The results of China's audit of local government debt came in well below many estimates.
While it is impossible to tell whether the audit covered all possible local government liabilities, the results make one thing clear--it is becoming much more difficult to identify what qualifies as local government debt, and who is ultimately responsible for the rapidly rising debt levels in China's economy.
Between the audits in 2010 and June 2013, the avenues through which local governments raise debt have diversified, the types of institutions borrowing on behalf of local governments have increased, and the degree to which local governments are explicitly liable for debt raised by third parties has become significantly more fuzzy.
Firstly--and no surprises here--bank loans have become a less important funding source for local governments. They accounted for 79% of all local government debt at the end of 2010, but only 23% of the 7.2 trillion yuan ($1.19 trillion) of new debt added between the two audits. At the end of June, local governments had a total of CNY17.9 trillion in outstanding liabilities.
In 2010, the government instructed banks to pare back their lending to local government financing vehicles-which local governments set up to skirt limits on borrowing directly from banks. That resulted in a migration of funding to shadow-banking sources. Shadow banks--trust, securities, insurance and leasing companies, and other non-bank financial institutions--accounted for 27.8% of new debt added between the two audits. They didn't rate a mention in the 2010 audit.
The other major increase in funding came from arrangements by local governments to pay later for goods and services, effectively pushing the cost of financing projects onto developers and contractors. Such IOUs--which didn't appear in the 2010 audit, either--accounted for 37.8% of liabilities local governments accrued between 2010 and the end of June.
Another shift is that local government finance vehicles are no longer the primary avenue through which local governments accumulate debt. Of all the types of local government institutions that can borrow money, these vehicles carry more debt than any other: 39% of the total at the end of June, down from 46.4% at the end of 2010.
But in terms of new debt added between the two audits, local government finance vehicles were responsible for only 27.9%. The biggest mover on that front was state-owned enterprises, or 'wholly state-owned firms and state-controlled firms,' in the wording of the audit. They accounted for 43.7% of new funds raised.
State-owned enterprises, or SOEs, didn't even make an appearance in the 2010 audit. According to a document issued by the National Audit Office soon after publishing the results, SOEs weren't a significant avenue through which local governments raised debt in 2010. What appears to have happened is that SOEs are taking responsibility for building projects mandated by local governments, and hence any debt they raise to complete such projects is ultimately backstopped by local authorities.
In a written response to questions from The Wall Street Journal, the National Audit Office said the rise of SOEs as a funding channel is explicitly linked with another major shift: local governments have moved from explicitly guaranteeing debt (which they're not supposed to do) to extending implied guarantees. In other words, SOEs raise debt on local governments' behalf, but there is no formal relationship that obliges local authorities to cover that debt.
Both the 2013 and 2010 audits divide local government debt into three categories. Debt the local government is directly responsible for repaying and debt guaranteed by the local government are common to both audits. In the 2013 audit, however, the third category is debt the local government 'could possibly' be expected to bailout. The 2010 audit simply referred to the third category as 'other.'
In the period between audits, there was a significant increase in the third category, and whatever we take 'other' to mean, it represents the growing importance of this fuzzy kind of guarantee. It accounted for 15.6% of local government debt at the end of 2010, but 37.2% of debt added between the audits. Most of the gain came at the expense of explicitly guaranteed debt, which accounted for 21.8% of debt in the 2010 audit but only 4.6% of the increase between audits.
The lingering question the audit doesn't answer is to what degree banks are on the hook for all this debt. On face value, it seems they have greatly moderated their pace of lending to the sector, but that might not be the case. Banks have been big buyers of investment products packaged by trusts and securities companies, much of which goes toward local governments. And how do companies with IOUs from local governments fund their projects up front? Quite possibly with bank loans.
China's financial system has become much more complex in recent years. But rather than a sign of maturity, the complexity is increasingly serving to obscure the risks.
在对《华尔街日报》(The Wall Street Journal)问题所作的书面答覆中，审计署说，国有企业充当融资渠道的情况越来越多，显然与另一重大变动有关系：地方政府已经从“明确负有担保责任”（实际上不应该如此）转变为“可能负有担保责任”。换句话说，由国有企业代表地方政府进行举债，但并不存在一种“地方政府有义务承担负债”的正式关系。