Friday, November 4, 2011

Following up on Groupon

In my last post, I made an attempt to value Groupon and came up with $14.62/share, before the voting right adjustment. Now that the offering is complete and the first day of trading is over, I thought it would be useful to take stock of general lessons that can be drawn from this deal. The offering price was raised to $20/share and the stock jumped another 40% during the course of the day. So, here are a few questions that I am fielding today...

1.     Is it possible that you are wrong in your assumptions and that other investors are far more optimistic about revenue growth/ operating margins?
There is no room for hubris in markets. Investors who do not admit to their mistakes, fix them and move on are doomed to pay a steep price. So, I will start with the presumption that I am wrong and the market is right and assess the likelihood, using two techniques.
  • Implied growth/margins: The two key inputs in this valuation are revenue growth and the target operating margin. In the table below, I changed both those numbers and assessed the impact on value per share. I also highlighted the combination of growth/margin assumptions that would get me to $28/share. For instance, if Groupon can maintain a revenue growth rate of 70% a year for the next 5 years (which will give them revenues of $64 billion in ten years) and a target operating margin of 23% a year, the value per share is $39.81. Notice that there is no scenario where the revenue growth is less than 60% a year for the next five that delivers a value greater than $28/share. (There are a few odd quirks in the table, where the value decreases as the margin increases, for a given revenue growth rate. That is because the NOLs that you accumulate spill over into the terminal value. Suffice to say that if your plan with Groupon is for them to lose so much money that they will never run out of NOLs, you should think again...)
(Revenues in year 10 in brackets next to five year growth rate)
  • Simulations: In the last few years, I have used an add-on to Excel called Crystal Ball to run simulations. In a simulation, you input distributions for key inputs where you feel uncertain about the future. In the Groupon valuation, I made revenue growth and operating margin into distributions – compounded 5-year revenue growth is uniformly distributed between 30% (pessimistic end) and 70% (optimistic limit) and the target operating margin is assumed to be normally distributed, with an average of 23% and a standard deviation of 4%.  I then ran 10,000 simulations, drawing from the distributions, and presents a distribution of values, shown below. Note that there is only a 15% chance that the value is greater than $28 (1500 outcomes out of the 10000 yielded values of $28 or higher). In fact, while the average value is around $14, there are far more outcomes under $10 than above... Put differently, you are not going to win on this stock most of the time, but if you do, you have to hope it is a big win.

2.     Is it a problem with the approach? Does DCF systematically undervalue young, growth companies?
     As many of you know, I am a staunch defender of discounted cash flow valuation, but here are a couple of criticisms I have heard of the model (especially in the context of valuing young, growth companies like Groupon) that I want to address.
a. DCF valuation is inherently conservative. It will under value growth companies.
An analyst that I was chatting with in the last day  was dismissive when I gave him the result of my Groupon valuation, and his claim was that  “DCF valuations always understate the value of young, growth companies”. But is that true? Those of you who have been reading my blog  and know that I valued Facebook, Skype and Linkedin earlier this summer (and found them all over valued) may very well conclude that I would find all social media companies to be over valued right now. That may be true, but it cannot be generalized to DCF as a technique. There is a bias that comes from the timing of these valuations: I chose these companies to value because they were in the news and were either going public or thinking about it. But when do companies in a sector think about going public or offering themselves for sale? It is when managers believe they will receive a favorable valuation. Put differently, I am valuing social media companies at a time when the market is most likely to be over valuing them. To provide some perspective, I valued dozens of dot com companies in the 1998, 1999 and 2000 and I found every one of them to be over valued. In 2001 and 2002, when I revisited the sector (or what was left of it), I found many of the same companies to be under valued. In the graph below, for instance, I have my value and the market price for Amazon.com each year from 2000 to 2003; notice how over valued it was in early 2000 and how under valued it looks in 2001. The bottom line: DCF is not inherently conservative, but done right, it is contrarian. You are likely to find stocks to be under valued when the market mood for a sector is darkest and stocks over valued when investors are enamored about a sector.

b. DCF valuation misses components of value
  1. There is some truth in this statement and I did cover one aspect when I talked about the “option” premium in some growth companies in my earlier post on the value of growth. What am I talking about? If I had valued Apple as a personal computer company in 2000, I would have missed its expansion into the entertainment business in the last decade. Similarly, if I had priced Google as a search engine in 2004, I would not have considered its expansion into other businesses in the last few years. If Groupon is successful in its core business, could it expand into other businesses? Sure, but there are two levels at which I would be skeptical in this case. First, I am not sure what Groupon competitive edge will be in these unspecified new businesses. Second, even when I have estimated a real option premium, I have never obtained an increase of more than 20-25% on my DCF value. 
  2. The other argument is that Groupon is issuing only 5% of its shares in the IPO and that the shares are therefore scarce: investors who want the stock therefore have to pay a scarcity premium. But shares in a company is not a Tiffany lamp or a Mickey Mantle baseball card. It is a claim on a set of cash flows and who generates these cash flows or how they are generated is not relevant. (As far as I can tell, a dollar you generate in cash flows from your Groupon investment buys exactly the same amount as a dollar in cash flows you generate from your Apple or Google investment.)
3.  If you  believe that the value is only $14 or $15 a share (or lower), how do you explain the $28/share price?
     My first response is that I feel no urge to explain the $28/share price, since I did not pay it. My second is that this is a snarky response and that I should be able to put myself in the shoes of those who did buy the stock today and explain why.  I could take the generous view and attribute their actions to  more optimistic assumptions about growth/profitability (and a higher value). I think that there is a simpler, more likely explanation. I would wager than most of the investors buying Groupon stock today have absolutely no idea what its value is and could not care less. They are playing a very different game than I am. With a time horizon measured in minutes, hours and days, they are buying the stock today, hoping to flip it to someone else at a higher price next week. Will some of them make money? Sure, and I don't begrudge them their profits. It is just a game I am good at and I won't even try!

   4.  If you believe that the intrinsic value is only $10-$15, should you short the stock?
In the Google shared spreadsheet that I put up in my last post, I notice that many of you, who found the stock to be over valued, plan to sell short the stock. You are far braver than I am! I did not sell short on any of the dot com companies that I found over valued in 1999 and early 2000. In hindsight, could I have made money by doing so? Eventually, yes, but eventually would have been a long time coming on some of those stocks... My problem with selling short has always been that I don’t control my time horizon; the person who has lent me the shares does. After all, even if you are right in your assessments of value, you will not make money until the market corrects its “mistakes” and that may take weeks, months or even years. With hot sectors, where prices are based on perceptions and the herd is “optimistic”, I think it is far more prudent to get out of the way and let the momentum investors have their day in the sun. My day will come!!!!

16 comments:

Unknown said...

Have you used the same model to valuate high flyers like fio, bidu, crm on their IPOs? Can you name a few overvalued IPOs that are still flying hi after X years as public companines?

Unknown said...

Have you used the same model to valuate high flyers like fio, bidu, crm on their IPOs? Can you name a few overvalued IPOs that are still flying hi after X years as public companines?

Aswath Damodaran said...

Isn't that a loaded question? If you premise the question by calling them "overvalued', then the answer will be obvious. The real question is whether we can tell the "overvalued" IPOs from the "high but fairly valued" ones.

Jiten Chadha said...
This comment has been removed by the author.
David Waltz said...

I find that the process covered in #1 is helpful because it provides the scenario under which you may be convinced to adjust the original valuation. The question then becomes re-framed: am I willing to invest on the basis there will be greater than 60% growth over the next five years?

This can lead to other analysis projects - looking at comparable historical situations, contagion models, etc.

@Gert_S said...

Thank you very much for the staunch defense of DCF with its broad-brush simplicity and robust logic. The scenario analysis above is particularly useful along the lines of Rappaport's Expectation Investing.

In this scenario analysis, however, you only sketch the difference of opinions and expectations regarding opportunity. Could it not be that a major difference of opinion exists regarding risk: early investors may have a lopsided opinion of the operational, strategic and financial risks? And what about the perceived duration of "unfair competitive advantage" (Pip Coburn)?

Such unbalanced opinions may not be straightened by short-selling: effective opinions and expectations in the capital markets may not be perfectly balanced when believers congregate in a small space such as an IPO. I think you describe that aspect very well: "My problem with selling short has always been that I don’t control my time horizon; the person who has lent me the shares does."

Loganthan said...

I wonder how many companies there are globally that after at least 10 years in existence have an operating margin of 23% or more and revenues in excess of $ 25 billion. It would also be interesting to know the compounded revenue growth of these companies over their best 10 year period. As one can guess this would give an insight into the companies there are which meet the above criteria and enable a comparison with Groupon to see if Groupon has any characteristics comparable to these companies.

Aswath Damodaran said...

Gert,
I did consider also making the initial cost of capital a distribution, but chose not to do so.. But it could very easily have been done..
Loganathan,
Good point. I may do something on these lines next year... when I update my global data.

Fff said...

Here is our SPX EOD WEEKLY: our WEEKLY TIMING MODEL went to a SELL last friday 04112011.
http://followmarketrend.blogspot.com/2011/11/spx-weekly-ind-weekly-timing-model-went.html

Anonymous said...

Aswath,
As I read this article it appears to me that the analysis takes into account only one side of the 2-sided market Groupon operaties in - the end consumers buying coupon. But how long will businesses keep giving away theie profits in the form of 50% off discount and another 25% to Groupon? Isn't that value destruction? Groupon's valuation should also be based on its share of net new value created, which isn't the case here as we only see value redistribution.

Regards
-Rags Srinivasan

Aswath Damodaran said...

It depends on your definition of value creation. If you have incurred the cost (setting up the dental office, buying the equipment), then the marginal cost of the added customer is close to zero or very low. Selling the product at a low price is still value added (not destroyed). That is why I framed it in terms of excess capacity.

Pranav Pratap Singh said...

Another great post! I got a ceiling for overvaluation over DCF for small high growth companies (20%), at my risk obviously. 9 years back, while doing MBA, I had read your paper on real options and still believe that it is the right method of valuation. If we could account for inputs that affect the value of real options like management quality it might make it more usable. But, would that mean modifying the assumption of geometric Brownian motion. Uff.. why did not I take the course on OFD in MBA!

Pranav Pratap Singh said...
This comment has been removed by the author.
Pranav Pratap Singh said...

looks like the stone is falling back to ground!
Groupon short position almost 3m shares: Nasdaq
http://www.moneycontrol.com/news/world-news/groupon-short-position-almost-3m-shares-nasdaq_625088.html

Angad Guglani said...

Making any sort of play on groupoun stock might be quite hazardous due the the few number of shares floating around. Market manipulation could be going on. Additionally the very high beta of the stock could hurt shorts the general market starts rallying on Europe news again.

anant krish said...

When is your next post coming up?