Is Chinese government debt 20% of GDP or 150% of GDP, and does it matter? The short answer is no, not really. Chinese general government debt, which comprises central government debt (including banking sector restructuring related debt) and should comprise local government debt, amounts to 20% of GDP. China does not provide (net) public sector debt statistics, which would consolidate the balance sheets of the general government sector and “public corporations”, including the central bank, state-owned corporations and banks as well as government-owned entities such as the China Investment Corporation (CIC).
Against the backdrop of China’s bank lending stimulus following the 2008 financial crisis, a lot of attention has been focused on (off-balance sheet) local government debt. Being barred from taking out bank credit, local governments set up funding vehicles to raise bank loans. This is presumably why they do not seem to show up under in the reported general government statistics. The National Audit Office (NOA) recently put local government debt (provinces, cities, countries) at RMB 10.7 tr (USD 1.6 tr or 27% of GDP) at end-2010. This would raise general government to nearly 50% of GDP.
Moody’s believes that this underestimates local government debt by RMB 3.5 tr (USD 520 bn or 9% of GDP), as the NOA only accounts for debt directly guaranteed by local governments. Other analysts (Shih) put the amount of total local government liabilities even higher, at RMB 15-20 tr (37-45% of GDP). But LGFV carried RMB 5 tr and “business units subsidised by the budget” another RMB 5.7 tr in debt. Rmb15,400bn (Rmb9,700bn in LGFV debt plus Rmb5,700bn in non-LGFV debt) and Rmb20,100bn (Rmb14,400bn in LGFV debt plus Rmb5,700bn in non-LGFV debt). If this estimate is correct, combined local and central government debt may be as high as 70% of GDP.
This does not take into account other public sector debt such as the debt of SOEs and other entities linked to the central government. For instance, the Ministry of Railways (5% of GDP) and policy banks (10-15% of GDP) have incurred substantial debts in recent years. The major “off-balance” item, however, is related to local government borrowing. If, unrealistically, all the debt of all government-owned entities were added up, public debt may add up to 150% of GDP or so. Naturally, failing to net out intra-government holdings, this greatly overstates the public sector’s debt position (e.g. lending by policy banks to local governments).
Even if the government were forced to underwrite all outstanding domestic debt liabilities, this would not break the “sovereign” bank. This would hold true even if economic growth declined substantially, real interest rates rose and the primary fiscal balance worsened. Naturally, the actual increase in consolidated public sector debt would be substantially less once intra-government holdings and the collateral transfer associated with the potential takeover of local-government debts are taken into account. As long as China maintains a growth rate of 5%, the government will be in a position to underwrite all outstanding debt liabilities.
With the financial system as a whole implicitly guaranteed by a very creditworthy central government, the banking sector faces little risk of a systemic destabilization. Significant government ownership, potential regulatory forbearance, a captive depositor base and financial repression further help limit risk. Similar to the bank restructurings of the late 1990s and early 2000s, the government can help out the banks by removing NPLs and recapitalizing the banks. Or it can bail out the local governments (and thus directly the banks) by guaranteeing or taking over local government debts and resolving the issue with the help of strong nominal GDP and cheap financing costs due to low/ zero real interest rates.
In practice, the resolution will likely be a mix of the central government taking on part of the local government debt and the banks taking limited write-downs in order not to jeopardize their financial stability. (While the government will not allow the write-off to be destablising, it would nonetheless impact the banks’ bottom line.) Another way of looking at this is that a chunk of the investment goes bad and gets absorbed, largely, by the central government – either though recapitalizing the banks or assuming local government debts. The direct financial costs to the central government of the second option will be larger, but the economic costs may be smaller. An economy with a savings and investment rate as high as China’s can afford to do this. The more important question is whether, going forward, government-driven bank lending and government investment is the best way to generate high productivity gains and economic growth and the return on investment to finance the debt. If the marginal productivity of investment were to decline, whether as a result of poor lending decisions or a structural decline in marginal productivity growth, rising debt levels could become a problem. But, in all likelihood, this day is still a decade or so away.