There is no denying that internet companies are the stars of the market today. Table 13.1 lists the top ten global companies by market value in Spring 2018. Two facts stand out. First, eight out of ten are American companies, and the other two are Chinese. Europe and Japan are entirely missing (they are included in the data, they just don’t make the cut). Second, the top six are internet technology companies.
TABLE 13.1
Top Ten Global Firms, Spring 2018
Company
Country
Market value ($ billion)
Apple
US
926.9
Amazon
US
777.8
Alphabet
US
766.4
Microsoft
US
750.6
Facebook
US
541.5
Alibaba
China
499.4
Berkshire Hathaway
US
491.9
Tencent Holdings
China
491.3
JPMorgan Chase
US
387.7
ExxonMobil
US
344.1
These companies are stars, undoubtedly. But there have always been stars in the economy. Are these stars different?
Carmen Reinhart and Kenneth Rogoff (2009) have famously shown that thinking “this time is different” is the shortest way to a financial crisis. In macroeconomics, there is no such thing as “this time is different.” But, perhaps, matters could be different where the internet is concerned. There are some technological reasons to believe this time might be different. Internet firms can grow very quickly. It took Snapchat only eighteen months to reach the $1 billion valuation that it took Google eight full years to achieve, a feat that, on average, takes twenty years for a Fortune 500 company. Digital data can be used in ways that data stored on paper cannot. Learning models trained on very large, fine-grained data sets go further than traditional learning models. There is the potential for more knowledge creation, although here we need to recognize the fundamental difference between data and knowledge. A billion tweets are a lot of data but not necessarily a lot of knowledge.
I have a slightly more mundane take on the topic. I have noticed that people who are the most convinced that the GAFAMs are different are usually the people who know the least about these companies. Most of them repeat what they hear without taking the time to look at the data. Conversely, the more people know about these companies, the more they tend to describe them using relatively standard business concepts.
Understanding the GAFAMs matters for many reasons, but in particular when we seek to understand market concentration. These companies are very successful and innovative, of course, and they also control large shares of their domestic markets. I think it is hard to look at the growing concentration in US airlines, telecoms, and health care without thinking that these are negative trends for consumers and the economy as a whole. It is less clear how we should think about growing concentration in the GAFAMs’ markets. Perhaps it’s efficient, much like the Walmart-driven concentration in retail. Perhaps it’s not fully efficient but a necessary cost of having large internet companies. At the very least, we should distinguish concentration as a result of the stars taking over the market from concentration driven by inefficient incumbents.
We thus need to understand whether these stars are different from those of the past. And more importantly, we need to understand if these star companies are going to create stellar performance in the economy as a whole.
In the next chapter, we will look at the lobbying and political influence of the GAFAMs. But before delving into their relatively new effort to flex their muscles in Washington, it’s worth looking closely at them as businesses—specifically to interrogate two assumptions that allowed them to escape regulation for so long: first, that they are a special breed of company that mustn’t be hindered by regulation, and second, that they are integral to the health of the US economy, and must therefore be treated with kid gloves.
The only way I know how to do that is by looking at the data. So, let us dig in.
The Business Models of the Stars
Let me start by discussing briefly the business models of these companies. Apple is a luxury manufacturing company. Google and Facebook are online advertising companies. Amazon is a marketplace and a retailer with a cloud service. Microsoft is a bit more diversified.
Apple sells iPhones. iPhones are desirable because they work well and because they are beautiful. They have become a status symbol, like Chanel bags or Hermès scarves. Apple also sells tablets and computers. Together, these three items account for 84 percent of Apple’s revenues. But Apple is a luxury brand. It makes money with high margins, not just high volumes. The Korean company Samsung sells more smart phones than Apple does: its market share is 27 percent versus 24 percent for Apple, but Samsung’s phones are relatively cheap.
An iPhone is to a smart phone what the Mercedes is to a car, except that Apple has a 24 percent market share, while Mercedes’s market share is ten times smaller. Mercedes-Benz sold approximately 2.3 million cars worldwide in 2017 (a record year for the company). That’s 2–3 percent of the global car market. Its revenues were about $120 billion and its profits about $13.5 billion, so its margin was around 10 percent. Apple sells about 220 million iPhones each year. Its consolidated revenues were about $230 billion in 2017, with $150 billion coming from iPhone sales alone. Its profits were around $50 billion, so its margin was above 20 percent. Its market value was more than ten times that of Mercedes. Apple, then, is a special kind of manufacturing company: it does luxury at scale. Another critical difference with other luxury manufacturing companies is that Apple develops a lot of software and services (iTunes, Apple Music, the App Store) that play a strategic role to support its manufacturing revenues. These services and software are used to attract and retain customers, and to create effective barriers to entry.
Google and Facebook earn most of their revenues from advertising: 88 percent for Google and 97 percent for Facebook. Their market shares are large. Together they collect about two-thirds of digital advertising expenditures. The ways in which they attract customers are different, however. Google helps people find stuff online. As of 2018, Google processes about 40,000 search queries per second worldwide, or 3.5 billion per day, or, if you prefer, 1.2 trillion searches per year. Google processes about two-thirds of global online searches, with Bing and Yahoo providing much of the rest. For internet searches initiated from mobile phones, Google’s share is above 90 percent. Google does not get paid when you do a search. It only gets paid if you click on a paid link. You can immediately see why advertisers like this model: they only pay if you click. Moreover, their ads do not appear randomly but rather in response to your search, so there is a higher chance that you are actually interested in what they have to offer.
Facebook provides content that grabs your attention and then shows you a bunch of ads. Facebook’s model is not so different from that of older media companies, newspapers, radios, and television networks. The main difference is in content and cost structure. Facebook does not produce any content. The content is you, your pictures and your videos, and your friends’ pictures and videos. In a sense, you buy from yourself and Facebook takes a cut. That is a pretty smart business model. As CEO Mark Zuckerberg stated, “We provide the social technology. They provide the music.” The other difference is that Facebook knows a lot about you, and your friends, and your family. As a result, Facebook can target its ads very precisely. That’s what advertisers like. It’s also what creates political risks for Facebook.
Microsoft has been around for longer than Google and Facebook. Perhaps for this reason, it also has a more diversified revenue stream. It derives significant income from Office, Windows, and Xbox, but also from its cloud services (Azure). Microsoft was the star of the 1990s. Its share price r
eached $55 in 2000. In 2009 it was down to $20, then increased to $25 in 2011. But it made a series of good business decisions, such as expanding its cloud business, and is now once again solidly among the stars. One important difference is that Microsoft is not a social media company. This could be considered a weakness, but it also reduces Microsoft’s exposure to risks in that sector.
Amazon is an online retailer with a media business and a cloud service business. It is also a marketplace. In some respects, Amazon is quite different from the other four companies. Amazon has many employees, including blue-collar workers, and it does a significant amount of tangible investment. Amazon still has fewer employees than Walmart, but that’s just because Walmart is so large. Amazon accounts for most of the growth in capital and research & development expenditures in the US retail sector over the past seven years. It is important to keep this fact in mind. We have discussed investment in Chapter 5. We have seen that investment and productivity growth have been weak, and that large and profitable businesses are largely responsible for the low ratio of investment to profits. Retail is an exception, as discussed in Chapter 2. By most accounts, it has remained quite competitive and productive.
Stars of Today, Stars of Yesterday
How do the GAFAMs compare to the stars of previous decades? We hear all the time that the market value of these companies is unprecedented. Apple was indeed the first company to pass the $1 trillion mark. But that’s comparing Apples and Oranges (pun intended). These are current dollars, and in a booming stock market.
Is Apple’s market value really unprecedented? No.
Table 13.2 contains the information we need to compare the GAFAMs to the previous cohort of stars. If you look at the data, the top five firms have typically accounted for about 10 percent of the total market value of US equities. In the 1960s, AT&T alone represented more than 6 percent of the market. Apple today is less than 3 percent. We are going to see that the GAFAMs are indeed different, but not in the way people think: they are … small!
Table 13.2 contains a lot of information, so let’s unpack it slowly. It highlights some key facts regarding the star companies of each decade. We start in 1950, because this is when the Compustat database starts. For each decade, we compute the average market value of each firm, and we select the top five firms for the table (the top twenty firms for the figures below). Each column contains different pieces of information.
The rank is the rank by market value of equity. General Motors was the second largest firm by market value in the 1950s. At the time, the number of publicly traded companies was quite small, and GM accounted for 5.81 percent of total market capitalization. It employed 0.89 percent of civilian employment. GM was also buying a lot of inputs from other firms in the economy. On average its cost of goods sold (COGS in the table) was 1.22 percent of US GDP. Cost of goods sold is an accounting item that includes the intermediate inputs (car parts, steel, energy) as well as the wages of production workers. GM production wages were about 0.5 percent of US GDP and it bought approximately 0.72 percent of GDP in inputs. That means that GM was deeply integrated in the US economy.
At the bottom of each decade, I show averages or sums for the top five. The average operating profit margin (Op. Inc. / Sales in the table) of the stars of the 1950s was 20 percent. They paid 51.7 percent of their operating income in taxes (Taxes / Op. Inc.). Together, they accounted for 27.95 percent of the market value of listed companies, 2.59 percent of civilian employment (Emp), sourced inputs, and production labor for 3.04 percent of GDP (COGS / GDP).
TABLE 13.2
Seven Decades of Stars
Profitability (%)
MV / Emp
ratio
Share of the Economy (%)
Decade
Rank
Company
Op. Inc. / Sales
Taxes / Op. Inc.
MV share
Emp share
COGS / GDP
1950s
1
AT&T
24.9
45.6
7.3
7.01
0.957
0.62
2
General Motors
16.9
57.2
7.5
6.71
0.891
1.22
3
ExxonMobil
16.8
38.2
24.7
5.70
0.231
0.57
4
Dupont
28.7
59.7
39.0
5.55
0.142
0.16
5
General Electric
12.7
57.9
8.0
2.98
0.373
0.47
Average
20.0
51.7
10.8
Tot.
27.95
2.595
3.04
1960s
1
AT&T
30.9
44.6
7.4
6.40
0.869
0.56
2
IBM
25.3
53.1
19.1
4.08
0.213
0.12
3
General Motors
16.3
51.9
4.5
4.25
0.952
1.25
4
ExxonMobil
13.5
43.0
14.5
2.98
0.206
0.69
5
Texaco
12.9
23.3
20.9
1.88
0.090
0.25
Average
19.8
43.2
8.4
Tot.
19.59
2.330
2.86
1970s
1
IBM
24.6
50.3
14.1
4.66
0.330
0.18
2
AT&T
25.5
35.0
4.4
3.91
0.894
0.69
3
ExxonMobil
17.5
66.6
15.6
2.46
0.158
1.03
4
General Motors
9.2
46.4
2.5
2.20
0.873
1.31
5
Eastman Kodak
24.1
47.5
12.6
1.72
0.137
0.10
Average
20.2
49.2
6.3
Tot.
14.95
2.391
3.30
1980s
1
IBM
19.6
42.6
9.4
3.31
0.354
0.31
2
ExxonMobil
9.8
44.5
15.8
2.08
0.132
1.14
3
AT&T
12.8
18.7
4.4
2.10
0.472
0.85
4
General Electric
11.5
33.5
4.6
1.48
0.320
0.42
5
General Motors
4.3
11.3
1.5
1.05
0.710
1.21
Average
11.6
30.1
5.0
Tot.
10.03
1.987
3.94
1990s
1
General Electric
22.5
17.4
10.1
2.12
0.209
0.49
2
Microsoft
39.0
35.5
93.6
1.28
0.014
0.01
3
ExxonMobil
7.7
38.1
23.9
1.71
0.072
0.67
4
Walmart
5.0
39.4
2.5
1.27
0.517
0.80
5
Coca-Cola
23.1
31.7
55.2
1.34
0.024
0.05
Average
19.5
32.4
9.2
Tot.
7.73
0.836
2.02
2000s
1
ExxonMobil
13.0
48.2
41.1
2.51
0.061
0.88
2
General Electric
23.8
10.3
10.5
2.35
0.223
0.44
3
Microsoft
40.7
31.6
44.8
2.05
0.046
0.03
4
Walmart
5.1
36.0
1.3
1.63
1.223
1.52
5
Pfizer
32.0
16.3
20.5
1.47
0.072
0.02
Average
22.9
28.5
6.2
The Great Reversal Page 26