Often times when analyzing an investment there is a single point of consideration that can make or break the thesis. When considering Facebook we believe that all the elements for success are in place: An enduring product with a wide economic moat and high levels of profitability. Unfortunately, the wonderful business characteristics of Facebook are widely known and the price of the stock reflects this. Due to the high price multiple the crux of the Facebook thesis hinges on one factor: How big can the company get?
When considering whether or not to purchase Facebook the Total Addressable Market, or TAM, is extremely important. The reason this particular metric is so important for Facebook is that the price multiple is high. If you purchase a business at a low price multiple it could be a low or no growth business and still be an outstanding investment. Purchasing a company with a P/E of 10 that pays out 100% of earnings as a dividend is equivalent to purchasing a 10% yielding bond – an excellent investment in today’s market. On the other hand, purchasing a company with a TTM P/E multiple of 40 means growth is a must. In this case the earnings yield is only 2.5% and to justify such a high price, not only does the business have to grow, it has to grow a lot. If the company is able to grow 400% over the next five years and then plateaus, the earnings yield would be 10% from your original purchase price five years earlier. This is identical to purchasing a 2.5% bond today that has the coupon payments rise to 10% and then is fixed. Obviously it’s better to buy a 10% coupon bond today. In light of the elevated price of Facebook we need to answer this challenging question: How big can Facebook get?
In 2016 Facebook generated $27 billion dollars in revenue and we estimate that the owner earnings (our internal calculation of profits, different than GAAP) was roughly $12 billion. The market capitalization today is $447 billion, but the company has $32 billion in cash that needs subtracted from this figure, for a net purchase price of $415 billion. The TTM P/E ratio is 34 for an earnings yield of 2.9%. If Facebook were to increase in size by 300% in the upcoming years and plateau it would be roughly equivalent to purchasing that previously mentioned 10% yielding bond. We believe Facebook is going to have to grow more than 300% in the next 5 years to justify the current price. Is this even possible? Let’s think out loud and see if we can estimate the TAM for Facebook.
How Big Can the Advertising Business Get?
There are a lot of different ways to think about the potential size of Facebook. We will start by focusing solely on the advertising business since this accounts for 97% of revenue today. When attempting to determine the size of the ad business 5 years from now, rather than forecasting growth one year at a time moving forward, we are going to reverse engineer the problem and start by thinking about the world five years from today. In order to do this we need to blend imprecise facts (such as how much advertising occurred on planet earth in 2016) and some educated guesses, known as SWAG. The blend of vague facts and SWAG should arrive at something very precise (that is a joke). So, without further ado, let’s see if we can estimate the future size of Facebook.
Angle #1. Total Global Advertising in 2021
In 2016, it is estimated that the total dollars spent on advertising worldwide was $542 billion. In 2016 Google generated advertising revenues of $79 billion, or ~14.5%, of all advertising on the planet. Facebook generated $27 billion, or ~5%. So the question is – how much of this pie can Facebook get? Let’s assume global advertising continues to grow at a rate of 5%, which is in-line with growth over the last five years. That would imply that the total global ad spend five years from today would be about $691 billion.
The next step is to estimate what percentage of the total global advertising spend Facebook will capture five years from now? Remember, Google currently has a 15% global advertising market share and Google does intentional advertising. This is the type of advertising where people hop on-line with the intent to do something. They type in key words and go buy stuff. It has proven to be the most successful form of advertising in the world. Facebook’s advertising is interruption advertising, meaning they interrupt your news feed with things you may or may not want to see. Now grant it, this interruption advertising is very targeted, since Facebook knows important things, such as your age, geography, interests, etc. This makes Facebook ads far more effective than the shotgun approach, where companies throw up a billboard for all to see, but perhaps not as good as the intentional advertising of Google. That being said…a difficult but important question – can Facebook capture a higher percentage of global advertising spend than Google? That’s not an easy question, but we think the intentional vs interruption comparison gives Google an edge. Now, let’s take an educated guess at Facebook’s market share five years from today and combined that with an assumed 35% net income margin (a bit higher than today). The table below is our first swipe at some high quality SWAG:
Estimated 2021 Global Ad Spending: $692 Billion
I think it’s very safe to assume the number will be bigger than 5%. The company is already at 5% and small businesses are seeing tremendous success on the platform and intend to ramp their spending. What about 20%? Not as clear. I personally think it will be challenging for Facebook to capture 15% of all ad revenue on the planet. If Facebook does capture 15% of the market by 2021 and matures the earnings yield would be 8.7% based on the original purchase price:
Earnings = $36 Billion / $415 Billion
Earnings = 8.7%
Before we pat ourselves on the back for performing some high quality guess work, let’s reflect a moment and think if this is a reasonable way to approach the problem. In this thought experiment we are comparing the ad revenue of Google and Facebook to the total advertising market, but is that an appropriate comparison? I’d say there is an argument to be made that Google and Facebook not only replace traditional forms of advertising, but also some traditional methods of sales & marketing.
Sales and marketing is fundamentally different than advertising and includes activities like cold calling. The sales and marketing budget at some firms is many multiplies of the advertising budget. In some instances, Facebook and Google are more effective at building a sales funnel and moving people through certain parts of that funnel than some traditional sales approaches. In-bound marketing has largely replaced cold calling and there is a good chance other elements of the sales funnel, especially towards the top, will be more effective using Google and Facebook. As this shift occurs dollars spent on functions like cold-calling become ‘ad’ dollars. This has the potential to change the growth rate for global ad spend from 5% per year to a higher number. For this reason, the global advertising market might not be the correct benchmark for our thought experiment. It might be better to compare the total advertising market and parts of the global sales and marketing spend.
Angle #1 Conclusion - Net Income in 5 Years: Between $24 and $48 billion?
Angle #2: It’s a Small Business World After-All
Perhaps Angle #1 isn’t the best angle to take. Let’s instead focus on Facebook as a different type of advertiser than Google. Certain products will probably always sell better on Google, such as car insurance policies. Geico purchasing leads for anyone typing in “car insurance” is a sure fire way to generate more business. However, for small and local businesses Facebook may have an edge. Facebook is aware that I enjoy chess and a well-placed advertisement about a chess club near me, with a group to join, a list of my friends already going, may be quite enticing. We’ve observed that small and local businesses are having a lot of success with Facebook campaigns. It appears to us that these businesses are seeing good ROI’s and strongly considering ramping up their spending. There are already almost 5 million small businesses globally advertising on Facebook. So how big can Facebook get focusing on small businesses worldwide? Let’s once again combine vague facts and some SWAG.
There are 27.9 million small business in the United States, however, roughly 22 million of these are individuals that are self-employed leaving about 5.9 million with more than one person. There are 120 million individuals working in small businesses, meaning 98 million in businesses with more than one employee. Doing a little math we can infer that small businesses, with at least 2 employees, have an average employee count of ~17. Companies of this size are large enough to do some ad spending. So let’s say all of these end up advertising on Facebook…how much would those 5.9 million small businesses spend per year? Random guess work builds the following table:
5.9 Million Small Businesses Facebook Ad Spend
So, the above table is a rough starting point. Really rough. But let’s move forward from here with some more assumptions. Let’s assume all businesses globally end up advertising in the same way as US small businesses.
The US GDP in 2016 is close to $18 trillion and the Global GDP is about $76 trillion. We have to exclude China, roughly $11 trillion, because Facebook has an immaterial presence in China. This means that the United States represents about 27% of the ex-China Global GDP. So basically, if all countries end up advertising like US small businesses…the global opportunity is 4x than the US opportunity alone. Most of this would be in Europe and APAC regions. So let’s multiply our previous table by 4 to arrive at the global Facebook revenues from small businesses around the world in a fully saturated market:
All Small Businesses Facebook Ad Spend
So what will the average annual spend per small business be? According to MerchantCircle, more than 50% of small businesses spend less than $2,500 per year on advertising in the US. This is very close to our $2,400 guess, which equates to about $23 billion in net income. To finish off our TAM calculation we’d have to add in the big businesses as well. To do that let’s consider these two pieces of information: In developed nations the % of GDP from small businesses and the % of GDP from big businesses is almost 50/50 (It’s 44% / 56% in the US). We could assume they advertise in equal amounts, but I suspect big businesses advertise more per $ of GDP contribution. We’ll double the ad spend per % of GDP for the big businesses and add this to our estimated small business number:
Income from $2,400 Per Year Ad Spend
Small Business + Big Business = $23 Billion + $46 Billion = $ 69 Billion
This result is reasonably close to Angle #1 – only off by a factor of 1.5x. It assumes a lot, such as 100% advertising for all businesses, a $2,400 per year average, etc. But this does help shed some light on a ball park global opportunity. However, if small businesses did end up spending $6,000 or $12,000 per year our estimate is way off and buying Facebook could be a steal. This seems to be a bit too speculative an assumption for our taste.
Angle #2 Conclusion - Net Income in 5 Years: Between $33 billion and $58 billion?
Angle #3: Digital Advertising is Traffic Bound – Let’s Look at ARPU
The approach I like best deals with analyzing Facebook’s overall traffic. Online properties like Facebook and Google are traffic bound. This means that the amount of revenue they generate is a function of the amount of traffic they generate. Our objective in this approach is to estimate how traffic and monetization will develop leading up to maturity. The key elements are to determine the number of users by region and the average revenue per user in each region.
Let’s start off by showing where Facebook and Google currently stand:
Facebook, Google and Global Internet Information (Ex-China)
Google currently has an ARPU of $133 per internet user in North America. This is a relatively mature market, but even Google has posted 20% growth in this region in the last 12 months. However, it might be safe to assume this rate of growth should slow down. Google just recently started putting tremendous effort into monetizing YouTube and much of this growth may be from this ‘new’ monetization challenge. It seem possible that the ARPU should stabilize for Google in the near future. Our reason for making such a bold statement is based off a few assumptions.
First, let’s assume that the adverting spend in the US will continue to grow between 3% and 7% per year. We make this assumption because we believe that the advertising dollars spent in the United States is ultimately a function of GDP – and it cannot grow much faster than GDP for long periods of time. In 2016 there was $200 billion in advertising dollars spent in the United States, or $547 per US citizen. Google and Facebook combined captured $210 of this $547. If both companies increased ARPU in the US at a rate of 20% per year for the next 5 years they would combine for $522 in adverting per US citizen and represent almost 74% of ALL ad dollars in the US. This seems unlikely…so we believe something has to give…meaning Google will probably slowdown in the near future.
Assuming Google does slow down, we can the build a table of what Facebook would look like at maturity. We perform this by assuming Facebook ends up with a monetization rate, at maturity, which is some % of Google’s. The tables below show results for a mature 50%, 75%, 100%, and 200% relative monetization rate compared to Google:
So where will Facebook’s ad business end up relative to Googles? Here at Selective we think its bold to assume it’s more than 100% of Googles due to our previously mentioned intentional vs interruption advertising comparison. This would imply that Facebook’s income at maturity might be between $29 billion and $38 billion.
Angle #3 Conclusion - Net Income in 5 Years: Between $29 billion and $38 billion?
Conclusion from our Advertising TAM Analysis?
Our three angles of attempting to derive a TAM for Facebook’s ad business produced similar results that suggested Facebook’s net income from advertising might mature in the range of $30 billion to $58 billion. While this is a broad range, and the guess work performed was rough at best, it provides extremely valuable information. Most businesses with growth rates in excess of 30% could have TAM where the net income for the business could potentially rise by a factor of 10x, 100x, or even 1,000x before the TAM is saturated. When the market opportunity is sufficiently large a high P/E ratio and traditional valuation tools, such as the PEG ratio, can be useful. However, a PEG ratio is pretty useless if a business is going to grow 35% for 3 years and stop. We don’t know if or when Facebook will decelerate, but it appears obvious the TAM for the advertising business is less than 10x – perhaps even about 5x. It’s safe to assume that in the next decade the ad business for Facebook should ‘mature’. So our next question…
What Will Growth Look Like When Facebook is Mature?
We believe that even a ‘mature’ Facebook will have a reasonable growth rate. The company will benefit, almost for free, from a few factors working in tandem: Inflation, Population Growth, Real Increases in ARPU, and Increased Internet Penetration Rates. Below we estimate the annual growth from each of these factors, at maturity:
Inflation: +2% per year
Facebook generates revenue in countries all over the world with various rates of inflation. As US citizens, our primary concern is the rate of inflation of the US dollar on an annual basis and how the ad revenue will change based on this inflation. We estimate that the ad dollars spent on Facebook’s platform worldwide will increase roughly 2% per year solely based on inflation. This particular type of ‘growth’ does not increase real purchasing power, but due to the capital light business model of Facebook it would compound the Goodwill carried on the balance sheet, or in this case the Market Capitalization, quite nicely (see Warren Buffet’s 1983 Annual Letter, Appendix A, Goodwill and its Amortization: The Rules and The Realities).
Population Growth: +1% per year
Facebook should benefit from the 1% growth per year in population for the foreseeable future as the global population increases.
Real Increases to ARPU: +5% per year
Facebook should be able to win additional business from advertisers by increasing the efficiency of its ad placements. If ads have higher conversion rates companies would increase their advertising spending. Facebook will also benefit from the real GDP growth of countries everywhere. As Asia-Pacific, Africa, and Latin America continue to develop the ARPU should trend upward toward the US and European ARPU. Let’s say the worldwide GDP grows at a rate of 2% per year in real terms - this should help add to Facebook’s advertising dollars overtime. The combination of improved placements and the closing of the monetization gap between developed countries and Rest of World could add as much as 5% per year at maturity.
Increased Internet Penetration Rates: +1%
Most of the world is still not online. The current number of internet users is about 3.5 billion, leaving 4 billion people without internet. Surprisingly only 84% of Americans are online and US internet user penetration increased 6% year-over-year in 2016. This goes to show the very long run-away for the rest of the world. We assume Facebook would benefit about 1% per year from more users coming online globally. At this rate it would take about 72 years to get everyone online.
Total: 9% Growth Rate at Maturity
This actually compares very favorably to most S&P 500 businesses current growth rate and the increase would be very close to free. At maturity Facebook may be able to distribute 100% of earnings as share repurchases or dividends and still compound at 9%. This is similar to how Visa has been growing for the last 10 years and shareholders have been wonderfully rewarded. If the company traded at 33x earnings at maturity it would still be growing EPS at a rate of 12% per year (9% from the above sources + 3% from share repurchases). If the P/E ratio contracts due to a lower growth rate, which is probably would, the EPS growth at maturity would be even higher due to the increased effectiveness of the share repurchase program.
Don’t Forget China
Facebook is still not widely adopted in China and our forecasts and projections exclude this particular country. It is probably wise to continue to exclude the country since the government policies in the future are unclear at best. However, if China allowed access there is a very high probability that it would be a boon to Facebook. The country has 1.3 billion people, or 17% of the world’s population. I suspect the country would rapidly move on to Facebook to be connected globally which could add about 300 million new users quickly (this is about 45% of all of the internet users in China), and substantially more over time. With an ARPU of perhaps $5 per year, a little less than the rest of APAC, it would add about $1.5 billion to Facebook’s revenue. A nice bump, but not even an additional 10%. The real benefit would be the long-term participation in the growing internet penetration rate, increasing GDP per capita, and the closing of the monetization gap between China and the US.
Don’t Forget Non-Ad Revenue
Our analysis of this point is focused solely on the ad revenue, but the value of Facebook’s platform goes well beyond ads. One way Facebook could increase the value of the platform would be to acquire a CRM, such as Salesforce, and integrate all the knowledge about its user base into the Salesforce product. Imagine the value of knowing quickly and easily a few personal interests of a potential client prior to a meeting. You are aware he is a huge fan of Ludovico Einaudi and look up his concert tour to inform your client when he is in the area. Being thoughtful and knowledgeable prior to a client meeting has been a sure-fire way to success for more than 50 years (read Dale Carnegie’s How To Win Friends and Influence People). Facebook could become a platform for B2B thoughtfulness! That is a high quality CRM. This represents one of a plethora of ways Facebook could create value for individuals and business outside the realm of advertising. It is possible that the revenue opportunity from non-ad sources could rival the ad business overtime. We would view this as a free ‘option’ on the upside.
Focus on the Downside
Howard Marks famously stated that “If you don’t lose money on an investment, all the other alternatives are pretty good.” For this reason it’s important to focus on the downside risk. We believe the biggest risk, by far, with the purchase of Facebook is overpaying. Assuming we are correct about the competitive position of the business being rock solid (if not, it can go to zero) – how much can we lose based on overpaying? The table below assumes the business hits ‘maturity’ at our worst case scenario for Angle’s #1, #2, and #3 and then compounds at 9% for the next 5 years. The market capitalization five years from today is highlighted below:
It appears that there is a pending ceiling for the advertising businesses of both Facebook and Google. Sometime in the next half decade the growth rate for both businesses should slow and for Facebook the growth rate will probably settle in a low to mid-double digit range. The TAM for Facebook appears to be about 4x to 5x the current size of the business and not 10x or 100x. If we are correct about the TAM for Facebook, under the base case, the company would be earning between $24 billion and $60 billion at maturity. If the P/E multiple contracts to 20 at maturity the market capitalization would range between $480 billion and $1.2 trillion. This represents an upside of 10% to 300% over the next 5 years. If the TAM is much smaller and business matures sooner…it appears the downside is about 34% over a five-year period, but long-term growth would likely overcome this temporary decline. This analysis excludes non-ad sources of revenue, which could be material in the future.