China's Unsustainable Investment Boom
The country has to boost household consumption to maintain heady GDP growth.
The Chinese fixed-asset investment boom of the past decade has been unprecedented. While all low-income countries have required outsize capital stock additions to make the leap to middle-income status, China's current boom is unmatched by anything on the record books. In fact, by some measures of physical capital, China already looks more like one of the world's leading developed economies, rather than the middle-income economy it is. We don't believe China can continue to rely on building more skyscrapers, highways, and manufacturing plants to sustain the kind of GDP numbers its citizens and global investors have grown accustomed to. In the next 10 years, the onus for growth will rest on Chinese households: their willingness and ability to consume. If consumption fails to grow at a rate well above historical norms, the economy may be able to muster only 5% growth at best, a far cry from the 10% average from 2001 to 2010.
In 2000, before the boom really kicked off, gross capital formation (or GCF, the GDP accounting term for investments in physical capital) accounted for 35% of Chinese economic output. While large by developed economy standards (the U.S. 10-year average is 19%, Japan 23%), it wasn't atypical for a high-growth emerging economy. Nor was it unusual for China: A decade prior, GCF also had a 35% share. The consistent share reflected the balanced growth China enjoyed in the 1990s. By 2000, the Chinese economy was 170% larger than it was in 1990, driven by a 173% increase in investment and a 156% increase in consumption.
Chinese GDP expanded at a similarly impressive rate in the 2000s. By 2010, the Chinese economy was 171% larger in 2010 than it was in 2000. But the sources of growth were anything but balanced. A massive share came from a surge in fixed-asset investment. By 2010, China was spending 273% more on physical capital than it was in 2000, with GCF accounting for a 49% share of output by decade's end. Yet Chinese households were "only" consuming 97% more than they were in 2000, depressing the share of household consumption of total GDP to 34% by 2010, a paltry share by any standard and the lowest level seen in China since the founding of the People's Republic.
The extreme imbalance between investment and consumption appears even more exceptional in the context of the three biggest investment-led success stories of the past 50 years: Japan, Korea, and Taiwan. By any measure of fixed-asset investment intensity--growth rates, share of cumulative GDP growth, or share of GDP--China has far surpassed the precedents set by Japan, Korea, and Taiwan. We estimate that, relative to the starting size of the economy, cumulative additions to Chinese capital stock in its boom decade have been 43% greater than Japan's, 33% greater than Korea's, and 49% greater than Taiwan's.
Just as striking is the relatively feeble contribution of consumption to Chinese growth rates. While GCF grew faster than consumption in the boom decades of Japan, Korea, and Taiwan, consumption still accounted for at least half of total economic expansion in each case, versus the paltry 26% share we see for China. While the investment share of Chinese GDP has exceeded the consumption share in every year since 2003, this was not achieved in any year by Japan, Korea, or Taiwan.
Various hard measures of GCF confirm the outsize role of fixed-asset investment in the Chinese growth story. The Chinese economy is roughly 170% larger than it was a decade ago, but it now consumes 383% more aluminum and 393% more steel. Total expressway mileage is up 354% over the past decade, the number of tunnels is up 338%, and floor space under construction, a concrete measure of real estate activity, is up 337%.
One might argue that this rapid buildout of Chinese real estate and infrastructure makes sense to the extent it correctly anticipates a commensurately rapid increase in consumption. Even if a good deal of the new floor space goes unoccupied at the moment, continued urbanization and rising incomes will eventually fill it. And while some of the new expressways, bridges, and tunnels may see only a trickle of traffic today, rising automobile ownership rates will ultimately generate a steady stream of vehicles. From this perspective, today's investment is nothing more than a down payment on tomorrow's consumption.
Consumption Has a Lot of Catching Up to Do
But given the immense investments made in the past decade, consumption has a lot of catching up to do. Even if the massive capital additions have correctly anticipated the consumption growth we'll see in the decades to come, the economic rationale for further outsize capital outlays grows increasingly weak with each passing year. Any nation, even a rapidly growing one, needs only so many airports, highways, high-speed rail lines, and luxury apartments.
By some measures, China's physical capital base already looks like that of a major developed economy. Consider the installed capacity of China's steel industry. Now roughly 5 times the size of what it was a decade ago, capacity is nearly twice that of the United States, Japan, and the European Union combined. Notably, the latter collection of economies has nearly 1 billion people of its own and collective GDP of about $36.6 trillion, making it more than 6 times the size of China's economy.
China's expressway system, the national trunk highway system, also looks rather overbuilt. By year-end, the NTHS will have quintupled its 2000 length. Heading into the year, the total length of the NTHS (45,554 miles) was already on par with that of the interstate highway system (46,876 miles) in the U.S., a country of similar size but with 3 times as many cars on the road. While growing Chinese vehicle ownership rates are likely to trim some of the bloat in the next couple of decades, the rationale for additions comparable to what we've seen in the past 10 years is very limited.
The biggest contributor to China's fixed-asset investment boom, particularly in the past few years, has been residential real estate. Throughout the 2000s, China built housing at a blistering pace, adding a cumulative 120 square feet in residential floor space per person. This is understandable, given the significant additions China made to its urban population over the period. What is less understandable is the roughly 80% surge in the rate of floor space additions we've seen in the past few years, which has not been accompanied by a comparable surge in urbanization. On a per capita basis, China now has nearly 5 times the amount of residential floor space under construction as the U.S. in its peak housing boom. This is particularly remarkable since, despite enormous gains in wealth and income, Chinese remain on average much poorer than their American counterparts and tend to occupy residences that are much smaller.
Perhaps the biggest counterargument to the overbuilding thesis goes as follows: Despite its massive urbanization of the past couple of decades, China remains relatively rural by global standards and will continue to require large additions to its capital stock as it accommodates new urbanites. According to Chinese government figures, even after an influx of 207 million new urban residents in the past decade, only 50% of the population resides in urban areas. An increase to 70% urban--a level typical of the high-middle-income status to which China aspires--would add 272 million to China's urban total. That's equivalent to adding 33 cities the size of New York.
While that notion is intuitively powerful, we'd strongly caution investors against taking the data underpinning the "stronger for longer" urbanization story at face value. Countries use very different definitions of what constitutes urban. As a result, relying on headline data alone can lead to ill-informed conclusions. China's self-reported urban share of 50% is equivalent to reported figures from Ghana (50.7%) and apparently well below that of North Korea (60.1%), which remains largely a subsistence agriculture economy.
As it turns out, China's definition of urban is stricter than most, effectively portraying the country as more rural than it might otherwise appear and potentially overstating the remaining runway for further urbanization. China's statistics bureau generally requires a density of 1,500 persons per square kilometer for a population to be deemed urban. By this hurdle rate, 4 of the top 10 largest U.S. cities would fail to meet China's definition of urban, not to mention hundreds of suburbs. While the data necessary to reformulate China's urban/rural breakdown on a more apples-to-apples basis with that of the U.S. aren't made available, it seems fair to assume that such an undertaking might add at least 10 percentage points to China's stated urbanization level, significantly curtailing the urbanization upside promoted by China bulls.
All told, we expect to see significantly lower GCF growth in the coming decade than we did in the decade just past. As a result, for China to sustain robust economic growth in the coming decade, household consumption will need to grow at above-trend rates, rebalancing its economy toward a more normal composition of consumption and investment. Chinese policymakers are not blind to the need to increase consumption, as evidenced by the recently articulated objective of boosting consumption's share of GDP to 50% by the end of the decade.
Two Ways to Rebalance China's Economy
There are two ways to achieve this rebalancing: stronger growth in consumer spending or weaker growth in fixed-asset investment. History suggests the latter outcome is more likely. Consider the average GCF growth turned in by Japan, Korea, and Taiwan in the decade following each country's investment boom: 1.2%, 2.4%, and 5.9%, respectively. If China were to achieve the average of those three (about 3%) in its own post-boom decade and household consumption continued to expand at the rate it has in the past decade (7.0%), total GDP would grow at 4.8%, well below the 10.5% average of the 2000s. Even if consumption were to expand at 12.0% annually, a level it achieved not once in the past decade, GDP would grow at 7.5%--stellar by developed market standards, but somewhat below what Chinese citizens and global investors have come to expect.
In the case of Japan, Korea, and Taiwan, GCF growth post-boom was inversely related to the size of the boom. Japan and Korea had larger booms (GCF averaged a 37% share of GDP in the boom decade), leading to weaker GCF growth post-boom. Taiwan had a relatively smaller boom (31% share of GDP) and saw fairly robust GCF growth post-boom. With this in mind, it seems reasonable to believe that, since China's investment boom has been both longer and stronger than its antecedents, its post-boom decade will require more modest additions to the capital stock than the cases of Japan, Korea, or Taiwan. This suggests even 3.0% GCF growth would be a fairly rosy outcome by historical standards. Assuming instead 1.0% GCF growth and 7.0% consumption growth would put GDP growth at 3.8%--a potentially troubling outcome not only for China, but for the global economy as a whole.