CHINA’s Bad Loans: Fudge in the Pudding
In the imprecise science of divining China’s banking health, the fine print is sometimes worth a closer look than the final statistic.
This week, the McKinsey Global Institute published a new estimate of China’s nonperforming loan ratio, measuring it at 7% as of last year. The indicator shows bad debt as a percentage of the total loan book, and is the most basic brush stroke in an analysis of China’s economic state.
McKinsey’s bad-loan estimate sits almost squarely between the Chinese government’s own projection—1.67% at the end of 2015.
At the Chinese end of things, it isn’t hard to understand why the figure seems low. Chinese banks have lots of discretion in deciding what is a bad loan.
The global norm recognizes nonperforming loans as ones that are more than 90 days past due. China doesn’t. Commercial banks don’t even bother to regularly publish that 90-day indicator in their earnings reports.
Instead, Chinese banks deem a loan nonperforming only if they think it might incur a loss. This causes consternation among those who prefer less fudge in the pudding, as it leaves out masses of past-90-day loans—particularly those to state-owned companies—that the banks deem sound as long as their deadlines are extended.
The disparity between McKinsey’s 7% and the others’ roughly 15% estimate is intriguing, given that the methodologies appear similar. So how did they end up with such different results?
For its analysis, McKinsey said it looked at 2,300 Chinese public companies in seven major sectors, including manufacturing, mining and real estate. It then examined debt loads at these companies, homing in on the debt-to-Ebitda ratio as a means to judge a company’s ability to cover its leverage. Ebitda means earnings before interest, taxes, depreciation and amortization. McKinsey then adds to the estimate how much banks may be exposed to losses on the higher-risk investment products they sell that are tied to trusts, guarantee firms and other shadow-banking institutions.
CLSA used much the same method, described in great detail in its May report, but came up with a current ratio of 15% to 19%. It examined data from China’s domestically listed companies—there are about 2,800—as of the end of 2015, estimating whether their cash flow or operating profit can cover interest expense. It added some tweaks to smooth out indicators, such as averaging Ebitda over a number of years, rather than use single data points that could be particularly volatile in isolated years.
The IMF, using largely similar methods, reported a 15.5% NPL rate in April Global Financial Stability Report.
McKinsey’s senior fellow and analyst Seong Jeongmin said the disparity is “largely driven by assumptions—how to categorize companies at risk, how much will be lost and then recovered.” He said the McKinsey assessment also relied on “perspectives based on our own knowledge on different sectors and interaction with industry experts.”
