Determining Valuation Multiples
Last week on MBA Mondays, I talked about valuing an internet marketplace business. In that post, I talked about using 1x gross marketplace transactions and 20x EBITDA as multiples to determine value. In the comments, I was asked about multiples for other sectors. That's a good question so I figured I'd show how to calculate multiples for various sectors.
For this exercise we will focus on the software as a services (SAAS) sector. The first thing you need to do is find a universe of publicly traded companies to use for your model. I found this blog post and used a subset of the companies on the list.
The next thing you do is create a spreadsheet with a bunch of companies on it. I decided to use five SAAS companies in my model; Salesforce, NetSuite, Constant Contact, Taleo, and RightNow. The spreadsheet with these five companies is here.
Please don't get too caught up in the numbers in the spreadsheet. I believe they are accurate but I did this work in about twenty minutes this morning and there could well be errors in it. The point is to show you how to do this work, not to build the absolutely most accurate valuation model.
For each company, I collected revenues and EBITDA for 2011 and 2012 and current values for market cap, cash and debt. I used two free services to get these numbers. I used Google Finance to get market caps and current cash and debt levels. Here is the Google Finance page for RightNow, for example. To get revenues and EBITDA, I used a combination of Google Finance and Yahoo Analyst Estimates. Here is the Yahoo Analyst Estimates page for Right Now.
I then put all the numbers down and used formulas in the spreadsheet to calculate enterprise value which is market value minus cash plus debt. This normalizes the market caps for companies with high cash levels or high debt levels. Then I divided enterprise value by revenues to get to enterprise value/revenue multiples for 2011 and 2012. And I divided enterprise value by EBITDA to get enterprise value/EBITDA multiples for 2011 and 2012. Please take a look at the spreadsheet to see how all of this was done.
The results of this exercise are as follows:
SAAS Price/Rev 2011: 4.8x
SAAS Price/Rev 2012: 3.9x
SAAS Price/EBITDA 2011: 66x
SAAS Price/EBITDA 2012: 31x
The EBITDA multiples are based on a smaller sample size (the 2011 sample size is one!) so they should be understood in that context. It is also true that most of the companies in this sample seem to be investing heavily in sales and marketing to grow revenues at the expense of profits and the public markets seem to be accepting of that approach (given the valuations they are carrying).
The price/revenue multiples seem about right given my cursory understanding of the SAAS world. If you have a SAAS business, then your company's valuation should roughly be 5x this year's revenues and 4x next year's revenues. These are for public companies. Investors will typically take a 20-25% discount for private company valuations because private company investments are not liquid. So maybe 4x this year's revenues and 3x next year's revenues is an appropriate multiple for a privately held SAAS business.
We led a follow-on in one of our most mature portfolio companies last month. It is a SAAS business and these multiples are almost exactly what we paid. That makes me feel good. It means we got a fair deal and so did the company. And that is why this kind of exercise is valuable.
When you published last week’s MBA Monday article I sat down and began working up a valuation for my own start up. I make t shirts, but you cannot use the valuations of a t shirt manufacturer because I also add a value through screen printing. But you cannot use the valuations of Threadless and or Cafepress, because they do not maintain inventory of blank shirts. Then you also have the fact that a community is being created that is unique, because membership is based on the consumers physical size.While I have three offers to buy the company if I hit my five year projections, I still am struggling with the fact that on one hand I am dealing with people who manufacture and think social media is just a fad and on the other hand you have people who do not comprehend manufacturing and want to avoid that.Its been quite an educational experience explaining social media and communities to one group that thinks its all a fad while also trying to get the social media and communities group to understand that a product can be the basis of a community.I hope that by focusing on valuations and comparing them that it might be easier to understand…if not I need to find a job as a lighthouse keeper.
earnings is one way that these people will take notice
Fred, I have made presentations to a variety of different funding sources and I realize that in every situation they attempt to grasp the business model from their own perspective and experience. Most angel investors I have talked to are more comfortable having someone else make their decisions for them, such as a local VC firm. Most VC firms in my area are focused on healthcare due to the local concentration. Most business leaders that I have talked to have analyzed the presentation from the perspective of their own industry and most of the people in my own industry are looking at units not margins.
Developing your own metric for valuation can be powerful.If it is reasonable and fair, it sticks.If it’s perceived as one-sided (a la GroupOn’s ACSOI that ignored marketing costs), it may not.
Aaron, one of my first blog posts was about GroupOn and how the valuations were absurd, and I didn’t have a clue about the math.I took three ex partners to a company last week that I tried to get them to invest in in 2000 and the CEO was kind enough to open his books to us (actually he loved the opportunity to rub his success in the face of my two ex partners) and they still cannot grasp the number of units sold, gross revenue, and net margins of this company! So, they spent the rest of the day explaining to me exactly why this company will not be successful; hmm, sales of 40 million a year, increasing by 20% a year, with NET margins of 34%….I believe they are way over the hump!They are netting more now than what we did on sales of over 100 million! Looks to me like the internet, B2C, and individualization/consumer centric is the way to go!But then every generation has their dinosaurs…..
Could you elaborate on why you chose Enterprise Value rather than Market Cap? I understand the rational behind including debt but theoretically the market has already priced the company’s debt into the Market Cap making the use of EV seem like double counting for this analysis.Also, how often are valuations based on two years of expected revenue?
the valuations are based on one year of revenue, either this year (2011) or next year (2012)enterprise value is the value of the business as if it had no cash and no debt. most companies have both so you have to back them out from market to get to enterprise value.think of it this way. two identical companies both have enterprise values of $100mm. one has $400mm in cash. the other has no cash. the one with all the cash will trade for $500mm and the one with no cash will trade for $100mm
Sorry, I meant should you be looking two years out (2012) or five-ten years out (2015-2020) for the valuation?I understand EV but don’t understand why you would use it for a valuation.
My guess: An early stage investment is interested in growth and is to ‘buy’ a fraction of the part of a company that looks like it will grow. The ‘enterprise value’ (EV) is a stab at the value of the part of the company that can grow.The cash and debt won’t change much over time so are just carried as ‘constants’ and not part of what will grow and not what the investment is for. E.g., might pay a big multiple for the company but certainly will not pay a big multiple for the cash the company has.Or I’ll start a company and capitalize it with $1. If you want to pay big multiples, then maybe you will pay me $2 for the company. SOLD! Then I will start two more such companies …!Or, what’s of interest is the founder, the team, the code base to date, the ‘attractiveness’ of the software at present, the current ‘traction’ and its rate of growth (e.g., ComScore numbers), and the general promise of a ‘home run’. The investment is essentially to buy a piece of that part of company. That there is some cash or debt on the balance sheet is important in various respects but has next to nothing to do with the real promise of the company. That is, putting another $5 million in cash on the balance sheet from, say, the company winning the lottery, doesn’t get more code written or get traction faster!
By definition, comparing equity capitalization market cap to EBITDA would require applying interest and tax effects there of the denominator cash flow. More fundamentally, EV / EBITDA allows several parties to discuss valuation of the business despite any differences of opinion about capital structure.
This makes sense for valuations of mature (or at least revenue generating) companies, but what about your earlier stage investment when there are no real revenue generating comps? How do you determine value something like Twitter in its earlier years?
it is a negotiation between the entrepreneur and the investor. there is no science to that
I’m still grappling with how to apply this to an emerging start-up situation. Especially when you transition from being valued based on the future potential/prospects to actual revenue multiple, i.e. when your revenues are small but starting to pick-up but a pure revenue X doesn’t make sense. Example: company did a Series A at 8M with no revenues a year ago, then is now tracking $100K rev/month. At 1.2mil/year and even at X5 that’s $6 million valuation which doesn’t make sense.
that’s the problem with revenues. they can reduce your valuation. going from hopes and dreams to a real business often exposes the difference between what someone thought the business was worth and what it is worth. best thing to do is get the existing investors to fund you through this phase and go back to market once your valuation based on metrics is well north of the last round price
Good point of course if you can time it that way. But wouldn’t a smart VC/investor see the blend of potential + revenue ramp-up as a proof that it’s working. What I’m thinking of is that there would be a premium on the multiple equivalent to a growth rate, e.g. if it’s 100% growth, then double that valuation multiple or something like that?
I’ve come to believe that (a) that definitely exists and (b) it’s less about math and more about belief, so there is no rule of thumb. It’s about you and several potential investors sharing that belief enough for them to write a check based on it.
Exactly. A strict revenue multiple at an early stage doesn’t make sense. I’m not sure what the right starting number is, but probably $4-5 mil/year should be the minimum for when revenue multiples should kick in.
I agree that revenue multiples at an early stage are not a good guidelines for valuation.But…if you have a product and a price list and a few first customers, you do start to define your market size by realistic expectations of what you ‘should’ be making over time.Still difficult even from that perspective. It’s a delicate dance.
“that’s the problem with revenues. they can reduce your valuation”I absolutely love that quote.
i should have said “temporarily”
No, its true. Once you have a history, that’s how you’re evaluated.Its what concerns me about all the startup funding lately. We both agree more money for startups is good.Where its gets scary is like many financial cycles you get the innovators, the imitators, and then the idiots (somebody who puts up their money with no due diligence, defines an idiot).As Mark Suster points out, if you need follow on investment to stay alive, you better have somebody like you that is going to put in more money. When you have to go out and it looks like its going to be a down round, people don’t want that Brain Damage.You always have to remember you have to grow into your valuation, Its kind of similar to buying too much office space if you err to big its a really big problem that’s hard to fix. If you err to small its also a problem but an easily solvable one.
Fred’s multiple ranges are still useful for your example, you just are early enough you need to add in one more layer. Where are you going and what is your probability of getting there? If the math of where you are going and your odds of getting there mean that the Series A numbers don’t make sense, it means someone either overpaid or under-delivered and the next round would likely be a down round.
When there is real high growth that’s happening, revenue is such a moving target that valuations should carry a premium on that multiple, just like stocks are valued on a premium above average multiples based on growth %.
Exactly Fred’s point.I think that’s why some of us that have done this several times, look at valuations of $8M with no revenue, a business plan, and a dog, and shake our heads.
These multiples are derived from US companies and most are focused / active on relatively mature US market. If you were investing a SaaS company in Turkey ( i.e. highly dynamic, higher growth expectations, better EBITDA margins) what would be the multiples?
As the world gets flatter, the differences should get smaller and smaller.The point of the multiple is to get a baseline. You can adjust the multiple up or down based on margins, revenue expectations, etc.
they would probably be lower but the best way to know would be to do this kind of analysis on turkish companies
My former Corporate Finance and Valuations professor posts data sets every year with most of this data (even multiples) compiled. Useful for folks to double-check your work (note: EV/Sales = EV/Revenue). http://pages.stern.nyu.edu/… (go to the second table, and download the XLS for the country/region of interest in the Current Year)@a21647211cb09c6814bafa4f6aea7df8:disqus , you should be able to find some comparisons in the Europe / Emerging Markets data sets. But, less developed financial markets (particularly those with limited regulation) will yield less reliable information so take any findings with a grain of salt.
That’s awesome! What a great resource.
wow. that is awesome
That is beyond cool. Thank you very much for that.
Great post, Fred.Too many revenue-less Series A investments are ignoring these numbers and forgetting that eventually one of their financing rounds will need to be valued this way.I look at it this way: a good check on the reality of a Series A valuation is to increase it to what you think your Series B or C valuation might be. Then calculate the revenues backward from there.Are you setting yourself up for a down round, or can you build that kind of company with what you’re raising?That is the math that can lead you to want a more reasonable Series A valuation as an entrepreneur.
that is one of the reasons i am doing these posts. to make that very point
Aaron, Mike, Fred,Thanks – this is extremely helpful as I start down the road of being a founder. We’re working through our valuation now and our advisors have made similar points. Having the spreadsheets and the additional resources from these comments has been huge addition to their suggestions.Sujal
Great post Fred, people definitely need to understand the basics of how multiples are used. I wrote a short answer on Quora ( http://qr.ae/7n5au ) addressing the question: How do you determine an appropriate EBITDA exit multiple for a given company?”The answer to your questions is all in the word “comparable”.You tend to need a multiple of some type when a new set of owners are looking to take a growing company into it’s next stage of growth. The same key areas that an earlier stage investor evaluates are relevant here. The market, the team, the product, the customer base, the revenue streams, potential catalysts, competitors, regulatory risk, etc. The unique combination of these is what determines what people are willing to pay for a company.What typically happens is that a partner-level person tells an associate/intern to “pull some comps”. They log in to CapitalIQ and search for transactions on other businesses in the same space. They try to figure out what was paid and what the underlying financials indicate are the multiples that are being paid. They will often come up with a range (2.5-4x or 5.5-7x for example). Then the REAL question is “That’s great they got a 4x multiple on their transaction, but how similar are these companies really?” People who know how to do serious due diligence and network well excel here. Interns on CapitalIQ might get you in the ballpark but it is talking to suppliers, customers, former employees, that helps you uncover real value / avoid overpaying and creates the foundation that allows the investment the best chance to be profitable down the road.Quick Caveat: EBITDA is not always the right multiple. Depending on company stage/profitability/industry, other multiples may be much better ways to estimate the value of a company.”
“It means we got a fair deal and so did the company”I like that statement in the end, given that there is a bit of creative art that is at work in terms of valuation multiples despite all the use of numbers etc.I like that Mr.Wilson states that USV got a fair deal , not a great deal. My dad always stated that in life there are no great deals at best you can hope for is a fair deal.Mr.Wilson and USV are content with a fair deal, so it seems. Wish more were like him.
Right…remember, a fair deal is when both sides slightly dislike the price, but like it enough to shake hands and get the deal done!
So HP’s bid for Autonomy at 10x revenue is good right 😉
What about for more baby companies?
it’s art at that point. certainly not science
Maybe I’m crazy, but this seems like a weird way to do valuations. Investors in the public markets are buying at those prices on the hope/assumption that the price will rise (at or above the market) or the dividends will earn them returns.For VC, isn’t the investment based on the eventual, potential exit price of the company (whether going public or being bought or having a private equity-style liquidity event). Thus, wouldn’t it make the most sense to judge by the expected outcome and whether that reaches the investors’ risk and return goals?I’m pretty familiar with this common wisdom of 4X for SaaS, but I don’t quite grasp the full logic behind it. It seems like reversing out the return potential and controlling for things like riskiness, growth rate, team, acquisition appetite in the sector, etc. would be a wiser way to determine if you’re getting a “fair deal” as an investor.
investors need to know how the markets will value your company when it eventually exitsthat is why you need to do this exercise
Rand- these models are most appropriate in the context of buyout deals- i.e., companies with mature operations. Venture deals would be valued differently, ie. discounted cash flow, revenue/subscriber valuations, or some other relevant metric.
The X sales value has been around forever.Reason being is that you can ratchet all sorts of other numbers a bunch of ways. Are you marketing ahead of sales, are you spending money building infrastructure ahead of new sales, etc, etc, etc.All those are up for debate, will they work won’t they work? One thing that you cannot debate is what are your sales. That is what the actual market says your value is worth.Its hard because it works against you the older you get….just like many things. Yes you can hope for the fantasy buyer, but while tremendously celebrated like lotto winners they are very rare. Same as believing your company is worth what the post round financing is if is not a straight cash for common stock transaction.
This was a very terse and accurate illustration for conducting proper multiples analysis. In many of the deals we pursue there is typically a broker involved who attempts to feed us valuation guidance. They are almost always WRONG and they are somehow right, they have no mathemetical support for their valuation. One limitation of this model is that it doesn’t take into account extraneous factors affecting a business. For example if a business’ product as a consequence of it market position dominates its supply chain we may give that company a premium based on market control exclusivity. These kinds of factors sometimes merit paying a substantially higher multiple.
Like (lack of) liquidity risk getting discounted by 25%, wouldnt investors seek similar discounts for other risks like product, market, team, revenue etc.?
Sure. These factors provide good guidance on pricing but also on transaction structure as well. Maybe liquidity risk merits a 25% discount but in lieu of discounting the purchase price- we extend the terms of the earn-out.
For SaaS companies many public investors value the properties off of OCF or Operating Cash Flow. Some value off of DCF or Discounted Cash Flow and some off of FCF or free cash flow. For some SaaS companies examining revenue doesn’t provide a great view as many customer contracts are multi month or multi year with the revenue recognized ratably over the length of the contract. Since many customers pay up front of the contract term,a view into cash flow offers the growth dynamic of the company.
Shouldn’t you be paying a higher multiple for a company that expects more growth than a more mature company?
I didn’t see risk come in beyond public/private liquidity. I presume risk would have to be weighted in with projections if they are to be utilized. IE how confident are (you as the investor) in the projections based on past performance, company and market changes, etc.
I had the same thought, but noticed on rereading that he specified it was “one of our most mature portfolio companies”. A mature company getting one more round before going public is probably different than an early-stage company.I thought Aaron Klein’s post from above was spot on — you can at least make some guesses at what your Series B or C valuation might look like, and justify your Series A accordingly.
Wouldn’t it be better to approach the valuation as a multiple of stabilized EBITDA plus cash, minus funded debt? Cash can be paid out as a dividend or stripped by a new buyer if the sellers are foolish enough to leave it on the seller balance sheet. Debt must either be paid off by the seller, or assumed by the buyer…in either case that is certainly a valuation deduction, not an addition.Then there is the matter of contingent liabilities of the seller, which are in the end just like funded debt in that someone must be responsible for them, even though they don’t appear on the balance sheet under GAAP..
I am confused aboutthis method, despite having invested as an angel in YCharts.com, which providesdata public company comps. Where is thediscount for small company risk (i.e. public companies are usually less riskybecause they are larger, better capitalized, more diverse, and receive heavierregulatory scrutiny)? Does this methodassume that the early-stage, private company should get the increase invaluation multiples that come from going public? How do you correctfor the natural optimism built into everybody’s projections? This method assumes that one can adjust fordifferences in growth rates/opportunities simply by eyeballing multiples. Spreadsheets showing the growth, and thendiscounting the cash flow back, can get a metric taking into account varyinggrowth rates and margins. The problem Ihave had with this method is an appropriate, risk-adjusted discount rate is sohigh as to dominate other inputs. Well, the above isprobably fodder for multiple posts. Andadjusting projections is totally an art, and perhaps not even as definitive asthat. Love your blog. Have lurked on it (and reposted it) forseveral years.
Can anyone please discuss/explain valuations for real estate brokerage companies?
Fred, What about free cash flow? Revenue recognition and deferred revenue rules combined with differences in bookings/collections processing across comanies will mean revenue is very different from cash and possible very different across companies. While Ebitda is a proxy for FCF, it can get quite muddy as well so I recommend EV /FCF if possible. A few years ago as a banker I did an analysis for a Saas client that showed that EV to FCF was a better predictor of value for saas cos with a higher correlation coefficient than other multiples. I realize though that Vc investments can be pretty loosy goosy on the multiples though since comparing big public companies to small start ups is not fair either. (correlations breakdown)Perhaps a 3-5x rev guideline is good enough.
Fred,> It is also true that most of the companies in this sample seem to be investing heavily in sales and marketing to grow revenues at the expense of profits and the public markets seem to be accepting of that approach (given the valuations they are carrying).I think this phenomenon: “the companies in this sample seem to be investing heavily in sales and marketing to grow revenues at the expense of profits” – may be because they are SaaS companies. Dharmesh Shah (of Hubspot) said in a post on OnStartups.com, that B2B SaaS companies (including Hubspot) tend to have a high cost of customer acquisition up front (e.g. $1000 per customer – this is for B2B and assumes a marketing and sales team and its costs, etc.), while their revenues (unlike the case of traditional enterprise software, where a big license fee is paid at the start) come in over a period of time, the period for which the customer stays as a subscriber, say 4 years, at a monthly subscription fee of, say, $50, which works out to $600 per year and $2400 over the 4 years. So while they do make a profit, overall (assuming the cost of acquisition plus all other costs add up to something less than that $2400), in the first year they make a net loss on that customer, and so on. This may be why the public markets seem to be accepting of that approach, because they understand that that is the case.
How did you get the EBITDA values for 2011 and 2012 from google finance and Yahoo?
i used operating income as a proxy
Good article, though I would have liked to hear more about what premium is placed on growth for private companies, as we’re all aware of discounts for liquidity and size.I’m the CEO of a SaaS startup and keep a similar spreadsheet for the investment committee on our BOD, though we have 15 companies on the list and the ratio is a little higher due to high fliers Salesforce and OpenTable.We segment the 15 companies into low/med/high growth and I think that tells a very interesting story…Low growth (i.e. <20% past 12m, avg 14%) = 4xMed growth (i.e. 20-25% past 12m, avg 23%) = 5xHigh growth (i.e. >25% past 12m, avg 32%) = 7xThe above is as of close 9/30 and so includes the “bad quarter”.My analysis of the above is that you get 1x for each 4.5% of growth rate. We’re growing about 70% this year, so does that warrent the 15x multiple the above suggests? (i.e. 70 – 32 = 38, 38 / 4.5 = ~8, 8 + 7 = 15)