Written by John Lounsbury
Jamil Anderlini, writing in the Financial Times, brought our attention to some Chinese government research about the debt in China and how it has been used for investment in that country. See China has ‘wasted’ $6.8tn in investment, warn Beijing researchers. Looking deeper into the subject, including looking at the paper Anderlini reported on (Inefficient and ineffective investment causing a huge waste by Xu Ce and Wang Yuan, Shanghai Securities News – in Chinese), we think the question of waste is more nuanced than the $6.8 trillion figure quoted would indicate.
The problem that China is dealing with starts with the unique position they occupy in the world with the level of investment much higher and consumption much lower than everyone else. The change (reversal) of this relationship is at the core of what is meant when the term “rebalancing” is applied to China.
The comparisons to other large economies shows China has gotten into a league of its own.
Note: The very high savings and investment level in China is achieved at the expense of wages (and therefore lowered consumption, which is how wages are spent). This situation is a form of what is referred to as “financial repression“. More common uses of the term applies to (1) “austerity” wherein government channels what might have gone to personal incomes into sovereign debt reduction through tax hikes or government expenditure reductions or (2) conditions of abnormally low or negative interest rates. These more common uses are processes whereas our use of the term for the situation in China refers to the end result: High government expenditures and low consumer expenditures.
The data was segregated into two time periods by Xu and Wang, shown in the next graph:
They identify a period of higher growth per unit of debt (1979-1996) and a lower growth period (1997-2013).
We suggest that the segregation into the two time periods may have been rather arbitrary. Although we have not calculated the correlation coefficients for the two medians defined above, by inspection we predict both would be weak. We suggest that another arbitrary subdivision of the data would show much better correlations, and have displayed those in the four graphs below:
One of the time periods is still weakly correlated to the trend line (1979-1994). It is likely that two better correlations would be obtained by subdividing the data into 1979-1989 (probably fair correlation with a shallow positive slope) and 1989-1994 (probably good correlation with a steep positive slope).
Bottom line, the structure of China’s marginal return on debt is much more complex than the Xu and Wang have discussed in their report. There have been periods where productivity of debt was much better (improving over time) than at other times (degrading).
In early 2009 we looked at similar data for the U.S. and found a more uniform progression toward lower marginal productivity of debt (although there was a 6-year period from 1986-1992 where the marginal return increased from about 20% per dollar of debt in 1986 to nearly 50% in 1992). Otherwise the fluctuations (each data point is one calendar quarter of data) were less than 2 years in duration.
Note (U.S. data): This 2009 work needs to be updated. Others who looked at this data at that time and since have used linear curve fitting (nearly as good correlation as our quadratic fit). It would be worthwhile to determine if the additional five years have diminished the correlation of the quadratic fit. It has also been our contention that the analysis should be attempted with a segregation of public debt and private debt; this has never been started.
Note (China data): Much of the “wasted debt” in China has been for under utilized infrastructure, excess industrial capacity and unoccupied housing. How can it be determined if this is wasted? Because it hasn’t been fully utilized to date? By such a measure (over a few years) how can a final determination be made? Will these still be underutilized in 10 years? In 20 years? If China rebalances with an increase in consumption as is desired and if GDP increases by an average of 3% a year, the needs of the country will be doubled over the next 24 years. How much of this waste will actually end up recycled into productive use? By applying short-term assessments wouldn’t the Suez Canal, the Panama Canal, the Erie Canal, the western U.S. railroads, the U.S interstate highway system, etc. have been “wasted debt”?
Bottom line: Are we trying to evaluate a book after only reading the introduction?