Facebook: Playing the "IPO pop" game?

In my last post on Facebook, I provided my estimate of value (about $ 70 billion) and concluded that I would not be a buyer of Facebook shares even if the company was valued at close to $ 70 billion. A few of you have taken me to task for leaving what you see as easy profits on the table, noting that if I were able to get Facebook shares at the offering price, that I would be guaranteed (or close to guaranteed) a substantial profit. More generally, there is the perception that investing in IPOs at the offering price (and making money on the offering day pop) is a low-risk, high return strategy that can be used to augment your portfolio returns. Like most "sure" things in investing, this one comes with implicit assumptions and costs and is definitely not "sure". Here is my list of caveats for those considering the "Facebook IPO Pop" strategy.

1. You have to be able to get the shares at the offering price
Let's say that when the Facebook IPO does get "priced", you bid to buy shares at the offering price. One of two things will happen: you will either get your requested number of shares or you will not, and ironically, it is the former that should worry you. If you are right in your basic hypothesis, i.e., that IPOs are under priced, the demand for the shares at the offering price will exceed the supply and the investment bankers managing the IPO will have to ration the shares. If the process is fair, you should get only a proportion of the shares you asked for, with the proportion getting smaller as the shares get more under priced. If the process is fixed and the investment banks give first dibs to their preferred customers, the proportion you get, if you are not one of the preferred customers, will be even lower. Thus, the only scenario you should dread if you are not a preferred customer is one where you get your entire allotment filled, because that is an indication that the shares are over priced.

This allotment process has always been the achilles heel of strategies that try to invest across all IPO offerings. You may bid for $100,000 worth of shares of every IPO for the next year, but you will end up with a portfolio that has too little invested in the most under priced IPOs (since you will get far fewer shares than you requested in these) and too much in the most over priced IPOs (since you will get all of the shares you asked for with these).

2. The stock has to pop on the offering date
Once you have been allotted the shares at the offering price, you have to hope for a stock price pop on the offering date. You do have history on your side. Looking across all IPOs, there is evidence of an offering day increase in stock prices. The figure below graphs out the average "under pricing" across all IPOs, by year, for US stocks:

Note, though, that this return presupposes that you can invest an equal amount in each IPO, under priced or not, but it is still impressive. Over the entire 50 year time period, there have been only four years (1962, 1973-1975) where the average returns were negative, and there are periods, such as the late 1990s, where the average return is close to 100% (doubling of price on the offering date). That is good news for the "Facebook IPO Pop" strategy.

Before you get too excited, though, note that the under pricing is greatest with the smallest offerings, as evidenced in the graph below:

Given the size of the Facebook offering (even 5% of $100 billion makes this a big offering), this graph should lead you to lower your expectations of the price pop on the offering date.

There is also evidence that this under pricing is by design. The IPO pricing cookbook at most investment banks includes a "take 15% off the value" ingredient in every pricing recipe, since the positive news stories that accompany the pop are viewed as a sales pitch for future stock issues. The under pricing also is consistent with the incentives for investment banks, who generally guarantee the offering price to the issuers, and thus have fare more to lose by over pricing than from under pricing.

This IPO under pricing has been the source of angst for some, who feel that the under pricing is unfair to the founders/owners of the company going public, since they are leaving money on the table, by offering their shares at a discounted price. That argument, though, becomes less persuasive, when you recognize that only a small portion of the outstanding shares is generally offered on the offering date. The discounted price on these shares operates the same way a loss leader operates in a retail store: it is designed to whet your appetite and get you to buy more. You hope that those who buy these shares (and feel good about the profits they make) will be back in six months or a year to buy more shares in the company. So, if you are feeling sorry for poor Mark Zuckerberg on the offering date, don't worry! There are plenty more shares in Facebook that will be hitting the market in future years.

From an investor's perspective, what can go wrong? Note that the average is across all IPOs (and is skewed by the smallest IPOs) and that a subset of IPOs see a drop in price on the offering date. These are of course the IPOs where you got all of the shares you asked for. If you are not a preferred customer, your odds of making money on IPOs decrease. Even the preferred customers are offered a mixed bag. Not only does their preference stem from the large amounts that they pay the investment banks for other services rendered, but they are expected to be "good" investors. In other words, if they flip the stock soon after the offering, it may endanger their preferred status on future IPOs. More on that in the next section!

3. You have to time your exit well
Assume now that the first two pieces of the puzzle have fallen into place. You have been allotted shares in the Facebook IPO and the stock has popped 15% on the offering date. Should you sell now or should you wait? That eternal  question has particular resonance with IPOs, because the gains on the offering date can be fleeting. Remember that Groupon's 20% jump on the offering date is now all gone!

Studies that track the post-offering performance of IPOs suggest that they do are not good investments in the aftermath. In the figure below, we compare the returns in the first five years after an initial public offering, with the returns on non-IPO stocks.

The returns on IPOs lag the returns on other stocks in the market and do so much more in the first few years after the offering. Thus, if buying at the offering price requires you to hold the stock for a year or two (which may be required of you as a preferred customer), you may not be getting a great deal.

In fact, my valuation of Facebook is predicated on the assumption that you may want to hold Facebook for more than a day in your portfolio. If you do, once the buzz fades and the IPO paparazzi leave, the company will be judged increasingly by how it performs relative to expectations. If price and value are on a collision course, value always wins out.

4. You are playing a sector and momentum game, even if you don't want to...
If you bid for shares in the Facebook IPO offering, I do believe that the odds may be in your favor for winning the game, if you define winning as getting the shares at the offering price and flipping them very quickly after the IPO. Much as I am tempted to join you, I am afraid I will sit out that game and not because of any noble impulses (such as wanting to be a long term investor or not speculating).

The IPO game is a subset of the momentum game, on which I have posted before. It is a game that produces big winners but momentum always turns, and when it does, it creates big losers. To see the link, note that IPOs go through hot and cold phases, with years in which you have hundreds of IPOs and years in which you have a few dozen.

In addition, also note that IPOs in any given year tend to be concentrated in a few sectors and those sectors generally have momentum on their side (dot com stocks in the 1990s, social media companies today). Thus, the success of IPOs in a sector is closely tied to whether the sector maintains it "hot" status.

One worrisome aspect of IPOs is that they may operate as canaries in the coal mine in signaling momentum shifts; the loss of enthusiasm for dot com IPOs (and the withdrawals of some) coincided with the epic collapse of the sector that year. The momentum driving social media companies will end one day and it may very well be the day that Facebook goes public. "Unlikely", you say, and I agree, but the tough part of being a lemming is that you never know where the cliff is coming. Of course, you may be able to sense momentum change much better than I, in which case you should be able to play the game successfully.

Bottom line
A strategy of investing in IPOs at the offering price looks much better on paper than it works in practice. All of the academic studies that show the average underpricing are implicitly based upon the assumption that you can create an equally weighted portfolio of all IPOs, when in fact, a non-discriminating investor will end up  will be with too much invested in all of the worst IPOs.

The strategy can be made to work by an investor who uses analysis (valuation or information) to invest only in IPOs that are most likely to be under priced and follows through with timely selling to capture profits after the offering. Even for that investor, it is a supplementary strategy since there will extended periods where there are no or very few IPOs in the market.

So, if you are bidding for those Facebook shares, good luck. I hope you get only a fraction of the shares you ask for (see part 1 for why), that the stock price pops on the offering date, that you get out before the you-know-what hits the fan. and that you are not the unfortunate soul who helps ring in the end of the social media circus.

The IPO of the decade? My valuation of Facebook

The Facebook IPO gets closer and I don’t think I can put off this valuation much longer. While we don’t have an offering price yet, the preliminary estimates are that the company will be valued somewhere between $75 billion and $100 billion. As with my Skype, Linkedin and Groupon valuations, I will present my assumptions and valuation of Facebook, with the admission that I have no crystal ball and know that your estimates will be very different from mine. So, with that disclaimer out of the way, here are my valuation assumptions for Facebook.

1. Where Facebook stands right now: I started with the Facebook S-1 filing which contains their financials from last year. The pdf version is available here, with my highlights and annotations (just ignore my snarky comments... I cannot help them). Looking at the most recent year's numbers, here is what I see:
(a) Revenues in 2011 were $3,711 million, up  88% from revenues of $1,974 million in 2010, which, in turn, were up  150% from revenue of $777 million in 2009.
(b) The firm's pre-tax operating income increased from $1,032 million in 2010 to $1,756 million in 2011. The firm's net income increased from $ 606 million in 2010 to $ 1 billion in 2011, though a third of that net income was set aside for participating securities (convertible preferred and restricted stock units... More on that later...). Incidentally, Facebook paid 41% of its taxable income as taxes in 2011.
(c) The company is primarily equity funded and its book value of equity at the end of 2011 was $5,228 million; the only debt on the books was $398 million in capital leases. They did have operating lease commitments, which when capitalized yielded a value of $776 million. The total debt is therefor $1,174 million.

2. Future revenues: Facebook is on a "high growth" path, with revenues growing by 150% in 2010 and another 88% in 2011, but as even that sample of two observations suggests, the big question is how that growth rate will hold up as the firm becomes larger. I estimate a compounded revenue growth rate of 40% for the next five years and a scaling down of that growth rate to the nominal growth rate in the economy (set equal to the risk free rate of 2.01%)  by the end of ten years. While both assumptions may strike you as conservative, I am effectively assuming that Facebook will follow a revenue growth path close to Google's over the next 8 years, as evidenced in the chart below, where I compare Google's actual revenues in the 8 years since their IPO with Facebook's forecasted revenues for the next 8 years:


Since advertising revenues are the drivers of both firms' growth engines, and they may very well be competing for the same advertising dollars, I think a comparison of their competitive advantages is in order. Facebook's primary advantage is that they can use what they know about their users (which is a lot... scary thought!) to offer focused advertising. Google's advantage is that it has a more direct and easy business model, since its revenues come from user clicks. In contrast, Facebook has to be careful about making its focused advertising too obvious, since some users will find this creepy. Google has added other products to its mix, with the Android as the most prominent example, and Facebook also has potential avenues for expansion.

3. Operating margin: Facebook has a phenomenal pre-tax operating profit margin in excess of 45%. To provide a contrast, Google's operating margin is currently about 31% and has seldom exceeded 35%. However, Facebook's margins will come under pressure as they actively seek out more revenues and I am assuming that the pre-tax margin will decrease to 35% over the next decade. Even with this assumption, I am estimating operating income for Facebook will exceed Google's by a wide margin over the 8 years following the IPO:


4. Reinvestment: In one of a series of posts on growth, I argued that growth does not come free (or even cheap). That is true for even a company with the pedigree of Facebook. There is some information in the financial statements about reinvestment: the company had net capital expenditures of $ 283 million, an acquisition that cost $24 million and an increase in capital leases of about $ 480 million. To estimate reinvestment in future years, I assumed that the firm would be able to generate about $1.5 million in revenues for every million in additional capital investment. At this stage, it is impossible to tell what form the reinvestment may take, but looking at Google over the last few years should provide clues; the company has moved increasingly to using acquisitions to augment growth. Lest you feel that I am being too conservative, I am estimating that Facebook will generate a return on its capital of about 32% in year 10, up from just over 26% now.

5. Risk and cost of capital: Facebook is a company that is funded almost entirely with equity and while it is a young, growth company, it does have a business model that is working and delivering substantial profits. While we can start from the bottom and work up to a cost of capital, using parameters estimated for Facebook, I will employ a far simpler approach. Looking across the costs of capital of all US companies at the start of 2012 (you can find this on my website), I estimate a cost of capital of 11.42% for advertising companies. I will assume that Facebook will face a similar cost of capital to start. The median cost of capital for US companies is roughly 8% and as Facebook grows and matures, I do adjust the cost of capital down to 8%.

6. Cash and Debt: The assumptions above are sufficient to estimate the value of the operating assets. Discounting the cash flows back at the cost of capital (with changes over time) results in a value of $71,240 million. To get to equity value, I subtract out the outstanding debt ($1,174 million) and add the current cash balance ($1,512 million). While I would normally augment the cash balance with any cash proceeds from the IPO, Facebook is open about the fact (See S1, page 7) that the proceeds will be going to Mark Zuckerberg to cover tax expenses from option exercise and will not be coming to the firm.
Value of equity = $71,240 + $1,512 - $1,174 = $71,578 million
Based on my estimates, the values being bandied around ($75 billion- $ 100 billion) are not unreasonable. As with my Groupon valuation, I ran a simulation,making assumptions about distributions for my key assumptions (revenue growth, operating margin, cost of capital and reinvestment). The results are summarized below:

Note that the median value of $ 70 billion is close to the base case estimate (as it should) but there is a 10% chance that the value could be greater than $ 117 billion and a 10% chance of a value of $ 43 billion or less.

7. Value per share: At some stage in this IPO process, Facebook's investment bankers will have to arrive at a value per share (offered) and you and I will have to decide on whether to buy or not. That could be messy because Facebook has multiple claims on equity, starting with:
a. Equity options: There are 138.54 million options outstanding, from earlier year compensation schemes, with an average maturity of about 2 years and an exercise price of $0.75. My estimate of the value of these options collectively, net of the tax benefits that I see Facebook getting from the exercise, is $3,782 million. I will net this value out against the equity value to get to a value in the shares:
Value in shares = $71,578 million - $3,782 million = $67,795 million
b. Restricted Stock Units: In the last few years, Facebook (like many other tech companies) has shifted to granting restricted stock units. These are regular shares but the holders who receive have to first stay long enough with the company (vest) to lay claim to them and often face restrictions on trading. The liquidity restrictions, in particular, should make these shares less valuable than regular shares. There are 380.719 millions class B shares, in restricted stock units, that will eventually become regular shares and I will add them to current shards outstanding.
c. Class A and Class B shares: After the IPO, there will be 117.097 million Class A shares (with one voting right per share) and 1758.902 million Class B shares (with ten voting rights per share). Other things remaining equal, the latter should trade at a premium on the former, though I don't think that the expected value of control in this company is significant.

If I take the equity value, net of the value of options, and divide by the total number of class A, class B and RSU shares outstanding, the value per share that I get is $29.05. Allowing for a slight discount (3-5%) on the non-voting shares, I would anticipate that the class A shares in the IPO will have a value of about $28 (assuming that my share count is right... I will wait to get a firmer update as we get closer to the offering, before I close in on a per share value). You can access the excel spreadsheet with the numbers by clicking here. If you don't like my inputs or assumptions, don't stew about them. Go in and change them and see what you get as the aggregate value of equity in Facebook. If you can post it in the Google spreadsheet that I have created for this purpose, even better... Let's see if we can get a consensus value for the company.

If you are investing in Facebook, give credit to the company for being upfront and honest about where the power rests in this company. On page 20 of the filing, you will find this "Mr. Zuckerberg has the ability to control the outcome of matters submitted to our stockholders for approval, including the election of directors and any merger, consolidation, or sale of all or substantially all of our assets. In addition,  Mr. Zuckerberg has the ability to control the management and affairs of our company as a result of his position as our CEO  and his ability to control the election of our directors. Additionally, in the event that Mr. Zuckerberg controls our company at  the time of his death, control may be transferred to a person or entity that he designates as his successor." A little later on page 31, you will find this "We have elected to take advantage of the “controlled company” exemption to the corporate governance rules for publicly listed companies. Because we qualify as a “controlled company” under the corporate governance rules for publicly-listed companies, we are not required to have a majority of our board of directors be independent, nor are we required to have a compensation committee or an independent nominating function." Let's be clear about this: this is Mark Zuckerberg's company and you and I are just providing him with capital.

For those of you who are familiar with my valuations of Linkedin and Groupon, you will note that I am more positive about Facebook than those companies. Part of that can be attributed to Facebook being further along in developing a business model that works and delivers profits. Another reason, though, is that Facebook has a real chance at being the next “winner take all” company. What am I talking about? In conventional businesses, a company that gets a large portion of the market is subject to competitive assaults that cap the market share and reduce profitability over time. In some parts of the technology business, controlling a large share of a market seems to give the winner the capacity to take over the whole market. Consider three big winners from the last 30 years. Microsoft started off in the “office suites’ competing with many players in the word processing, spreadsheet and presentation program businesses, but at some point, its dominance drove the competition out. To a lesser extent, Amazon’s dominance of online retailing and Google’s ownership of online advertising (so far) reflect similar “winner take all” phenomena. I am not suggesting that Facebook has a lock on social media advertising, but it has a chance to get a big chunk of it, and if it does, the value that I estimated will be too low. Note that the simulation does yield values of $120 billion or higher for the company, if the stars align.

Would I buy Facebook stock, if its equity were valued at $75 billion? No, and not because I believe that the price is outlandish, but for two other reasons.

  • The first is that the price reflects the expectation that Facebook will become a phenomenal success. Anything less than superlative will be viewed as a failure.
  • The second is that what Facebook is brazen about the fact that they don't see any need for input from stockholders. In effect, they want my money but don't want me to have any say in how the company is run. This does not jell with the notion that stockholders are part owners of the companies that they owned stock in. You may be comfortable with Zuckerberg as CEO for life but I am not. I am sure that I am in the minority on this one, but different strokes for different folks....
In closing, Facebook has immense promise as a company and it is being priced on the premise that the promise will be delivered. Could it be worth $ 100 billion? Sure, but you are fighting the odds as an investor. Social media companies today collectively and Facebook in particular resemble stores with tremendous foot traffic (850 million users in the case of Facebook) but with nothing on the shelves. You are buying access to the foot traffic and hoping that you can get something on the shelves that they will stop and look at and buy. Given that social media is still in its infancy, we really don't know whether this promise will pan out, and that remains the basis for the uncertainty, and why short cuts that are based on value per member (a metric that I see with social media companies all the time) are fraught with danger.

    Options and Taxes: Is a "Facebook" tax next?

    Facebook is in the news and I will do my usual pre-IPO valuation and posting in a few days on the company but I want to focus in on an interesting story in this morning's New York Times about option exercise and taxation (at both the individual and corporate levels).

    The story itself focuses on two tax issues. The first is that Mark Zuckerberg is planning to exercise about $ 5 billion of options ahead of the offering, resulting in a tax bill of roughly $ 2 billion for him, about $1.5 billion in federal taxes and $ 500 million in California taxes.  The second is that Facebook will be claiming a tax deduction of roughly the same value, which will shelter them from taxes this year and allow them to claim tax refunds of about $ 500 million from prior years. All of this has some in Congress in full "indignation" mode, with Senator Carl Levin saying "“When profitable corporations can use the stock option tax deduction to pay zero corporate income taxes for years on end, average taxpayers are forced to pick up the tax burden,” he said. “It isn’t right, and we can’t afford it.” Before we embark on another tax policy change predicated on a sample of one (Facebook), it is worth examining the broader question of how employee options get taxed, especially at the corporate level.

    At the moment, if you are a company that grants options to its employees, the accounting laws require you to value those options as options (rather than at exercise value) and expense them when you grant them (though you can amortize these expenses over a period of time). Thus, what you see reported as operating or net income for a company today is after employee options have been expensed. This, of course, is a sharp reversal of accounting policy prior to 2006, when firms had to show only the exercise value of the options at the time of the grant. Since at the time of the grant, employee options were usually at the money (strike price = stock price), this effectively meant that option grants had no effect on earnings when they were granted. However, if and when the options were exercised, companies were required to show the difference between the stock price and the strike price as an expense. 

    To illustrate the difference, assume that you grant 100 million options with a strike price of $10, when the stock price is also $ 10, in 2008. Let's also assume that the options get exercised 2 years later when the stock price is $ 40. With pre-2006 accounting, you would not have shown an option expense in 2008 but you would have shown an expense of $ 3 billion [$40 - $10) (100)] in 2010. In the post-2006 time period, the company would have had to show an option expense in 2008, with the expense computed by valuing the options at the time. (For instance, an at-the-money option with five years to expiration on a stock with a price of $ 10 and a standard deviation of 40% would have a value of $3.36. Carrying this through, the company would have to record an expense of $336 million in 2008 and revisit this expense in subsequent years, as stock prices go up or down.  If you want to, you can try your hand at valuing options with the attached spreadsheet).
    [Update: I have been taken to task by the accountants among my readers for being simplistic (and wrong) on the accounting rules, since they are far more complex than what I have described in this example. I confess to the crime but I feel no remorse. I think that I am being truer to the underlying accounting principle of matching expenses up to revenues than the current accounting rules claim to be. My point is that accounting has moved grudgingly to accept the fact that options have to be expensed when they are granted and not when they are exercised, though the accounting obsessions with smoothing and back filling finds their way into the rules. In fact, more on that in my next post]

    So, what does this have to do with today's story? While the accounting treatment of options changed in 2006, the tax treatment did not. In effect, the tax authorities still use the pre-2006 convention, not allowing companies to expense options when they are granted but only when they are exercised. This creates a disconnect between accounting earnings and tax earnings, which can make valuation more difficult. But is it a loophole? Seems like it, if you only consider Facebook, which will save a billion dollars in taxes because its options will be exercised at a time when its stock price is sky high. But let's add to this sample of one. Take a company like Cisco, which has granted hundreds of millions of options over the last decade. Since the stock has stagnated over the period, many of these options are now under water and will either end up un-exercised or exercised for far less than the value at the time of the grant. If Cisco had been able to deduct these options at the time they were granted (at option value), they would have saved hundreds of millions of dollars, which they may now will lose forever, if these options remain under water. In the aggregate, with the current tax treatment of options, the government collects less in taxes from Facebook and more in taxes from Cisco.  

    Do I think that the tax rules on options should be changed? Perhaps, but it is not because the tax law is unfair or because I think it creates a loophole. As I see it, here are the three choices:
    1. Continue with the existing policy of taxing options when they are exercised. The net effect is that the most successful companies (at least in terms of creating market value) will get bigger tax deductions from option expensing than the least successful companies.
    2. Change tax law to match accounting law and allow companies to expense options in the year that they grant them. It will smooth out tax collections, level the playing field across companies and create more consistency. But here is the follow up question that gives me a little pause:  Should employees who receive the options then have to show them as income in the year that they receive them? If you are being consistent, the answer is yes, but where will they come up with the cash to pay the taxes? After all, employee options are not liquid and the employee while wealthy (in terms of options) may be cash poor.
    3. If you follow Senator Levin's logic that this is a loophole, and you try to craft legislation, I am not sure where it leads you. Should we ban the expensing of options by companies? I would accept that, if you stop taxing employees who receive these options. (If that had been in place this year, Facebook would have to pay about a billion more in taxes but the government would be collecting two billions less in taxes from Zuckerberg...)
    As someone interested in valuation, I have wrestled with options and their effect on value for a while. I think that options are mishandled in many valuations, with flawed arguments (Options are non cash expenses...) and perilous short cuts (treasury stock and fully diluted value approaches) overwhelming common sense. In fact, I wrote a paper on the topic that you can download by clicking here