philip Lang

    When we talk about EBITDA as a proxy for free cash flow (which is typical in the investment banking world), it is also important to mention that capital expenditures have a major impact on the gap between free cash flow and EBITDA.  While capex and EBITDA are related, a growing business will have much higher capital expenditures than depreciation.  From a valuation perspective, industrial companies are often trade on an EV/EBITDA basis because it is capital structure agnostic, whereas tech companies are more often traded on a P/E or P/E/G basis because EBITDA is often negative (or 0) for early-stage companies.  Your example of $500mm of debt on $25mm of EBITDA implies a 20x leverage multiple, and it would be likely impossible to find a bank willing to provide that level of leverage.

  16. PhilipSugar

    I used to think that EBITDA was devised as a way for unprofitable telcom, cable, and high tech companies to at least put a positive number up on the board.However, I’ve come to see it does have its uses.  One is that hopefully as you both grow and learn you become more capital efficient.  Also from a tax perspective you’d like to depreciate as fast as you can so that can help reduce the DA as well.Pulling out interest and taxes are removing two variables.that really aren’t dependent on the health of the business.

  17. Kyle S

    IMHO, the reason that buyouts often blow up (and they do) is not due to the finance people not understanding that depreciation isn’t a free lunch. Models always account for capex in some way, and mostly use realistic assumptions. The problem is usually that a business that has been doing well is already employing a good deal of operating leverage; once it covers its fixed costs, additional revenue that comes in does so at a high margin. Examples of businesses like these I’ve seen include examples as discrete as casinos and ready-mix concrete businesses. The trick is that in a downturn (say, people stop coming to the casino, or home building drops 30%) that operating leverage disappears and decreases in revenue have a disproportionate impact on EBITDA. A business that has 30% EBITDA margins at a given revenue level might only have 15% margins at 75% of that revenue. If you paid 8 times at 30% margins, all of a sudden that transaction is over 20 times at the company’s new (lower) revenue number. Hope you didn’t use too much leverage!Of course, lots of internet businesses (like salesforce) have incredible operating leverage and they’ve been growing like crazy, so they trade on very high ebitda multiples. If that growth ever stops (or even slows dramatically) the train jumps the tracks. Witness Opentable, which is down 50% from its highs earlier this year despite reporting 50% y-o-y revenue growth and 100% y-o-y operating cash flow growth last quarter. just the HINT that growth might slow faster than anticipated made the company worth 50 percent less in the mind of the investing public.

  18. Kasi Viswanathan Agilandam

    I love this tribe….let us find that land to live.

  19. Xander Mahony

    Worth noting that if valuing a company as a multiple of EBITDA (or EBIT or revenue) you’re looking at an enterprise value (debt plus equity) because all of these are earnings available to all financial stakeholders (because you haven’t subtracted interest – the debt holder’s share).  When you get to net profit, it’s just the profit left for shareholders, hence multiples of net profit give you the value of equity only.  For capital intensive businesses EBIT is preferred because depreciation is a large factor, but for most internet/software businesses EV/EBITDA is fine.  Obviously if EBITDA is negative or trivial, stick with EV/revenue to get a sensible value.

    1. JLM

      Damn good comment and very insightful.  Well played!

  20. jeff trudeaux

    Fred,Great post that makes this very, very simple.I have an other question regarding valuation… what is your take on valuation caps while raising seed funding through convertible debt?  I’m currently raising funds and one investor has asked about this (the others have not).   It seems like it is forcing me to come up with a valuation (pure art at this point) to provide some peace of mind for the investors.  and also, if a val cap is set, why even do a convertible note?

    1. fredwilson

      ah, that last question is the besti don’t like convertible notes for seed funding and refuse to do them, cap or no capi prefer a priced round. it’s better for everyone

      1. Kerim Baran

        Why? What is the downside if the convertible note has a reasonable cap?Kerim

        1. fredwilson

          i just want to know how much i own when i part with my moneyand i want the entrepreneur to know how much it cost them in dilution

          1. Jeff Trudeaux

            Thanks for the response, and reading it I just realized that asking you I should always get the response that you don’t like convertible debt! For me, and my ownership, conv debt is the best option.  But, if holding out for that means not being able to create my vision, my stance will change.In the end, I will still be happy having some equity in a real business vs. all the equity in something that never got off the ground.

  21. Rohan

    Nice. makes it simple. Thank you! 

  22. David Smuts

    Hi FredLove your MBA Monday’s. In answer to your question about interest in a post or two about buy-outs; YES would love a post or two. In particular, would love to know more about the process of an acquiring company using its’ shares as capital for a target acquisition. How that is structured? How is it valued (discounted?) etc.., I am particularly interested in the process of utilising a SPAC as a means of acquiring a target company.

  23. Lbartlett16

    Random posting, not sure where else to put ithttp://horriblehiphop.blogs…

  24. Kyle Pearson


  25. Kyle Pearson

    another vote for a post or two on LBOs. Love MBA Mondays!

  26. JLM

    It is very important to “derive” EBITDA line by line at the beginning of any discussion about a particular use of EBITDA.By starting with a GAAP term (earnings) and deriving EBITDA by adding back interest, taxes, depreciation and amortization you can look at each line item and number and know its relative impact.It is also important to know the actual basis for the inclusion or calculation of each number — depreciable asset basis and schedule, amortizable asset base and schedule, tax assumptions, principle/interest rate/term — in order to evaluate the relative sensitivity of each number.More and more, I am convinced that depreciation is an actual operating cost of a company particularly given the literal consumption of hardware and the functional obsolescence of technology.Look at a number of different assumptions as a means of evaluating a range and probability of numbers.  Often there is not a single number but rather a range of outcomes that is valid.This modeling should also be fairly broad.It is A tool not THE tool.

  27. Bill Earner

    Hi Fred,Thanks for the great post and the awesome MBA Mondays series, I really have been enjoying it.One issue regarding EBITDA that hasn’t really been brought up yet: it isan earnings-based measure and is thus prone to manipulation via “creativeaccounting” much like net income is. Things like aggressive capitalization of costs, inappropriate revenue recognition, and understated accruals can artificially inflate EBITDA and drive it out of whack with the true underlyingcash flow.  I believe Global Crossing got caught playing games with its EBITDA in order to avoid violating debt covenants that had specific EBITDA targets. One other related that I think you ought to think about when using EBITDA as a measure that hasn’t already been talked about in depth in the comments:EBITDA ignores working capital changes (for instance, you wind up thinking you’ve got higher cash flow then you really do in high working capital growth periods), thus making EBITDA a less-than-ideal measure of liquidity or actual cash. 

  28. Shawn

    Hi Fred, sorry if I am being stupid, but just wondering why you didn’t take out the depreciation expense of 5M to get to your net income?Wouldn’t it be 100-20-60-5=15*0.6= a net income of 9M?I was under the impression that +5 in depreciation will result in -3 in net income.Thanks.

    1. fredwilson

      because i assumed depreciation was in the operating expense number

  29. NedStata

    I’d benefit from a post on Buyouts if you could ask your friends to contribute.thanks and keep up the very useful/informative blogs!

  30. Finance

    Thank you for simplifying the application of EBITDA Fred. Small plug for interested finance students: If anyone is interested in learning more about understanding financial statements with a private tutor, please let us know at http://finance-tutor.com/. 

  31. Rob Bakshi

    Is there any book available which gives a primer on corporate finance without getting into much detailed calculations, more like principles of corporate finance then hands on approach. I liked Financial Intelligence by karen berman and joe knight so far.

  32. Stefano Barazzetta

    Just discovered this blog, it’s really amazing. Thanks Fred, and thanks to all the community!