Wednesday, 26 February 2014

Facebook's acquisition of Whatsapp

This note looks at Facebook's acquisition of Whatsapp. Before writing the note, I was inclined to believe Damodaran's conclusion that the acquisition price was astronomical. However, after running some models I have concluded that the acquisition is okay given Facebook's high current valuation. The key issue for Facebook is to justify its current valuation, not justifying the acquisition of Whatsapp.



Facebook's valuation

Facebook's future is not determined by whether the Whatsapp acquisition is successful or not. Facebook's success is dependent on the following three factors:
  1. Growth in total advertising. Facebook is in the advertising industry, which has total revenue of USD 514B or $70.5 per person (worldwide, 2013, Zenith data). In 2000, revenue was $56 per person. The key driver of total advertising revenue is economic growth. Most of the increase in the last decade was driven by countries catching up to the western living standard. 
  2. Online advertising share. 20% of total advertising revenue is attributable to online advertising (worldwide, 2013, Zenith data). In 2000, that share was 3%. The key driver of online share of advertising is likely to be time spent on the media. If consumers spend more time online and less time reading print newspapers, that will be reflected in the amount of advertising online. 
  3. Facebook's market share. In 2013, Facebook had around 7% market share. This is a clear indication that Facebook has a workable strategy. 
The following three discounted cash flow calculations are not meant to be fully realistic. They are simplified, but serves an important purpose in understanding how the factors above are linked together. I will solve the models for a break even market share. All numbers (including discount factor) are in real terms without inflation. I have used a 50 year forecasting period.
  • Model A. Key assumptions: Population growth according to United Nations' medium forecast, real advertising consumption $70.5 per capita remains fixed, online advertising share 20% remains fixed, Facebook profit margin 44% remains fixed (actual Q4 2013), tax rate 30%, discount rate 8%.
  • Model B. As model 1, except real advertising consumption increases gradually to $490 per person in 50 years, i.e. the average worldwide GDP/capita will then be equal to the current US GDP/capita. This is equivalent to 3.9% per capita advertising growth per year.
  • Model C. As model 2, except online advertising share increases with 1 percentage point per year until 50% is reached.
In Model A, Facebook requires 52% of all online advertising revenue in perpetuity to justify its current share price, in Model B 23%, and in Model C 10%. Do remember that Model C requires 8.6% yearly market growth for 25+ years.

My objective with the number crunching is to show that Facebook is very highly valued today. These models can be improved by making better assumptions regarding Facebook's market share and profit margin in the future. In Model C, only 20% of Facebook's revenue is attributable to the coming ten years. A lot can happen thereafter. There is for instance political risk associated with a large portion of advertising revenue going to US companies. Other countries could make it difficult for these advertising giants by regulation. China is doing this today by banning Facebook, but in the future regulation would be much more sophisticated. The focus would lie on reducing Facebook's profit margin.


Why did Facebook buy Whatsapp?

We do not know the true reason for buying Whatsapp for $19B, but I will speculate. Here is what they actually bought:
  • Largest instant messaging application available on multiple smartphone platforms. 
  • 450M users, of which 71% are daily users (source: Whatsapp). Around 25-30% of the world-wide users of instant messaging are monthly users of Whatsapp (source: rough estimate). This is user share and will add to more than 100% since some use more than one instant messaging application.
  • User-friendly, minimalist application. Slow in adding new features (e.g. push voice was late, no ability to make voice calls, no application available on the PC).
  • Around 30 engineers and 60 employees in total (source: Internet)
  • Top management retained for for three year through special compensation.
Facebook already has Facebook Messenger so they are not buying a new product. It is also unlikely that the capabilities of the software engineers are the reason to acquire Whatsapp. This is in contrast to many other smaller acquisitions, which Facebook has justified by talent acquisition. 

There are three more likely reasons (1) gaining market share in online advertising, (2) gaining market share in instant messaging, and (3) uncertainly about Whatsapp as a competitor. The next sections will provide a commentary on the three reasons.


(1) Gaining market share in online advertising

Facebook's current strategy is to sell advertising space in the news-feed in their social network service. In 2013, advertising revenue per user was $5.99 per monthly user (plus an additional $0.76 for in-game purchases). 61% of the monthly users are also daily users. This is a working strategy, so a key reason to acquire Whatsapp is to add more users to the Facebook ecosystem. More users translate into advertising revenue. We need to assess two different sources of revenue. First, additional advertising revenue by turning Whatsapp users into Facebook users. Second, additional advertising revenue by advertising on the Whatsapp platform.

Additional Facebook users. Whatsapp is strong in some of the countries in which Facebook is weak (e.g. Africa, India, Indonesia;, but notably not China and Russia). This might make it easier to create new Facebook users. However, turning Whatsapp users into users of Facebook's social network site is not going to be easy. Most PC users are already familiar with Facebook and have probably decided not to use the service. There is for instance hardly any growth in users in the US. However, there will be a large number of smartphone users in developing countries that did not own a PC. Facebook will be novel to many of them. Already 24% of Facebook's monthly users are accessing the service only on mobile devices.

If Facebook can convert a proportion of Whatsapp's users to become regular Facebook users, there will be additional advertising revenue. Whatsapp has 450M users. We can assume a large portion of these users are already Facebook users, maybe 50%. We can further assume that some users do not want to become Facebook users, maybe another 50%. With this guesstimate there would be an additional 112M customers for Facebook's social network service. Facebook would have had an acquisition cost of $170 per new user (if there are no other benefits with the acquisition). 

Advertising on Whatsapp. If Facebook starts adding advertising on Whatsapp, there will be additional revenue. The management of Whatsapp has stated that there will be no advertising on Whatsapp, but such a statement is not legally binding. In fact, many companies have retracted similar statements (e.g. Google promising "do no evil" and no advertising on the search page).

Users spend a substantial time on Whatsapp, which means that advertising is likely to have an impact. Facebook should be able to target Whatsapp advertising as efficiently as it does on the Facebook platform. If all customers are potentially advertising customers, then the acquisition cost per user goes down to $42 (or $59 per daily user). However, another perspective is to note that Facebook can also sell advertising by turning current Facebook social network users into users of Whatsapp. 

The industry structure will not currently make it possible to add advertising, because users will move to other applications. However, in three to five years, the industry will have consolidated and users will have created stronger habits. And most users would have forgotten the promise not to use advertising. The introduction of advertising will then not result in a major loss of customers. Whatsapp can at the time also offer an advertising free subscription service.

Conclusion. To keep growing its market share in online advertising, Facebook needs products that customers use. If a substantial amount of time is spent on non-Facebook products, its advertising revenue will suffer. Acquiring services like Instagram and Whatsapp makes sense from this perspective. The direction of the migration flow does not matter. What is important to what extent users spend time in the social network service.

The Facebook Messenger application has not been very successful compared to the dedicated instant messaging services. Since users spend a lot of time on instant messaging especially in Asia (=trendsetter?), it poses a risk for Facebook's advertising revenue. Paying $42 per user of a very successful instant messaging service is not cheap and the payback period is 24 years (with Model C assumptions). However, we might instead conclude that Facebook paid an insurance premium of $16 per Facebook user. If too many Facebook users would spend too much time on Whatsapp, the core advertising revenue would be threatened.

If Facebook acquired Whatsapp to create new users of the social network service, we can expect to see a name change of Whatsapp in the near future. I don't think this will happen.


(2) Gaining market share in instant messaging

One of the oldest reasons for an acquisition is to increase market share to gain pricing power and/or economies of scale. To understand pricing power we need to understand the industry structure. To understand economics of scale we need to understand the cost drivers.

Pricing power. The so called over-the-top instant messaging market (e.g. Skype, Whatsapp, Viber, WeChat, Line, Kakao, Facebook Messenger) is growing fast in terms of volume. All of these services are currently free. Here is an analysis of the industry structure
  • Suppliers bargaining power is weak. There are several companies offering web hosting services alternatively a server farm can be created by using generic hardware and software. Programming skills are offered by thousands of people. Internet service providers might in the future charge for access. It will be technically difficult to single out low-bandwidth instant messaging, but possible.
  • Buyer bargaining power is medium. There instant messaging product is around 20 years old, but it is still a growth business. Buyers do not have any individual power but the industry has made buyers price-sensitive by offering the service for free. There are very few switching costs since the applications automatically have access to the telephone contacts in the smartphone's address book. Without this feature there would be strong network effects leading to a winner take it all situation. However, there will still be a certain degree of positive network effect.
  • Substitute power is weak. There are no new substitutes.
  • Threat of new entry is high. There are very few entry barriers. A new entrant can easily enter the world market by hiring 10 engineers and buy space from a web hosting company. There are network economics at play, but they are limited as the applications are using a lot of functionality from the operating system layer (i.e. it is not a major nuisance to simultaneously use two-three service providers, alternatively to set up another service provider)
  • Rivalry is high. The current price for the service is $0 (Whatsapp charges $1) so price competition cannot get any worse. There are a handful of companies worldwide with a sizeable market share.
Once the industry matures, it will be possible for the established companies to start charging for the service. Given the low marginal cost of providing instant messaging, we will see intense competition. The ease of entry into the industry will ensure that profit margins will stay low. This is a contestable market using terminology from economics (Baumol, 1982). I think the end-result will be that instant messaging is bundled together with another product, i.e. a separate instant messaging service is unlikely to remain viable. Facebook Messenger is for instance an integral part of the advertising supported Facebook website. Skype's instant messaging is supported by revenue from voice telephony services.

One argument in favour of pricing power is network externalities. People want to be on Facebook because their friends are already on Facebook. There is a lock-in to a specific platform that is difficult to get around. Google Plus is facing this problem. I do not think this is a valid argument for instant messaging on smartphones. Prior generations of instant messaging services (e.g. ICQ, Yahoo Messenger, Skype) required the user to set up a separate account and add contacts. The current generation of instant messaging services utilise the contacts already present in the smartphone so it is very easy to a user to use more than one instant messaging service simultaneously or switch to another service.

Economies of scale. Whatsapp's accounts are not public, but it is likely that their cost of operations is much lower than $1 per user. The economies of scale in operations are likely to be small. More computing power is needed to manage more users. There will be economies of scale in R&D and marketing, but this is likely to be a small portion of total cost.


Conclusion. To justify the purchase price based on a subscription fee, Facebook needs to charge all daily users of the service $15 per year in perpetuity (71% daily users of 450M monthly users). Given the industry dynamics and a cost below $1, this is not a realistic assumption. There could very well be a niche market willing to pay this amount for a non-advertising based service, but that is not relevant for Facebook.



(3) Uncertainty about Whatsapp as a competitor

Facebook is dependent on mobile online advertising and users are likely to spend a fair amount of time on instant messaging. The Asian competitors (i.e. Kakao, Line, WeChat) have started to build social network like features and Facebook is likely to have been afraid what an independent Whatsapp would do. From this perspective it would be even more dangerous if Google were to acquire Whatsapp. Google's market share in mobile online advertising is much smaller than its market share in search online advertising. The rumour that Google was involved in a bidding war is probably correct. Acquiring Whatsapp can be seen as an insurance policy.



Conclusion on intrinsic value

I started out by noting that the success of Facebook will be determined by how well it does in the online advertising market. The acquisition of Whatsapp is just one step in the direction of making the overall strategy successful.

The initial Models A to C show that Facebook is highly valued. Without going into much detail I would like to note a major problem with the discounted cash flow analyses described above. The models assume that Facebook will grow organically. If it is necessary for Facebook to acquire companies to grow, an additional investment component needs to be added to the models. Facebook has now acquired Instagram and Whatsapp. I do not know whether this is the beginning of a string of acquisitions during the coming decades or whether they are one-time preempting acquisitions in key product categories. Looking at the instant messaging market, there has been several generations of products. Earlier generations included ICQ, Yahoo Messenger, Windows Messenger, and Skype. Non of these products have been able to stay relevant for long. Skype missed the smartphone market most recently. My guesstimate is that Facebook will have to continue to acquiring new products.

The current valuation of Facebook is high. However, Facebook should remain on track as long as the following benchmarks are being reached in Model C:
  • Growth in total advertising revenue should stay above 5%, after inflation.
  • Share of advertising revenue going online should increase with 1 percentage point per year, until it reaches 50%.
  • Facebook's market share of online advertising revenue should increase from 7% to 10% in the next three years.
All three factors taken together are equivalent to a 10% increase in Facebook's revenue per year (I assume profit margin remains at Q4 2013 forever). However, Facebook needs to grow revenue with this amount per year over several decades and that can only be accomplished with the three requirements largely being fulfilled independently. There is some give-and-take between the components (e.g. smaller growth rate, but higher market share).



Sentiment 

The share price is a function of intrinsic value (discussed above) as well as sentiment, which is driven by emotions. When Facebook was introduced in 2012 sentiment was negative. The sentiment began to change in January 2013 when short sellers gave up (see figure 1). Following strong earnings reports, sentiment turned more positive in the second half of 2013. Sentiment has remained positive as evidenced by a three day cumulative abnormal return of 2.8% (adjusted for index, but not beta) after the acquisition of Whatsapp.


Figure 1. Share price of Facebook and short interest as percent of outstanding shares. The short interest for a normal stock falls between 1% and 15%, but levels as high as 50% are quite possible. Source: Capitaliq.



The consensus forecast among investment analysts is that Facebook's revenue will increase 24% per year until 2018. This is equivalent to reaching a market share of 15% in 2018. In other words, the expectations on Facebook's five year growth are much higher than in my intrinsic value calculation above. Irrespective of the long-term trajectory, Facebook is going to experience trouble along the way. The initial trigger of sentiment change can be a macro factor (e.g. slowed growth in global advertising revenue) or a micro factor (e.g. no growth of Facebook's market share). Both type of factors will be reflected in Facebook's quarterly earnings. 

At the time of writing Facebook's share price is 70.78.

No comments:

Post a Comment