China is reverting to the mean, predictably so

A photo of one of China's ghost cities, from the Daily Mail.

Lately, China's economics growth figures have disappointed to the downside. To many observers, this wasn't a surprise. A reversion to the mean was bound to happen, as described by Lant Pritchett and Lawrence Summers in Asiaphoria Meet Regerssion to the Mean:

However, predicting future levels of output by the extrapolating the current growth rate into even the near, much less the distant, future is on very weak scientific footing. The single most robust and striking fact about cross-national growth rates is regression to the mean. There is very little persistence in country growth rates over time and hence current growth has very little predictive power for future growth. Hence, while it might be the case that China continues for another two decades at 9 percent per capita growth, given the regression to the mean present in the cross-national data this would be a 3 standard error anomaly. - (this and following emphasis mine)

So it is not at all surprising that China failed to achieve its 7,5% target rate of growth. Indeed GDP growth in China has been falling for several years. It is surprising that China managed to keep such a high rate of growth for so long*.

China has had growth rates of over 6 ppa for 33 years starting in 1977 and this data set ends in 2010. - (that's 37 years if the figures were updated to include 2010-14)

But rare is this so-called super-rapid growth?

First, episodes of super-rapid (>6 ppa) growth tend to be extremely short-lived. The method of dating growth episodes doesn’t allow an episode of less than 8 years. The median duration of a super-rapid growth episode is 9 years, only one year longer than its possible minimum. There are essentially only two countries with episodes even approaching China’s current duration. Taiwan had a growth episode from 1962 to 1994 of 6.77 ppa (decelerating to 3.48 from 1992 to 2010). Korea had an acceleration in 1962 to 1982 followed by an acceleration in 1982 until 1991 when growth decelerated to 4.48 ppa so a total of 29 years of super-rapid growth (followed by rapid growth).  So China already holds the distinction of being the only country, quite possibly in the history of mankind but certainly in the data, to have sustained an episode of super-rapid growth for more than 32 years.

Reversion to the mean, by definition, means that the longer something deviates from the mean, the higher the probability of it reverting back to the mean. Betting against China hasn't been a profitable bet for the last 30 years*, but it might change. How far down might a reversion to the mean take China's GDP growth? Based on historical figures, the authors offer this:

Third, the typical (median) end of an episode of super-rapid growth is near complete regression to the mean. The median growth of the episode that follows an episode of super-rapid growth is 2.1 ppa. So the “unconditional” expectation (or central tendency) of what will happen following an episode of rapid growth is a reversion to not just somewhat slower growth but massive deceleration of 4.65 points (more than twice the cross-national standard deviation).  A deceleration of that magnitude takes growth in India from 6.29 to 1.64 and in China from 8.63 to 3.98.   

A slowdown in overall growth is bound to affect some sectors more than others. Investment (think housing, infrastructure and manufacturing capacity) in particular will take a beating. Investment as a share of the economy has reached 50% and most of it was financed with debt. The overall debt-to-GDP ratio went from 153% in 2008 to 243% in 2014 (see this and this). Debt has grown faster than debt servicing capacity and this is bound to end up in tears for investors, as it already has for investors in some of China's ghost cities. Empty apartment buildings, nearly empty high-speed trains and shopping centers only mean one thing - no revenue, thus nothing to service debts with. China's banking system is intoxicated with bad debt and up to a point it was manageable by simply refinancing the old loans. Every speculative bubble finds its limit and looks like China has met its limit at the 240% debt-to-GDP mark. 

Without substantial growth in I, G or NX, the growth has to come from somewhere else. More on that in my next post.

*There is evidence that China has been manipulating inflation, income and spending statistics, see this and this. But even if we assume a quarter or even half of the GDP growth is fake, the China's growth is still unprecedented. Some have pointed to the quality of leadership in China, or the extensive industrialization (China was a rural economy before the 70s), or the size of China which allowed them maintain a high rate of growth by pumping money into less-developed regions as wealthier regions slowed down, or the quality and efficiency improvements of their manufacturing. A plethora of explanations have been offered, but the persistence of super-rapid growth remains a mystery.