China Debt Pile: Hangover or Bust?

China has seen a large secular buildup of debt relative to GDP across government/corporate and household sectors since 2007 (here) and, in absolute terms, it has now exceeded the U.S. and EZ. Property developers and Local authority financing vehicles (LGFV) debt is now so large that managing the debt will be a headwind to long-term China growth, but will any such balance sheet adjustment be a headwind or an economic bust in the coming years?

Market Implications: China’s policy stimulation will have to get more aggressive even in the baseline scenario and this includes lower rates that can cause some further weaken of USDCNY to 7.30. Weak domestic assets are likely to show more distress and intermittent tensions in the banking and financial markets, though we feel that China authorities have sufficient tools to avoid a systemic crisis. Spillover to global markets will occur intermittently.

Figure 1: Scenario Analysis for China Debt Overhang

Source: Continuum Economics 

Case for a Debt Hangover Rather than Crisis

Our baseline (Figure 1) is not a healthy economic picture for China, as we see long-term growth slowing to 3-4% in the 2024-27 period, despite moderate government fiscal stimulation efforts and easy monetary policy. Figure 2 shows that the build up of debt since 2007 has been rapid and across government/corporate and household sectors and is worse than the other two main debtors (Japan and France). Household debt/GDP will likely not now increase much, as it high by EM standards and as the residential property markets remains less exciting that the boom from 1990 to 2019. The corporate debt/GDP ratio could stabilize at current very high levels or fall, as property developers reduce leverage and other private sector companies are less willing to expand debt leverage with more state intervention in the economy in the present common prosperity era of statist policies. This is a hangover of debt that slows economic growth in China. We are sufficiently concerned that we are revising down our 2024 GDP growth forecast to 4% from 4.5%, as we feel that the authorities are not being aggressive enough with fiscal and monetary policy stimulation.

Figure 2: Change in Total Non-Financial Sector Debt/GDP Since Q1 2007 (%)

Source: BIS/Continuum Economics

A number of reasons argue for the baseline scenario (Figure 1), where the debt overhang slows long-term growth in the 2024-26 period, but does not cause a Japan or GFC dramatic hit to growth. China M2 growth (Figure 3) is still around 11-12% and consistent with circa 5-8% nominal GDP growth. It is also worth remembering that China’s banking system is directed by the authorities to meet certain lending targets or alternatively to rollover debt (as occurred for all borrowers during the 2020 COVID crisis). This quantity guidance is more powerful than the pricing of money (interest rates). In contrast, Western economies quantity of lending is normally a function of individual bank decisions and the level of interest rates, which after a huge debt built-up and bust normally see rising non-performing loans (NPL’s) curtailing bank lending – e.g. post GFC credit supply crunch.

Figure 3: China M2 (Yr/Yr %) 

Source: Datastream/Continuum Economics 

Additionally, as we have argued in the past, China’s authorities want to avoid a banking and financial crisis and have the tools to do this. Forced mergers of weaker banks have recently been used and will likely be used again to consolidate bad debts into stronger entities. If the bad debts become too large for the system, then China can use the bad bank asset management company model that it used 1997-2003 to clean up the largest four banks. Clean-ups can also be used with key borrowers (e.g. Evergrande), where a slow restructuring occurs to ensure stability but also protect other stakeholders (e.g. sub contractors/employees and local governments). The critical issue for the size of non-performing loans and recovery of any defaulted loans and bonds, will be the scale of the decline in residential property prices. So far the decline have been small and China can restrain property developers from cutting prices too far. It is an uncertainty that we will discuss further below.

Looking at borrowers, corporate debt is 158% of GDP, but LGFV are 47% of GDP (Figure 4). Additionally, SOE form a major part of the remaining corporate debt.Though China authorities want to leave the impression that these debts are not 100% guaranteed by central government for every borrower, a rescue is extremely likely if system wide problems occur. An example of this is the 2015 swap of LGFV debt for central government debt, which amounted to a huge RMB14trn (here). Big SOE were bailed out in 1998-2000 by the government (here), which prompted heavy NPL’s for the bank that was then recapitalized by the government. 

Figure 4: Local Government and Additional LGFV debt/GDP (%)

Source: IMF Article IV 2022 (here)

If China’s authorities are likely to partially bailout LGFV’s and SOE in the coming years and swap into government debt, does China have fiscal space? Official government debt is 51% of GDP in 2022 (Figure 5), while the debt is in domestic currency and the domestic creditors (banks and other investors) can be convinced to accept debt with low yields. The IMF calculating government debt including LGFV debt and other quasi debt and this is estimated at 110% of GDP in 2022, though if LGFV debt were reclassified then the corporate debt/GDP caulcutions come down substantively. 

Additionally, the PBOC could always launch a QE government bond buying program, which would reduce the effective net government bond supply and could also keep the government debt burden low by financial repression. This is a policy that the U.S./EZ and Japan have pursued in the last two decades. Though the PBOC has been reluctant to aggressively ease, in a crisis they do have space to do much more on interest rates and potentially QE.  The government would then fiscal dominant the PBOC. 

What about China’s external position in a debt deleveraging phase? As long as this is low it should not be an issue, as China restrictions on capital movements curtails capital flight. Global investors would be conflicted. While some bond and equities investors would worry, other more passive funds could still be forced towards China government debt by benchmark weighting increasing. Additionally, China’s huge FX reserves provide a smoothing influence, alongside the natural current account surplus. Indeed, it is worth remembering that the 1997-2003 big bank bailout saw 10% of China U.S. Treasury holdings being transferred to the banks as part of a recapitalisation exercise. China has the capacity to sell U.S. Treasuries in the future to support its banking and financial system if needed. 

Overall, we feel that the wide array of tools, plus the dominance of the government over domestic borrowers/creditors and the PBOC means that a major systemic bank and financial crisis is unlikely provided intervention is timely (Figure 1 – we attach a 5-10% risk to this scenario).

Figure 5: China Government Debt/GDP is 51% and Nationalising LGFV/other Quasi Government Debt Is Feasible 

Source: IMF Article IV 2022 (here)

Arguments For 2^nd Scenario of More Serious Slowdown

The 2^nd scenario is for a more serious slowdown for the economy and is based on an economic bust caused by a deeper decline in residential property prices, deeper cutbacks in property’s contribution to GDP, less property purchases by individuals, weak growth causing private sector investment to be curtailed and insufficient policy support for the economy. This could see circa 1-3% China growth in the next few years!

The baseline argument in China is that the authorities can control property prices, which avoids a western style bust in prices. Moreover, with no ongoing property taxes in China, the incentive to sell in a downturn is less than in Western countries. Additionally, Chinese culture places great importance on property ownership, with expectations that a man will own a property before getting married.

However, the 30 year boom in property and associated 2^nd and 3^rd property ownership could be bursting. Property developers are under acute pressure still from the three red line policy and as property purchases are increasingly reluctant to pay upfront for new builds. Figure 6 shows that property developers are already paying down debt relative to GDP, which is unlikely to stop anytime soon. If the three red line polices were abandoned (politically unlikely) then it would likely slow rather than stop the deleveraging.

Figure 6: Developers Loans/GDP is Shrinking (%) 

Source: Datastream/Continuum Economics

More important is that household loan growth has slowed dramatically (Figure 7), both due to the property downturn and also as the population has become more cautious after the COVID lockdowns (China weak social security net means lots of precautionary savings for ill health and old age). The optimism of the 1990-2019 period is a memory. Official policy is still that housing is for living rather than speculation, though this was omitted from a recent Politburo statement. The official stance dampens demand for housing. China households love affair with property is thus under severe pressure and this has resulted in a fall in home sales and property under construction. It is become more difficult to sell property (Figure 8).Could this be followed by a panic and a nationwide 15-30% decline in property prices that causes an economic bust?

Figure 7: Household Loan Growth Slows Dramatically (%) 

Source: Datastream/Continuum Economics

Figure 8: Years Taken to Sell Residential Property is High in Some Key Regions (%) 

Source: Datastream

One worry is that residential property prices are overvalued, after a 30 year surge in prices. Disagreement exists among economists over how overvalued that market is due to insufficient data on house prices and incomes, but estimates have ranged from 15-25 times income for key tier 1 cities (e.g. Figure 9). This is among the most expensive in the world. This perhaps could be justified in a boom when nominal wages were growing quickly, but the more cautious 2020’s will likely see China wage growth slow noticeably and this will likely also reduce the willingness for individual households to stretch financially. Big house prices declines have also occurred as the house price/income ratio gets to such overstretched levels (e.g. Japan 1990, US 2006). Thus a scenario does exist that nationwide property prices could come down 15-30% over a number of years and cause a property bust and economic crisis in housing related sectors.

Figure 9: House Price/Income Major China and U.S. Cities (%) 

Source: Harvard Joint Center for Housing Studies/Bloomberg 

Meanwhile, the starting point for corporate debt is also very high as previously noted. China government/household and corporate debt have all surged since 2007 (Figure 10) and boosted total non-financial sector debt dramatically and now exceeds US and EZ. Such a rapid build-up has normally been followed by a debt crisis for the economy where the total debt/GDP stops rising and worst case falls as private sector pays down debt quicker than government increases debt. In Japan after 1990 corporate debt/GDP declined dramatically (Figure 11), while after 2008 U.S. household debt/GDP declined and ensured a major drag on growth.

Figure 10: Total Non-Financial Sector Debt/GDP Since Q1 2007 (%)

Source: Datastream/Continuum Economics

China corporate debt ex SOE and LGFV has still surged and property developers are deleveraging. Other private businesses are seeing a less business friendly policy stance from China’s authorities and could slow borrowing. If this ratio fell, then it would be a major headwind to China growth. 

Figure 11: China Corporate Debt Surge Similar to Japan Boom and Bust (%) 

Source: Datastream/Continuum Economics

Thus a deeper property bust could spill over to the wider private sector. This would mean that China’s authorities would have to be very aggressive in easing policy. The question is timing and what tools are used. Increased SOE and LGFV borrowing is a short-term fix, but would eventually need central government to absorb the debts and the economic effectiveness could be curtailed by inefficient SOE/LGFV operations. Using rate cuts is also perhaps not the best policy, as China households could be less sensitive to ultra-low mortgage servicing costs alone in a property bust. Fiscal policy can be the most effective, but Beijing worries about being too aggressive and this could delay action.

While this 2^nd scenario is not our main scenario, the probability has increased in recent years and that is why we wanted to highlight this alternative view of the world.