Friday, January 28, 2011

Tax policy

As some of you may be aware, I report average effective tax rates for US companies, by sector, at the start of every year. Yesterday, that data was picked up by the New York Times and has got plenty of publicity since.
Today, I have heard from both sides of the debate. from tax lobbyists  that feel that the low tax rates reported for some sectors do not reflect reality and also from those who believe that companies in the US don't pay their fair share.

Before I dive in, I want to be clear that the tax rates on my website were never intended for a tax policy debate. My interests are more mundane - computing cost of capital and value - and tax rates are raw material that I use to these numbers. Here is how I compute the averages. I compute the effective tax rate for a company by dividing its taxes paid by the taxable income; if the company is losing money and pays no taxes, I set its tax rate to zero.  For my purposes, I need an average tax rate for all companies in a sector, money making as well as losing, to compare valuation multiples and costs of capital across sectors. That is the number (a simple average of effective tax rates for all firms in a sector) that was reported on my site and picked up by others.

However, that average may not be an indicator of what a profitable firm in that sector pays as taxes, especially if there are  large numbers of money losing firms (as is the case with the biotechnology sector). To remedy this, I have decided to expand my tax rate table to include an additional statistic - an average tax rate for only money making firms. If you get a chance, you can see the data here and download it.
Note that these tax rates are much higher than the original averages. The average tax rate across companies that have taxable income is more than 29%, whereas the overall average across all firms (including the money losers) is just above 15%.

You can draw your own conclusions from the data, but here is my reading:
1. The average US company pays its "fair share" of taxes: I know that there will be many who disagree with me on this premise but the facts back me up. I computed the average effective tax rate for companies globally and here is what I got:
Global Region Number of firms Effective tax rate Taxes as % of Revenues
Australia, NZ and Canada 3834 26.65% 3.38%
Developed Europe 4818 28.49% 2.51%
Emerging Markets 17079 21.63% 3.15%
Japan 3584 38.30% 2.39%
United States 5472 29.35% 3.01%
World 34787 27.17% 2.82%
The average money-making firm in the US pays about 29% of its taxable income in taxes, which is higher than the tax rate paid by most European, Asian or Latin American companies. Only Japanese companies have higher effective tax rates.To be fair, the taxable income may be lower in the United States than in other countries because of sundry deductions, but here is a comparison that dispenses with this problem. The average US company pays 3% of revenues in taxes, roughly similar to what companies in other countries pay.
2. There is much higher variance in tax rates across companies in the US than in other countries: Here is where I think our excessively complicated tax code has an effect. There is much higher variance across the tax rates paid by companies in the US, as companies in some sectors are given tax deductions and credits and others are not. I will wager that US companies spend more on tax lawyers and consultants than companies elsewhere.
3. There is a much bigger difference between the effective and marginal tax rate in the US than in other countries.  Note that the average effective tax rate across companies is less than 30% whereas the marginal tax rate in the US is close to 40% (with state and local taxes). The difference is far smaller in countries with simpler tax codes. In Japan, for instance, the marginal tax rate is 41% but the average effective tax rate is 38%.

Having laid that data question to rest, here are my thoughts on the tax policy questions, for what they are worth.
1. Investment decisions should be driven by economics and not the tax code:   The more complexities (and goodies) that we build into the tax code, the more we risk having investment decisions determined by tax law and not by economics.  I would rather have all companies pay a 25% tax rate on taxable income, with no special deductions and credits, than have an average tax rate of 25% with wide variations across investments and companies.
2. Borrowing decisions are driven by marginal tax rates: Decisions by firms on how much to borrow are driven by the marginal tax rate and having a high marginal tax rate will induce more borrowing across the board. If you use load the tax code with deductions and credits, you have to raise the marginal tax rate to compensate and with it, you raise the amount of debt that companies will carry. If we truly want to bring down financial leverage at US companies, we have to start by making the marginal tax rate lower.

Do I think that this latest move to simplify the tax code and lower tax rates will work? I am not hopeful and here is why. To keep that change revenue-neutral, we also have to start stripping the tax code of some of the deductions and credits. Unfortunately, that will make it a zero-sum game, at least in the short term, where some sectors will have to pay more in taxes while others will save. In the long term, I believe it will make the economy stronger, but who has the patience of the long run, when it comes to taxes?

Tuesday, January 25, 2011

Buybacks and Stock Prices: Good or bad news?

XYZ Inc., a publicly traded company, has the following characteristics:
a. 100 million shares trading at $ 10 a share. (Market cap = $1000)
b. A cash balance of $ 200 million, earning 2% a year annualized.
c. Total net income of $ 40 million (giving the company a PE ratio of 25 today).
XYZ Inc. uses its cash balance of $ 200 million to buy back shares. What will happen to the share price after the transaction?
a. It will go up
b. It will go down
c. It will remain unchanged
Classic corporate finance question, right? Let's see what the answer will be at the two limiting extremes: an extremely "lazy" market and a completely rational one.

a. Markets are really lazy: Here is how it goes. Assuming that you can buy the shares back at the current price (unrealistic, but you have to start somewhere), you will buy back 20 million shares with the $ 200 million, reducing the number of shares to 80 million. The loss of the income on the cash (2% of 200 million = $ 4 million) will reduce net income by $ 4 million to $ 36 million.
Earnings per share = $36/80 = $0.45
Applying today's PE ratio of 25 to this earnings per share:
Price per share = $0.45 * 25 = $11.25
Even if you iterate and reduce the number of shares bought back (by dividing the $ 200 million by $ 11.25), you will still end up with a hefty increase in the price per share. In fact, for the math to work out, this is all you need for a buyback to increase price per share:
a. Current E/P ratio > Riskfree rate (Current E/P ratio for this firm is =1/25 =4% > Riskfree rate of 2%)
b. PE ratio remains has to remain unchanged after the buyback.
You are implicitly making the assumption that the market was mispricing cash prior to the buyback and here is why:
Net income from cash (prior to buyback) = $ 4 million
Since you are assuming that the PE ratio of 25 applies to all income, the estimated value of cash is $ 100 million (25*4), half of the actual cash balance of $200 million.

b. Markets are rational: Is equity in the pre-buyback firm as safe as equity in the post-buyback firm? After all, the firm is eliminating not just 20% of its assets, but its safest asset. Presumably what is left behind in the firm will be riskier than before and you will therefore pay a lower multiple of earnings for the stock. As a limiting case, assume that the market was valuing the cash correctly before the transaction. The implicit PE ratio for cash is 50 (1/ riskfree rate=1/.02). The observed PE for the company is then a weighted average of the PE for risky operating assets and riskless cash, with the weights based on the income you make on each one:
PE for stock = 25 = PE for operating assets (36/40) + 50 (4/40)
Solving for the PE of the operating assets, we get:
PE for operating assets = (25-5)/.9 = 22.22
New price per share = 0.45 * 22.22 = $10.00
In a rational market,  where assets are fairly priced, a buyback by an all-equity funded firm will have no effect on the stock price because it is a value neutral transaction.

So, what will actually happen after the buyback? I think that either extreme is unlikely to hold, but looking at both allows us to crystallize the factors that will determine the effects of a stock buyback:
a. Market's valuation of cash: A buyback reduces the cash balance at the company by the amount of the buyback. The effect it will have on the stock price will depend upon whether the market was pricing cash as a neutral asset (a dollar in cash is valued at a dollar). If the market was "discounting" cash in the hands of a company (viewing it as likely to be wasted), a buyback will increase the stock price. If the market was attaching a premium to the cash (viewing it a strategic asset), a buyback will decrease the stock price.

b. Financial leverage: In the example above, the buyback had no effect on the firm's debt ratio because the firm had no debt and was using cash to fund the buyback. However, a firm that borrows money to fund a buyback will change its debt ratio, and consequently may change its cost of capital. In what direction? It depends on whether the company was under levered, correctly levered or over levered prior to the transaction. An under levered firm will lower its cost of capital with a buyback funded with debt and thus increase its value (and price per share). A correctly levered or over levered firm will increase its cost of capital by borrowing money (the higher cost of equity and debt from the additional borrowing will overwhelm the advantage of using debt) and see value (and stock price) go down.

c.Value of operating assets: The value that the market attaches to operating assets is a function of expectations about future cash flows. A buyback can alter this value assessment by changing market expectations: this is the signaling story for buybacks but it can cut both ways. In its positive form, a firm that buys back stock is signaling to the market that it's stock is cheap and that investors are under estimating the cash flow potential from operating assets.  In its negative form, a firm that buys back stock is telling the market that its growth opportunities are starting to dry up and that future cash flows may not have the growth that investors had presumed.

The net effect of all three of these variables will determine the stock price impact of stock buybacks. Using them to make overall judgments, here is what I would expect to see in response to a stock buyback:
  1. The most positive impact on stock prices should be at mature firms that have a history of earning poor returns on operating assets and are under levered. You get a triple whammy at these firms: the market probably is discounting cash at these firms because it does not trust the management, the firm is under levered and there is little likelihood of the buyback being viewed as a negative signal (since expectations for growth were low to begin with).
  2. The most negative impact on stock prices will be at high growth firms with a history of generating high returns on operating assets and little debt capacity. Cash at these firms is unlikely to be discounted (and may be even be viewed as a strategic asset), there is little potential for value gain from financial leverage and the buyback is more likely to be viewed as a negative signal about future growth potential.
In my earlier post on Apple, this is why I argued against a buyback. Apple meets two of the three criteria for the second group: superb returns on operating assets and perceptions that there is still growth potential. It is true that Apple has some debt capacity (though its effect on value is muted). The debate about whether and when Apple should use this debt capacity is a good one to have, but I think that the argument for using debt right now is weak. That will change, as the debt capacity continues to grow, and returns on operating assets weaken (as they inevitably will).

A stock buyback is not a magic bullet. If you are a firm that should not be doing buybacks, don't go down that road, even if every other firm in your sector is doing so. You may be able to fool the market initially and get a stock price pop, but the truth will eventually come out to hurt you (perhaps after the top managers have cashed out their options and moved on.. but that is another story...)

Stock Buybacks: What is happening and why?

S&P's most recent update indicates that US companies, after a pause for about a year after the banking crisis, are back in the buyback game. In the third quarter of 2010, the S&P 500 companies bought back almost $ 80 billion of stock, up 128% from the third quarter of 2009. Note that this is part of a long term shift away from dividends towards buybacks in the United States, as is evidenced by the figure below which reports total dividends and buybacks at US companies starting in 1988:

Note that aggregate dividends amounted to $ 100 billion in 1988 and aggregate stock buybacks were $ 50 billion in that year. (In fact, if you go back to the early part of the 1980s, buybacks were just a pittance... an unusual occurrence even at large, cash-rich companies). During the 1990s, buybacks increased dramatically and in 1999, cash returned in buybacks exceeded cash paid out in dividends, in the aggregate. The trend continued uninterrupted through 2008, with 2007 representing a high water mark for buybacks. The market collapse and economic fears that followed induced companies to hold back on buybacks in 2009 but it was clearly just a pause, not a stop, as the return to buybacks in 2010 indicates.

So, what has caused this  movement away from dividends in the last two decades? It cannot be that dividends are taxed more heavily than capital gains: Note that dividends have been taxed at much higher rates than capital gains going back to the early decades of the last century. In fact, in 1979, the highest marginal tax rate on dividends was 70%, while it was only 28% on capital gains. The changes in the tax laws in the last three decades have reduced the tax disadvantage of dividends - in fact, they have both been taxed at 15% since 2003 - and cannot therefore be a rationale for the surge in buybacks. It also cannot be attributed to companies thinking that their stock prices were too low, since these buyback surge occurred during the bull markets of the 1990s and 2004-2007, not during down markets.

There are several possible explanations.
a. Management compensation: The first is the shift towards using options in management compensation alters managerial incentives on the dividends vs. buybacks choice. When you pay a dividend, your stock will drop on the ex-dividend day whereas a buyback should not have the same effect (I will talk about the price effect of buybacks in my next post). As an investor, you may not care because you get a dividend to compensate for the price drop. As a manager with options, you do care, since your option value will decrease with the stock price.
Testable hypothesis: Companies that reward their managers with big option grants or tie compensation to price per share should buy back more stock than companies that have more traditional compensation packages (bonuses tied to profits, for instance).
b. Uncertainty about earnings: The second is that dividends are sticky: once you initiate or increase a dividend, you are expected to keep paying that dividend. Buybacks are flexible: companies can buy back stock in one year without creating expectations about future years. Companies that are uncertain about future earnings will therefore be more likely to buy back stock than pay dividends. It is my contention that deregulation (of telecommunications, airlines and a host of other businesses) in conjunction with globalization (and the concurrent loss of secure local markets) has resulted in less predictability in earnings across the board.
Testable hypothesis: Companies in sectors with more unstable earnings should return a larger portion of cash in stock buybacks (and less in dividends) than otherwise similar companies (in terms of cash flow) in sectors with more predictable earnings.
c. Changing investor profiles: A more debatable reason is that the expansion of hedge funds and private equity have changed investor profiles in the stock market. These investors tend to be focused on price appreciation (rather than dividends) and often are unwilling to wait for their cash.
Testable hypothesis: Companies that have activist institutional investors or are held by hedge funds/private equity holdings should be more likely to buy back stock.
d. The Dilution Delusion:   A stock buyback reduces the number of shares outstanding and generally increases earnings per share. Applying the current PE ratio to the higher earnings per share should result in a stock price. If this logic holds, stock buybacks are magic bullets that companies can use to push their stock price up. There is a fatal flaw in this reasoning, which I will examine in my next post.
Testable hypothesis: The more focused a company becomes on earnings per share, the greater the likelihood that the company will buy back stock.

I don't see any reason to believe that these shifts will be reversed in the near future. In fact, I will go further. I think that you will see companies across the globe shift to buybacks or at least to more flexible dividend policies (tied to earnings).  My next three posts will revolve around the buyback question.  In the first, I will take a closer look at the alluring (but untrue) argument that a buyback is always good for stock prices. In the second, I will look at the consequences of the shift towards buybacks for traditional investment rules and valuation models that have been built around conventional dividends. In the last one, I will argue that dividends should not be abandoned, but that it is time we rethink how companies set and change dividends.

Wednesday, January 19, 2011

How much cash is too much? Looking at Apple

In the midst of lots of news about Apple - Steve Jobs taking a leave of absence and a 78% surge in profits reported today - I saw this news story in the Wall Street Journal. The gist of the story is that a portfolio manager who has about $700 million in Apple's stock feels that it should pay out some or all of its $ 50 billion cash balance to investors. I do not know the portfolio manager mentioned in the article, Mr. Bonavico, and I hope that he was misquoted because what he is quoted as saying borders on corporate finance malpractice.

Here is what Mr. Bonavico is purported to have said:"...they (Apple) are leaving money on the table by having such a large cash balance well below their cost of capital. The cash is earning near zero." That is an absurd comparison. Apple's cost of capital is close to 9% but that is for its operating investments in software, hardware and its iTunes store. Cash is invested in near-riskless, liquid investments and the appropriate benchmark (or discount rate) to compare it to  is therefore what you would make on riskless, liquid investment. The three-month T.Bill rate currently is yielding 0.16% and that is all that cash has to make to break even (or to be a zero net present value investment). Cash is not a good or a bad investment. It is a neutral one.

Does that imply that all companies should be allowed to hold on to as much cash as they want to? Not at all. Clearly, some companies accumulate too much cash and their investors would be better off, if that cash were returned to them. The question of how much cash is too much cash is  a debate worth having.  To resolve this debate, though, you have to start off with a clear sense of how or why cash balances affect equity investors in a company. No investor in a company is ever hurt by cash being invested in low return, riskless assets (commercial paper, treasury bills). What investors should worry about is what the company may do with the cash: take bad investments or overpay with acquisitions. I would rather that the cash earn 0.16% in T.Bills than be invested in projects earning 6%, if the cost of capital for those projects is 9%. To make a judgment on whether to attach a "stupidity discount" to cash, investors should look at a company's track record. They should discount cash balances in the hands of companies that have a history of over reacting, poor investments and bad acquisitions. They should not discount cash balances in the hands of companies where managers are selective in their investments and have earned high returns (on both projects and for their investors). In fact, over the last decade, there have been several studies that have looked at how the market prices cash balances and the results support this proposition.

In the case of Apple, a company that has seen its market cap rise almost thirty-fold over the last decade while generating a return on invested capital that exceeds 30%, this debate to me is a no-brainer. Do you trust Apple's managers with your cash? In fact, the real question should be if you don't trust Apple's managers with your cash, what company would you trust with your cash? As an Apple investor (albeit with a lot less than $ 700 million invested) since 1999, I have no complaints and here are the three scenarios relating to cash that I can see unfolding:

a. Do no harm scenario: In this scenario, which is the one that Apple has practiced for the last decade, it invests, when it feels that it has a good product (iPod, iPhone, iPad etc.) to promote and holds on to cash when it does not. Continuing with this scenario does not hurt me, as long as they keep the hits coming. I don't get dividends, but who needs them when you get that price appreciation?
b. Dream scenario: In this one, Apple finds a way to invest its entire cash balance of $ 50 billion right now and manages to earn a 30% return on capital on this investment. While this would clearly jump start and increase value, it does not strike me as viable. A company, even one as good as Apple, just cannot create new products and investments out of thin air and then staff them successfully.
c. Nightmare scenario: In this one, Apple decides that the return on cash (less than 1%) is too low and decides to take operating investments or acquisitions that generate returns that are higher than 1% but lower than the cost of capital. This would be devastating for value. If Apple goes on an extended buying spree, acquiring companies at outrageous premiums, I would join Mr. Bonavico in demanding my cash back.
d. Listen to Mr. Bonavico scenario: In this one, Apple returns $ 50 billion in dividends immediately. As an investor, I will get a big check in the mail (for dividends) on which I will have to pay taxes, but the stock price will decline by roughly the dividend. In fact, there is a very real chance that a big payout could be viewed by the market as a negative signal of future prospects and that the stock price could drop by more. (If the alternative is a stock buyback, the same problems exist though they will manifest themselves in a different way)

I have an extended paper on the value of cash and cross holdings (another widely misunderstood asset) that you can access at:
Have a go at it!

Saturday, January 15, 2011

Herding behavior: Why, so what and what if?

A news story from yesterday's Wall Street Journal on hedge funds and their herding behavior provides a good starting point for this discussion. In summary, the article notes the following:
(1) Hedge funds seem to buy and sell the same stocks, at the same time, and track each other's investment strategies.
(2) The correlation across hedge funds has increased over time. Hedge fund managers copy each other more than they used to.
(3) Hedge funds collectively are under performing the S&P 500 by more and more each year; for 2010, hedge funds generated 10.4% in returns and the S&P 500 earned 15%. On this point, you can take issue, arguing that hedge funds (or at least some of them) are less risky than the market and that the hedge funds may not lag the market on a risk/return basis. However, there is no denying that even on that dimension, hedge fund performance has deteriorated over time. I am not surprised by any of these findings, since they are consistent with existing research.

Why is there herding behavior?
We see herding in all aspects of human behavior, not just in finance. We tend to wear what other people wear, eat where other people eat and congregate in places that attract the biggest crowds. Herding may be more researched in finance than in other areas, but it is not unique to finance. Here are some reasons why herding is common.
a. Evolution instinct: It is not hyperbole to note that we have survived as a species by following the crowd. A cave person, when confronted with a crowd of cave people running in the opposite direction, would have been well advise to turn tail and run with them. Odds are that they were being chased by a mammoth. That instinct is still deeply embedded in our psyches. Faced with a wave of panic selling or buying in markets, it is very difficult to not join in.
b. Safety in numbers: Remember when you were a child, doing something stupid with a group of your friends. Odds are that when queried about your behavior, you used the standard excuse, "that they were all doing it". Put in analytical terms, a portfolio manager or CFO who makes a mistake is more likely to escape the consequences, if others make the same mistake, but will be punished if he or she is wrong alone. Think of all those bank equity research analysts who had buy recommendations on Lehman in 2007, who are still equity research analysts...
c. Information:  Assume that you are in a city you have never been in before and that you are looking for a restaurant to eat dinner at. Watching where the locals eat does give you information; you want to avoid the empty restaurants, because they are empty for a reason.
d.  Absence of competitive edge: It is easier to stand alone, if you know something that others do not or have a unique skill that gives you a leg up on the competition. The hedge fund story is revealing. Note that the herding behavior has increased as the hedge fund business has grown and collective performance has suffered. Much as we like to attribute superior skills to hedge fund managers, the herding behavior suggests that the average hedge fund manager has no competitive edge to speak off and seems to know it.

What are the consequences of herding?
If herding is a fact of life in both portfolio management and corporate finance, here are the consequences.  
In portfolio management: The bunching up of buying and selling on the same stocks will increase correlation over time in stocks (serial correlation), as an up day on a stock will attract more buyers in the near term, resulting in more up days. This will make momentum strategies more lucrative, at least in the short term. It will also make pricing bubbles and corrections more extreme and the latter will lay waste to the momentum strategies that looked so good before the correction. That "random walk" down Wall Street just became a lot more like a drunk walking down the street, overshooting in both directions.
Corporate finance:  If CFOs indulge in herd behavior, corporate financial practice will reveal "me-too" characteristics. In terms of financial policy, companies will try to set dividends and debt policy to be as close to their peer group as possible. If you are in a sector where everyone borrows money and pays dividends, you too will do so (even if you cannot afford to pay dividends or carry debt). An acquisition or buyback by one company is a sector should set off a wave of acquisitions and buybacks by other companies in the same sector. Not surprisingly, mistakes, when they occur, will be sector wide (like those telecomm companies that borrowed too much money in the late 1990s) or even market wide.

Can you take advantage of herding and if so, how?
 There are two strategies that you can adopt in a world where herding is the rule, rather than the exception:
a. The Yogi Bear strategy: The movie that just came out was disastrously bad, but Yogi Bear was "smarter than the average bear". To adopt the Yogi Bear strategy, you have to be smarter than the average investor. Essentially, you play the momentum game, reaping profits from herd behavior, but you get out just in time, before the correction hits. You get all of the upside of herd behavior and none of the downside. I had an extended post on momentum investing a while back, where I noted that while I don't think I can pull this off, there are others who can.
b. The Yoda strategy: Every investment sage will tell you that you should not be part of the herd and that you can make more money as a contrarian. Easier said than done. To succeed as an idiosyncratic investor, here is what you need:
a. A competitive edge: There is no point going against the crowd, if you have little to offer that is unique or different. It is a point that I have made several times before, but you need to bring something to the table before you bet against the crowd.  In Yoda's words: 
"You will find only what you bring in."
b. Self confidence: You have to believe in yourself. Without a core investment philosophy, it is difficult to hold on in the face of peer group pressure. As Yoda would say: “Do … or do not. There is no try.”
c. A patient client base: If you are investing for yourself, you have to answer only to yourself. If you are investing for others, you need investors who trust you to be right in the long term.

Wednesday, January 5, 2011

The Facebook Valuation!

One of the biggest stories of the last week was Goldman's $ 500 million investment in Facebook for approximately 1% of the company. Extrapolating from the transaction, we obtain an implied value of $ 50 billion for Facebook, a number that has been making the rounds in news stories over the last few days. There are three questions that emerge from this news story: (a) With private businesses, can you extrapolate from a single transaction amount to an overall value? (b) Why would a company worth billions choose to stay private, when it clearly has the option to go public? (c) How would you value a share in a non-listed, non-traded company (as opposed to a publicly traded company)?

a. Can you extrapolate from a single transaction amount to an overall value?
Sure, as long as the transaction is an arms length one and all you are getting in return for your investment is a share of the company's equity. If, as is common, there are side benefits or side costs that go with the transaction, extrapolation will yielding a misleading estimate of value. In the case of the Goldman transaction, there are plenty of reasons to be skeptical. In addition to getting a piece of Facebook, Goldman also gets the following benefits:
a. Investment opportunities for Goldman's clients: As part of the deal, Goldman will be raising $1.5 billion from its clients to invest in Facebook. While this may seem to be a favor that Goldman is doing for Facebook, the reality is that Facebook is a hot company to invest in and this will allow eager investors an exclusive entree into the company.
b. A front seat for the Facebook IPO: If at some point in time, Facebook decides to go public, Goldman is likely to be the lead underwriter and reap a big share of the commission.
c. Private wealth management services to Facebook's potential billionaires and millionaires: When Facebook goes public, Mark Zuckerberg and a number of other executives will have the capacity to sell their shares in the market. While I do not expect a wholesale cashing out of equity positions immediately after the IPO, it is likely to happen over time, at which point these very wealthy individuals will need some private banking help and Goldman will be there to provide that help.
The profits and fees from these added businesses could account for a significant chunk of the $ 500 million that Goldman paid in this transaction. Exactly how much will depend on the likelihood of an IPO and the fee structure for the transaction. If, for instance, the present value of the expected fees from these side benefits is $ 200 million, the implied value for Facebook will be $ 30 billion, rather than $ 50 billion.
One more note of caution. Strange though this may sound, I would trust a market price derived from a consensus of a thousands of buyers and sellers to get the value right more than I trust the price from a single transaction, even if the buyer and seller are supremely sophisticated.

b. Why would a company worth billions choose to stay private, when it clearly has the option to go public?
Facebook's reluctance to go public may seem surprising. After all, the conventional wisdom has always been that companies like Facebook should get a more favorable response from offering shares in the public market place than from private offerings to venture capitalists and large investors. Here are some reasons, rational or otherwise, for why Facebook may be holding back:
i. Extending the tease: Looking at the favorable publicity that Facebook has got in the last week from the Goldman deal, it does not look like waiting to go public is hurting Facebook, at least for the moment. In fact, it may be making Facebook an even more desirable investment to those who cannot invest in it right now.
ii. "Proprietary" information: While I don't think that this is a big factor for Facebook, there are some companies that choose to stay private because they are afraid of revealing proprietary information about their products/services to the general market. Instead, they can provide the information, with sufficient restrictions on disclosure, to a few wealthy investors who can then invest in the company.
iii. Founder idiosyncracies: If the founder and majority stockholder in a company decides that he does not want the company to go public, the company will not go public. In the case of Facebook, it is entirely possible that Mark Zuckerberg has decided that he does not want to take the company public and he does not seem the kind of person who can be dissuaded easily.
iv. Regulatory and information disclosure concerns:  From Sarbanes-Oxley to SEC restrictions, public companies are constrained in ways that private businesses are not.
v. No valuation scrutiny: As a publicly traded company, no matter how well regarded it may be, the market valuation will be questioned by skeptical investors. Scaling value to earnings or book value, investors will argue that the company are over priced, relative to other companies in the market. (Take a look at Apple, Google and Netflix, all big winners over the last year, and you will see this phenomenon at play). Facebook gets to have the best of both worlds, again at least for the moment. We get glimpses of its immense value, each time a transaction is made, and no real way to examine whether the value makes sense, since we do not have access to much of the information we need.
In summary, Facebook is in a unique position. It has the profile to raise capital from wealthy investors are favorable terms and is getting many of the benefits of being a publicly traded company without any of the costs. Could that change? Absolutely. If there is bad news (or even rumored bad news) about the company and some or even a few investors have trouble exiting the company, the estimated value could melt down quickly.

(c) How would you value a share in a private company (as opposed to a public company)?
Let's assume that you are one of those lucky investors that has a chance to invest in Facebook. How would you go about valuing the company?
i. Financial data: You have to get your hands on some operating numbers. All you have right now is rumor: Facebook supposedly will generate $ 2 billion in revenues this year and there is no word on how much earnings they will have. You cannot value a company based upon information that is this threadbare and you will need fuller financial statements.
ii. Future projections: Once you have the information, you have to make projections for the future, valuing  Facebook just the way you would value any young, high growth publicly traded company. I have a paper on the topic. Normally, with private businesses, you will discount the value for lack of liquidity but I don't think this is a concern with Facebook shares, even if privately held.
iii. Ownership protections: I don't know about you but I just finished watching Social Network, the movie, and I am not sure that I feel secure that my ownership rights will be protected by the controlling stockholders at Facebook. I would need to make sure that there are enough protections in place for existing stockholders in the event of new capital being raised or an IPO.

So, is Facebook worth $ 50 billion? Based upon current revenues of $ 2 billion, it is richly priced; 25 times revenues and god only knows how many times earnings. The justifications that I hear from analysts for the high valuation are:
(a) An unprecedented platform: The 500 million users provide a platform that could generate much higher revenues and earnings in the future, but a lot of things of things have to go right for this to work out. I am not a big user of Facebook, but my gut feeling is that an overt commercialization of the space will make it less attractive to many users. So, it has to be subtle and creative commercialization... while fending off competition. (Remind me again what happened to Myspace, another hot place to be not so long ago).
(b) Goldman knows best: Smart investors (like Goldman) think its worth $ 50 billion. So, it must be worth $ 50 billion. This line of reasoning is so absurd that it is not worth pursuing. If you think that Goldman does not make big valuation mistakes, you are wrong. What Goldman does well is cut its losses, if it does make mistakes.  You and I will not have that option.
(c) The Big Story: To those who use the big story justification, everyone will be on a social network in the future, and you need to pay a premium to be part of the movement. Having heard variants of the big story before used to justify other bubbles (dot com, telecomm, PCs), I don't buy this. I think the market may be right about the macro story but is being hopelessly over optimistic about the micro pieces. In other words, we may all be parts of social networks a decade from now, but can all of these social networking platforms (Facebook, Twitter, Groupon...)  be profitable? My guess is that there will be a few big winners and lots of losers, before the final story is written. (Remember that the market was right in 1998 about dot-com retailing being the wave of the future but most dot-com retailers never made it through to nirvana. Amazon did and it is worth almost $ 80 billion, but it is the exception.)