Empirical Analysis of Corporate Growth  167 To sustain high growth, companies need to overcome this “portfolio treadmill” effect: for each product that matures and declines in revenues, the company needs to find a similar-size replacement product to stay level in revenues—and even more to continue growing. Think of the pharmaceu- tical industry, which showed unprecedented growth from the mid-1990s, thanks to so-called blockbuster drugs such as Lipitor and Celebrex. Then growth plummeted as these drugs came off patent and the next generation of drugs didn’t deliver the same outsize sales as the blockbusters. Finding sizable new sources of growth requires more experimentation and a longer time horizon than many companies are willing to invest in. Royal Philips’s health technology business was a small corporate division in 1998, when it generated around 7 percent of total company revenues. It took 15 years of ongoing investments and acquisitions to become Philips’s largest business unit, generating half of its total revenues. After the carve-out of its light- ing business and other divestitures, health technology has now become Philips’s core business. Empirical Analysis of Corporate Growth The empirical research backs up the principles we have been discussing. This section presents our findings on the level and persistence of corporate growth for U.S.-based nonfinancial companies with revenues greater than $1 billion (inflation-adjusted) from 1963 to 2017. (The sample size for each year is different but amounts to 1,095 companies in 2017.) The analysis of their revenue growth follows the same procedure as the analysis of ROIC data in Chapter 8, except here we use three-year rolling averages to moder- ate distortions caused by currency fluctuations and M&A activity. We also use real, rather than nominal, data to analyze all corporate growth results, because even mature companies saw a dramatic increase in revenues dur- ing the 1970s as inflation increased prices. Ideally, we would report sta- tistics on organic revenue growth, but current reporting standards do not require companies to disclose the effects of currencies and M&A on their revenues. The overall findings concerning revenue growth are as follows: • The median rate of revenue growth between 1965 and 2017 was 4.9 ­percent in real (inflation-adjusted) terms. Real revenue growth fluc- tuated significantly, ranging from around 0 percent to 9 percent, with significant cyclicality. • High growth rates decayed very quickly. Companies growing faster than 20 percent in real terms typically grew at only 8 percent within five years and at 5 percent within ten years. 168  Growth Growth Trends Let’s begin by examining aggregate levels and trends of corporate growth. Exhibit 9.7 presents median revenue growth rates in real terms between 1965 and 2017. The average median revenue growth rate for that period equals 4.9 percent per year and oscillates between roughly 0 percent and 9 percent. Me- dian revenue growth demonstrates no trend over time, but over the past five years, growth rates declined to around 2 percent in real terms. Real revenue growth of 4.9 percent is quite high when compared with real GDP growth in the United States, which was at 3.0 percent during the same period. Why the difference? Possible explanations abound. The first is self- selection: companies with good growth opportunities need capital to grow. Since public markets are large and liquid, high-growth companies are more likely to be publicly traded than privately held ones. We measure only pub- licly traded companies, so these growth results are likely to be higher. Second, as companies become increasingly specialized and outsource more services, firms providing services will grow and develop quickly without af- fecting the GDP figures. Consider Jabil Circuit, a contract electronics manu- facturer. When a company like Apple or IBM has Jabil manufacture products or components on its behalf, GDP, which measures aggregate output, will not change. Yet Jabil’s growth will influence our sample. A third explanation is global expansion. Many of the companies in the sample create products and generate revenue outside the United States, so they can grow faster than U.S. GDP without gaining sales in the United States. Finally, although we use rolling averages and medians, these cannot eliminate but only dampen the effects of M&A and currency fluctuations, which do not reflect organic growth. Exhibit 9.7  Long-Term Revenue Growth for Nonfinancial Companies, 1965–2017 3-year revenue growth rate, adjusted for inflation, % –10 –5 0 5 10 15 20 25 30 35 1980 1975 1970 2010 2015 2005 2000 1995 1990 1985 –15 1965 Average Median 3rd quartile 14.7 14.2 Median 5.0 4.9 1st quartile –1.8 –1.9 Source: Compustat; Corporate Performance Analytics by McKinsey.