• Facebook is so ridiculously profitable that makes management blush under increased public scrutiny on security and privacy grounds.

• Hence the self-inflicted punishment: massive investments in safety and security resulting in 15 percentage points margin compression over a number of years.

• Facebook's multi-billion dollar commitment to safety may not appease regulators, but will elevate Facebook's barrier to entry even higher.

• We claim that business tailwinds will outweigh the headwinds in the mid to long term.

• We value the business based on very conservative assumptions and argue for a long position at $160/share.

There’s a lot going on at Facebook (FB) these days: Russia meddling, fake news, hate speech, Cambridge Analytica… Add to that management guidance for decelerating revenue growth, margin compression and rising capital expenditures, and you get a serious selloff.

Yet despite all the uncertainties, we see lots to like on Facebook at $160/share. In fact, if we had to summarize our take on the business in one paragraph, it would be this one:

Facebook is a ridiculously profitable business.

So ridiculously profitable, that under increased public scrutiny on security and privacy concerns, management is apologetic, almost ashamed of its profitability.

By ridiculous profitability, we are talking 50% operating margins and 45% ROIC for full year 2017. For context, operating margins at Google (GOOG) (GOOGL) and Apple (AAPL) are in the low to mid-20s %.

Now, this is not to deny that Facebook has been reaping the benefits of its stranglehold in social without bearing the full cost of its responsibilities. That is to say, that 50% operating margins and 45% ROIC are unsustainable.

But we have a hard time interpreting plans to build a permanent 20,000 safety workforce and expectations for a 15 percentage point hit to margins over a number of years, as anything but an exaggeration. A purposely overreaction to redeem past sins. “We’ve learnt our lesson, we’re willing to pay a hefty price, now let us preserve the social throne”.

 Facebook CEO Mark Zuckerberg testifying on Capitol Hill, April 10, 2018, in Washington, DC. (Photo: JIM WATSON/AFP/Getty Images)

On profits and shame

Part of the 15% margin compression is associated to investments in growth (video content, virtual and augmented reality, etc.) rather than safety and security measures. Still, even a 10% hit to margins to improve safety would require annual expenses to increase by $4+ billion based on 2017 revenues.

What’s more: by the time the full margin compression is realized (2021 at the earliest, in our view), revenues are likely to exceed $110 billion, hence a 10% hit to margins related to safety initiatives would necessitate an annual budget of $11 billion. Again, for reference, that’s 110,000 employees searching for fake news, hate speech, violent content and the like, based on a cost per employee of $100k/year.

In reality, machine learning will do the heavy lifting, and Facebook will replace labor with automation where possible, but the fact remains that $11 billion is an insane amount of money to devote to safety enforcement every year. And even if a large human team is necessary initially, won’t those employees be working on their own demise by building the multi-language labeled datasets required to train the neural networks to detect unwanted content? How many years of data labeling until they are let go?

In other words: while most companies struggle to turn a profit, we think Facebook is going to have a hard time keeping profits below the threshold of shame.

Competitive advantages

But there’s even more. Consider the following extract from COO Sheryl Sandberg’s Q2 2018 prepared remarks (emphasis ours):

We’ll keep looking for ways to improve, and we hope these tools become standard across the industry.

She was talking about transparency and accountability. But if we broaden the context to security as a whole, a fundamental implication becomes evident: by pledging tens of billions of dollars to raise the standards of safety in social platforms and having authorities come to expect those standards, Facebook is radically upping the fixed costs required to operate a social platform.

Part of the costs to maintain safety and security in a social platform are traffic related. But even a tenth of Facebook’s $11 billion hit to margins would be sufficient to bring Twitter (TWTR) to the red. Not to mention what Facebook’s future standards of safety would do to Snap’s (SNAP) financials.

So yeap. Look beyond the mea culpa, and you’ll see Facebook is building another massive barrier to entry around its social ecosystem, this time built around safety. Want to play in Facebook’s turf? Be ready to weather multi-billion recurrent expenses to enforce the revised expectations of safety, and even then, try to match the wealth of data Facebook will have to automatically detect abuse.

Just for completeness, other competitive advantages of Facebook include:

• The powerful platform effects inherent in social networks

• The switching cost to millions of SMBs that have built and grown their business in Facebook, and rely on the platform for customer relations and acquisition.

• An unmatched pool of personal data to inform targeted advertising.

• A superior ability to attract and retain top talent (despite all the scandals, Facebook is Glassdoor’s 2018 Best Place To Work and 96% of employees approve of CEO Mark Zuckerberg).

All these system dynamics make of Facebook one of the most defensible moats in Western tech.

Risks and opportunities

There are several reasons beyond investments in safety and security to temper short-term earnings growth expectations:

• Facebook active users have peaked in Europe and North America.

• Reputational damage from the Cambridge Analytical data scandal and more data privacy choices (in EU through GDPR and US by company’s initiative) will continue to pressure active users in most lucrative markets.

• Reduced ARPU (average revenue per user) for users opting out of targeted advertising.

• Increasing regulatory overhead and fiscal pressure.

• Add load at Instagram has catched up to Facebook: no more low-hanging fruit.

There are as well a number of long runway tailwinds and optionalities, which in our view, more than offset the headwinds. We list them below, in decreasing order of certainty:

• Active user growth outside of Europe and North America: India, Indonesia, Brazil, etc.

• Increasing ARPU due to continuous improvements in ad targeting technology increasing advertiser ROI.

• Increasing ARPU outside of North America, as foreign digital ad markets mature: quarterly ARPU in Europe ($9), Asia Pacific (below $3) and rest of the world (below $2) is markedly lower than in US and Canada ($26).

• Strong user and engagement momentum in Instagram, Messenger, and WhatsApp.

• Facebook stories have not yet been monetized.

• Facebook Messenger has barely been monetized, and WhatsApp remains unmonetized.

• Video initiatives: Facebook Watch, IG Video.

• E-commerce: we believe that Facebook platforms (especially Instagram) will be the most serious contenders to Amazon in long-tail categories such as luxury and fashion. Facebook is already a big e-commerce player in the discovery phase and is making inroads down the marketing funnel. A partnership with an inventory management and logistics player (JD?) would help it integrate fulfillment into the platform.

• Wildcard: Facebook transforming WhatsApp and/or Messenger into a mega-app à la WeChat (TCEHY).

• Wildcard: VR and AR initiatives (Oculus).

Valuation

Our readers know that we strive for simplicity when valuing our companies. The best investment theses rely on a few key pieces, and identifying those pieces is halfway towards profitable investing. The other half is getting those pieces right, and so we search for opportunities in which the key pieces can be assessed with relative certainty.

That’s the reason why, when possible, we value companies on an earnings power value basis (EPV). The EPV gives a rock-bottom estimation of value without too much reliance on future forecasts. Buy a first-class business with defensive competitive advantages at a moderate premium over EPV and you are likely to achieve satisfactory returns over the long term.

Now, EPV is not appropriate for Facebook, given the outlook for decelerating revenue growth and declining margins over the next years. Instead, we value the company based on (very) conservative assumptions for the period from Q3 2018 to 2021.

The guidance for full year 2018 provided in Q2 consists of:

• Revenue growth decelerating by high single digits in 2H relative to previous quarters (it was 49% in Q1 and 50% in Q2).

• Growth in annual costs and expenses in the 50% to 60% range.

• Effective tax rate in the mid teens.

• Capex of $15 billion.

Moreover, management guided to operating margins trending towards the mid 30s on a percentage basis “over the next several years” (they were 50% in 2017).

Reading conservatively into that guidance, for 2H 2018, we model:

• YoY revenue growth of 35% in Q3 and 33% in Q4.

• YoY total expense growth of 65% in both Q3 and Q4.

• Effective tax rate of 17% in both Q3 and Q4.

Furthermore, we assume that operating margins decline in each of the upcoming years, reaching 35% already in 2021. We model modest revenue growth of 30% in 2019, 25% in 2020 and 20% in 2021, and an effective tax rate creeping up to 17% in 2021, to account for increasing fiscal pressure.

The resulting NOPAT (net operating profit after tax) from 2018 to 2021 is shown in the table below.

 Source: Investment Works Research

Our conservative model implies YoY NOPAT growth of 22% for full year 2018 (it was 45% in H1 2018), trending down on an annual basis to a meager 13.5% in 2021. For the per share figures, we use the 2,930 million diluted shares as of June 2018 (the cost of SBC to shareholders is included in costs and expenses and we assume that the company doesn’t add value through repurchases).

Now, using 7% as discount rate, the net present value (NPV) of estimated NOPAT for the 2H 2018 to 2021 period is $29/share. For periods after 2021, we assume zero growth in perpetuity, hence the NPV of NOPAT in 2022 and beyond is equal to 14.3x (1/0.07) the present value of 2021 NOPAT, or $127/share. Once again, we are after a reasonably conservative estimation of intrinsic value, not our best expectation, since we are trying to determine the price at which we would be buyers of the stock. If the assumptions for NOPAT growth until 2021 are conservative, all the more the assumption of constant NOPAT in perpetuity backed into the 14.3x multiple.

Moreover, the books show some $14/share of non-operating current assets (distributable cash and marketable securities). Finally, to err on the side of caution, we subtract from intrinsic value the NPV of the amounts by which we expect capital expenditures to exceed D&A charges (the latter, part of cost and expenses) in the period from 2H 2018 to 2021: $9/share, or $26.5 billion. We neglect working capital expansion requirements during the period since Facebook’s net operating working capital is tiny, below $3 billion as of June 2018.

Putting all parts together, our conservative estimation of intrinsic value for Facebook is just above $160/share, roughly equal to the current stock price.

Source: Investment Works Research

Investor takeaways

The recent selloff has provided an interesting entry opportunity into Facebook.

At $160/share, investors can expect 7% annual returns for the foreseeable future even in the face of a drastic deceleration in earnings growth. Any upside relative to the bearish assumptions in our model would produce additional returns.

We may initiate a long position in Facebook in the coming days. Beyond the merits inherent in buying a strong-moat business with a long growth runway at a price below intrinsic value, an investment in Facebook would round off the exposure of the IW Portfolio to an inevitable mega-trend: the shift of advertising dollars towards programmatic.

Earlier this year, we initiated long positions in The Trade Desk (TTD) and Alphabet (long theses here and here). Google and Facebook are the two dominant walled gardens, with a combined 57% market share of US digital ad spend. The Trade Desk is poised to become the dominant demand-side platform for purchases of ad inventory outside the realm of the two gardens.

Facebook also brings to the table a wealth of valuable personal data and artificial intelligence know-how, which reinforces the AI-exposure of another major portfolio component: Nvidia (NVDA).

We will update our thesis on Facebook after the Q3 earnings release in October. We publish research on stock picks and the names in our coverage universe on a weekly basis.