Facebook shares could fall another 50%

With such a short history, the release of Facebook's (Nasdaq: FB) second-quarter earnings is an important opportunity to review the valuation in light of new data. Prior to the company's May IPO, I warned investors to steer well clear of the shares. While the near-40% decline the shares have suffered since the offering may tempt investors into believing they can now secure an attractive entry point, after consideration, my advice is unchanged today.

Facebook falls short
In an article published just after the IPO, I used a graph similar to the one below to show that Facebook's growth path was far outstripped by Google's (Nasdaq: GOOG). The graph matches up annual revenues for both companies over five-year periods: 2000 through 2004 for Google versus 2007 through 2011 for Facebook:

The trend is very clear and it suggests that in the following year, Google's revenue would surpass Facebook's. And that's exactly what is happening. Although we don't have Facebook's full-year results for 2012, we can add trailing-12-month revenue through June 30, 2011, instead. This is what an updated graph looks like:

Source: Company filings, S&P Capital IQ.

Google overtakes Facebook, with 12-month revenues of US$4.48 billion vs. US$4.33 billion! But that doesn't properly convey the gap in growth rates between the two companies. In the second quarter, Facebook's revenue grew by 32% year-on-year; the equivalent figure for Google during the second quarter of 2005 was 98% (in fact, in the second quarter of this year, Google's revenue ex-acquisitions increased 21%, which is remarkable for a company with trailing-12-month revenues that are now 10 times Facebook's.)

On July 22, 2005, the first full trading day after Google announced its results, shares closed at US$302.40, with a price-to-earnings (P/E) ratio of 88.7. Google shares have roughly doubled since then, for an annualised gain of 11.1%. Based on last Friday's closing price, S&P Capital IQ has Facebook's P/E at 82.2.

A high-risk bet just to earn the historic average equity return
What are the implications for investors? Let's assume that Google shares are currently fairly valued (or at least not significantly undervalued). In that case, if you're willing to bet that Facebook can achieve over the next seven years what Google has achieved over the past seven, then you can reasonably expect an annual return of 11%-12%. That looks — to me — like a very risky bet to achieve nothing better than decent returns. After all, we've already observed that Facebook is on a lower growth path than Google's. If I'm going to take punt on a long shot, I want to get paid in full if it comes in, and I don't mean 11% per year!

Another model I have referred to previously is that of econo-physicists Didier Sornette and Peter Cauwels (hereafter S&C), who valued the shares in a paper for the Journal of Portfolio Management, based on the following assumptions:

• Revenue per user per year is stable at US$3.50.
• Profit margins remain stable at 29%, for an annual profit of US$1 per user.
• Annual profits are discounted using an equity risk premium of 5%.

Based on these assumptions, S&C came up with a range of values for Facebook conditional on three scenarios for the maximum number of users the social network ultimately attracts ("base case," "high growth," and "extreme growth"). Under their extreme growth scenario, users top out at 1.8 billion (roughly twice the current number) and the company is worth US$32.9 billion.

Facebook's value: The hard numbers
On the release of Facebook's offering prospectus, S&C updated their model with new assumptions in terms of profitability (they increased expected annual profit per user to US$1.20) and user numbers. Under the new assumptions, the valuation associated with the base case and high growth scenarios increased, but the extreme growth scenario — investors' best-case outcome — yielded a lower value than previously, due to lower expected user numbers (S&C are now forecasting that Facebook's monthly active user base will stop growing in April 2013):

Source: Facebook IPO, Updated valuation and user forecasting, Cauwels & Sornette, May 2012.

With the quarterly earnings release, we have yet more data with which we can compare S&C's results.

• With regard to the ultimate number of total users, Facebook looks likely to surpass the base case scenario, since it is already at 955 million.
• With regard to annual profit per user, the revised assumption is holding up. Over the trailing-12-month period ending June 30, Facebook achieved a profit per user of US$1.24. Although we've got a very short history, this metric looks pretty stable — it was US$1.36 in 2011 and US$1.24 in 2010.

A 55% haircut
Facebook could well exceed S&C's base case valuation, but even S&C's most aggressive user growth assumptions produce a current value for the company of US$29.2 billion at the time of writing, or US$10.66 per share — a better than 55% discount to last Friday's closing price of US$23.70.

The nightmare isn't over
Facebook's IPO was botched, yes, but the shares have performed so poorly because they began trading at an outrageously inflated valuation. It's no more complicated than that. The same is true for Zynga (Nasdaq: ZNGA), which came to market without a hitch. In that context, Facebook shareholders should consider that horrific past performance does not mean their downside is now capped. While Facebook may increase its business value over time, shareholders will continue to suffer underperformance consistent with the shares' current overvaluation.