China's public finances, a tangled web

China’s public finances have been deteriorating for several years now, and the trend accelerated in 2020 with the Covid-19 crisis. Reforms introduced since 2014 have made the public sector’s accounts more transparent and improved the management of local governments’ budgets and debt. However, those changes have not stopped fiscal imbalances building up. In addition, large quasi- and extra-budgetary operations exist alongside the official budget, and there are many, sometimes opaque, links between the various public-sector entities. This means that analysing the public finances is often a complicated exercise.

Whereas the government’s “official” deficit has only showed a moderate increase in the last ten years, fiscal performance has deteriorated much more if we look at the data available for all government bodies. In addition to the increase in fiscal deficits and in the government’s direct debt, there has also been a rise in local governments’ indirect debt, which is notably taken out through their financing vehicles.

Although sovereign risk in the strict sense – i.e. the risk that the central government will have repayment difficulties – is not a concern in the short and medium terms, the structural worsening in the public finances has increasingly visible consequences. Firstly, the fiscal policy’s room for manoeuvre has narrowed. In 2021, the authorities have already had to give priority to the adjustment of public finances after the sharp increase in fiscal imbalances in 2020, while at the same time extending certain measures to support domestic demand.

Another consequence is the increasing interconnections between the financial health of the government and that of state-owned enterprises (including financing vehicles). The excessive debt of stateowned enterprises (SOEs) represents a growing contingent risk for the central and local governments, but the dependency runs both ways. The most fragile local governments (LGs) are more reliant on their financing vehicles to help cover public investment. Besides, the financing conditions of firms owned by those local governments are becoming tougher, especially since there is now an erosion of implicit State support. Reforms to strengthen the financial health of both local governments and state-owned enterprises are thus becoming increasingly urgent.

Budget deficits before and after Covid-19 shock

In 2020, the combination of the Covid-19 crisis, the economic growth slowdown and the support plan implemented by the authorities led to a sharp increase in fiscal deficits and public debt, after they had already worsened for several years. The deterioration has been widespread, affecting all government bodies included in the official budget as well as quasi- and extra-budgetary entities. Fiscal adjustment efforts have become crucial and increasingly constrain China’s economic policy.

A deterioration underway for several years...

The sharp increase in fiscal imbalances in 2020 followed several years of steady deterioration. The implementation of the new budget law in 2014¹ and the reforms that followed led to improvements in budget transparency and management. However, those changes have not prevented rising sources of vulnerability. Firstly, budget deficits increased gradually until the Covid-19 shock (see the box below for definitions of the various budget balances and what they cover). Between 2015 and 2019, the “official” deficit only worsened from -2.4% of GDP to -2.8%, but the general budget deficit rose from -3.4% to -4.9% and the total consolidated deficit of all government bodies doubled from -2.3% to -4.6%. See Chart 1.

On the one hand, the structural slowdown in economic growth and global trade as well as reforms of the tax system (widening of the VAT base², reduction in income tax and customs tariffs, cuts in social security contributions for enterprises etc.) have reduced budgetary revenue. The tax base (the general government’s tax revenue) fell in the years preceding the Covid-19 crisis, from 18.5% of GDP in 2014 to 16% in 2019, and then 15.2% in 2020³. On the other hand, total public spending and investment (of all four budgets) rose substantially to 41.3% of GDP in 2019 and 43.8% in 2020 from 37% in 2014, in order to support domestic demand. See Chart 2.

In addition, the reserves available to the government have dwindled. To help finance the general budget, the authorities have moved resources from the reserves of various public-sector funds (such as the stabilisation fund, various government funds and the fund financed through the profits of state-owned enterprises). Those transfers increased sharply in 2020, after already several years of rise (they took about 35% of the revenue of all the funds concerned vs. 25% in 2019 and 15% in 2016-2017⁴). For example, transfers from the stabilisation fund to the general budget rose from RMB 100 billion in 2014 to RMB 280 billion in 2019 and RMB 530 billion in 2020 (0.5% of GDP), and the stabilisation fund’s accumulated balance could fall close to zero this year. Meanwhile, the reserves of social-security funds reached almost 10% of GDP in 2019 after rising for several years. The government is not allowed to tap those reserves to fund the general budget, but yet most of the support measures introduced in 2020 were covered by those reserves.

The deterioration in public finances has mostly affected local governments. There is a structural imbalance between their revenue (due to their narrow tax base and central government transfers that are insufficient and based on a complex system) and their large spending responsibilities in terms of public services and investment. This has led to local government’s large deficits (averaging -9.9% of GDP before central government transfers in 2015-2019, and -2.1% after transfers), ever-growing debt and heavy dependency on alternative and less well controlled sources of funding. These mainly consist of land sales, which have fed speculation in the real estate markets, as well as various local taxes and indirect debt via dedicated financing vehicles (see Box and Chart 3).

... and that accelerated in 2020

The government introduced fiscal support measures at the start of the Covid-19 crisis in February 2020 and published all of its budget for 2020 in May. The official budget deficit target was increased from RMB 2,760 billion in 2019 to RMB 3,760 billion in 2020, which was due to represent -3.6% of GDP vs. -2.8% in 2019. At first glance, this seemed to indicate a moderate loosening of fiscal policy. In reality, the official deficit target announced in the spring of each year does not reflect all of the measures being considered. It can rather be regarded as a signal of the short-term direction of fiscal policy. In fact, the official deficit announced at the end of the fiscal year is always equal to the initial target (in 2020, the official deficit was indeed RMB 3,760 billion, which represented -3.7% of GDP in the end as GDP growth was slightly weaker than expected).

As a result, the fact that the official deficit in 2020 was historically high (above 3% for the first time) and that its increase was unusually large (RMB 1,000 billion, almost 1 point of GDP) pointed to a major fiscal easing. However, these figures underestimate the real extent of measures implemented in 2020. The analysis of data available for all government bodies and for the four budgets shows a larger increase in fiscal imbalances and a larger support plan, closer to 5% of GDP (which remains very modest compared with fiscal packages adopted in most developed countries).

The total consolidated deficit of all government bodies (the “four budgets”) doubled year-on-year, from RMB 4,600 billion in 2019 to RMB 9,200 billion in 2020, i.e. from -4.6% to -9% of GDP. It was lower than the authorities’ initial forecast (of -11.4% of GDP), since the economic rebound starting in the second quarter of 2020 allowed the government to limit stimulus spending and supported total revenue.

Over 2020 as a whole, the increase in fiscal deficits was explained primarily by the fall in total revenue (down 2.4% in 2020 compared with 2019), which in turn was mainly due to the decline in tax revenue (down 2.3%) and social-security contributions (down 13.3%). The increase in total expenditure (+9.2%) was moderate. It was mainly driven by LG investment (spending by government-managed funds jumped by 28.8% in 2020), while the increase in total current expenditure was very limited (+2.8%).

As a result, the general budget deficit rose from -4.9% of GDP in 2019 to -6.2% in 2020. With interest on debt rising very slightly and estimated to equal 1% of GDP in 2020, the general government primary budget deficit was -5.2% in 2020 vs. -4% in 2019. See Chart 1.

The balance of the three other budgets was in deficit for the first time in 2020 (-2.9% of GDP), because the post-Covid19 support package was implemented to a large extent through the quasi-budgetary government-managed funds (which posted a deficit of -2.4% of GDP in 2020), and through social-security funds, which posted an exceptional and probably temporary deficit equal to -0.7% of GDP. See Chart 4.

Around 40% of the total amount of the post-Covid19 stimulus package consisted of new public investment, mainly in infrastructure⁵. The rest consisted of one-off measures (some of which have been maintained in 2021) such as: healthcare expenditure (controlling the pandemic, medical equipment), tax and social-security exemptions and reductions, changes to the unemployment benefit system to accelerate payouts and extend coverage (particularly for migrant workers), a reduction in levies, and other measures to help the most vulnerable corporates and households.

In addition to fiscal support, LGs’ financing vehicles and SOEs also embarked on new expenditure (investments, recruitment). The resulting extra-budgetary deficit is hard to estimate. Based on available data and IMF estimates, it may have been around 4-5% of GDP in 2019, and it continued to rise in 2020.

Funding deficits on the local bond markets

The central government and LGs cover almost all of their official net borrowing requirement (i.e. after transfers from various public funds and excluding indirect extra-budgetary debt) through local bond markets.

The budget reports of the authorities plan the annual bond issuance quotas for: i) “general” bonds, issued by the central government (around two thirds of the total) and by LGs, and which usually finance the general budget up to the official forecast deficit, and for: ii) “special” bonds, which are mainly issued by LGs to cover specific expenditures of government-managed funds.

In 2020, total new general bond issuance exceeded the official budget deficit by almost RMB 300 billion, totalling RMB 4,040 billion. Moreover, new issues of special bonds by LGs increased sharply to RMB 3,600 billion (slightly less than the initially authorised quota), and were supplemented by an exceptional RMB 1,000 billion issue of special bonds by the central government (which had only carried out this kind of bond issue twice previously, once in 1998 and once in 2007). This means that the central government financed a larger share of fiscal deficits in 2020, in order to make up for the loss of revenue arising from the Covid-19 crisis. See Chart 5.

The issuance of general and special bonds does not pose any difficulty. Liquidity in the local bond market is abundant, supported by a large amount of available savings in the financial sector (national savings represent 45% of GDP and are still mainly invested locally). After a period of monetary policy loosening to respond to the Covid-19 shock in the first quarter of 2020, the central bank has cautiously tightened credit conditions since the fourth quarter, while maintaining comfortable liquidity levels in the local markets. The government’s funding terms have remained stable. On average, since 2019, local governments have issued bonds at spreads of around 20-40 basis points (bp) over sovereign bond yields of the same maturity. See Chart 6.

Less room for manoeuvre, requiring more careful adjustments of fiscal policy

Last year, public finances were solid enough to absorb the Covid-19 shock. However, the government has much less room for manoeuvre as fiscal policy is now constrained by the need to reduce deficits and mitigate risks. In its last budget report, the finance minister acknowledged the “serious” nature of the situation. In order to meet their various fiscal policy targets (maintaining some measures to support domestic demand while making fiscal consolidation efforts and containing public debt growth), the authorities have to adjust their instruments more carefully, including through closer monitoring of public capital expenditure. They may also increase their recourse to taxes in the medium term.

The 2021 budget plan

The official deficit target for 2021 has been reduced by only RMB 190 billion to RMB 3,570 billion, or -3.2% of GDP, as opposed to -3.7% in 2020. This suggests a cautious policy tightening. In addition, in their March 2021 budget report, the authorities projected a reduction by less than 10% in the total deficit of consolidated public accounts, to RMB 8,470 billion in 2021, or -7.6% of GDP (assuming nominal GDP growth of 10%) vs. -9% in 2020. The expected increase in total public spending was +5.6% in 2021, down from +9% in 2020. Meanwhile, total revenue was expected to rebound strongly, rising by 9% after the contraction in 2020, supported by the upturn in activity and the gradual elimination of tax and social-security exemptions and deferrals for corporates (some tax support measures are still being maintained for small firms). Therefore, the fiscal adjustment that was planned last March relied on social-security funds, which were expected to return to surplus. The general budget deficit was also expected to improve from -6.2% of GDP in 2020 to -4.6% in 2021, which is lower than its pre-crisis level (of -4.9% in 2019). See Chart 1 & Chart 4.

Meanwhile, the deficit of government-managed funds was expected to continue to widen. According to the official forecasts announced in March, it was expected to reach -3.3% of GDP in 2021 vs. an actual deficit of -2.4% in 2020 and less than 1% in 2018 and 2019. This projection was based firstly on the expected stabilisation of land sales proceeds, illustrating Beijing’s desire to cool the real estate market and, secondly, on a moderate slowdown in capital expenditure growth.

Mid-year adjustments

The sharp rebound in economic growth between the second quarter of 2020 and mid-2021 led to a solid recovery in government revenue and allowed a rapid adjustment of economic policy priorities. The authorities started tightening credit conditions from the fourth quarter of 2020 and revised public investment plans in the first quarter of 2021.

Tax revenue in the general budget recovered more quickly than expected in the first half of 2021 (up 22.5% year-on-year), already exceeding its level in the first half of 2019. Meanwhile, local government revenue from land sales also rose sharply (up 22.4% year-on-year), taking advantage of the recovery in the property market. See Chart 7.

At the same time, current expenditure in the general budget returned to normal in the first half of 2021 and capital expenditure by LGs was much lower than forecast. By the end of June, LGs had only issued 35% of their authorised quota of bonds for the year (and so the total annual quota was reduced slightly). That adjustment came alongside tougher credit conditions, and investment in public infrastructure quickly levelled off in the first half of 2021. See Chart 8.

Beijing is seeking tighter control over local governments’ spending. The greater discipline being imposed on them and the closer monitoring of their investments in the last few months contrast sharply with the strategy adopted in 2008 and 2009, when Beijing gave carte blanche to the regions to spend and stimulate growth in response to the global financial crisis. Investment in public investment projects remains a favoured countercyclical policy instrument, but it is being adjusted more closely in line with trends short-term activity data. These adjustments are intended to limit the rise in LG deficits and debt as much as possible, while retaining the ability to respond if domestic demand weakens. Given the sharper-than-expected and broad-based slowdown in activity in summer 2021, the authorities are likely to make further adjustments to bolster their monetary and fiscal policy support measures in the next few months.

Public-sector debt: the problem of indirect debt and contingent risks

Despite the general deterioration in public finances in recent years and the Covid-19 shock, the central government is still in good financial shape and its direct debt remains very moderate. The conclusion regarding local governments is more complex. Their fiscal operations have become more disciplined and transparent because of reforms adopted since 2014, and their “official”, explicitly budgeted, total debt remains under control. However, LGs are still making extensive use of “extra-budgetary” financing vehicles to cover certain public expenditure. This is helping LGs overcome the shortfall in their resources, but it is pushing up their “implied”, or indirect, debt, taken out by their financing vehicles, in an opaque manner. This indirect debt is high, and situations differ widely from one region to another. Moreover, the central and local governments are also facing large contingent risks associated with the debt of state-owned enterprises. The interconnection between public finances and credit risk has increased.

Central government solvency still good

The solvency of the central government remains strong. Its debt is very moderate and highly sustainable, it has large assets and it can easily cover its financing needs in the bond markets. As a result, sovereign risk in the strict sense is not a concern in the short and medium terms.

The central government’s debt rose quickly in 2020, by 24% in nominal terms, whereas it had increased by 11-13% per year between 2015 and 2019. Given the sharp slowdown in nominal GDP growth, the debt/ GDP ratio rose from 17% in 2019 to 20.6% in 2020, which remains very moderate. Debt interest costs are low: based on available data for the general government, interest payments represented only 4.4% of total revenue in 2019 (0.9% of GDP) and 5.4% in 2020 (1% of GDP).

Refinancing risks are almost non-existent. More than 80% of the central government’s debt is long-term, and it consists almost entirely of bonds denominated in local currency. Most of these are owned by Chinese investors (principally commercial banks), which represent a stable base of creditors. The proportion of government bonds held by foreign investors remains low but it is rising gradually, reaching 10% of total bonds in 2020 vs. 3% in 2014. The central government has also made slightly greater use of international bond markets in the last five years, but the amounts involved remain very low: its foreign currency debt amounted to 0.9% of total central government debt in 2020 (RMB 193 billion) or 0.2% of GDP⁶ .

Finally, the government debt dynamics benefit from a highly favourable differential between GDP growth and interest rates, and this will remain the case in the medium term despite the expected slowdown in economic growth⁷. The apparent interest rate on debt (interest payments on existing debt, calculated for the general government) was estimated at 2.5% in 2019-2020 and is lower this year. Based on our central medium-term macroeconomic scenario (with a very slight downtrend in both fiscal deficits and in the apparent interest rate on debt), central government debt is projected to increase slowly but remain below 25% of GDP by 2025. See Chart 9.

Local government debt: low clarity and high risk

Local governments are more indebted than the central government and there is a lack of clarity regarding their debt, since it is mainly taken out indirectly via their financing vehicles.

Since the new budget law was adopted in 2014, LGs have been authorised to borrow directly, subject to new debt quotas determined by the central authorities and specified in the annual budget report. As a result, the official, explicitly budgeted amount of total LG debt has increased since 2014⁸. It stood at 21.6% of GDP at end-2019 and 25.3% at end-2020. See Chart 9.

Taken as a whole, LGs’ official debt benefits from the same positive factors as central government debt, which make it sustainable over the medium term: a wide differential between GDP growth and interest rates, a highly favourable profile and moderate interest charges. LG debt consists almost exclusively of bonds issued in local markets, most of which are long-term and 80% of which are held by commercial banks, mainly regional ones. However, financial situations vary extremely widely from one province to another, and some local governments are already facing excessively heavy debt servicing charges.

Most importantly, direct bond issuance is not enough to cover the entire financing needs of LGs. This means that most of them are continuing to use financing vehicles. Although these vehicles have been banned from taking out debt on behalf of LGs since the 2014 budget law, their debt does in fact represent indirect, implied debt for their local governments.

It is a major source of vulnerability for public finances, primarily because this kind of debt has continued to rise rapidly in recent years and reached high levels. Moreover, local government financing vehicles borrow within an unclear regulatory framework and sometimes in a highly opaque manner. Their debt consists mainly of bank loans, along with bonds (the most “visible” portion of debt, estimated at 20-25% of the total in 2020) and other credits from non-bank institutions of the shadow banking sector. The total amount of debt is hard to gauge.

According to IMF estimates⁹, the total debt of local government financing vehicles and other extra-budgetary funds handling public investment on behalf of LGs has increased by 15-20% per year since 2018. It represented 43% of GDP at end-2019 and 48% of GDP at end 2020.

Therefore, the total (direct and indirect) debt of LGs amounted to 73% of GDP at end-2020, which is excessively high, including by comparison with other emerging economies or OECD countries. See Chart 10. Lastly, there is a high risk that local government financing vehicles will experience difficulties to refinance and then repay their debt. Indeed, returns on their assets (mainly infrastructure) are long to come and often not high enough to cover debt repayments. The IMF estimates that at least two thirds of their debt is destined to be recognised directly as LG liabilities.

Debt of state-owned enterprises: increasing again in 2020 after three years of improvement

The central and local governments face large contingent risks associated with the excessive debt of state-owned enterprises (including financing vehicles)¹⁰ and high credit risks in the financial system.

Total debt of non-financial corporates was estimated at 162% of GDP at end-2020 vs. 152% at end-2019 and 158% at end-2016¹¹. The increase in the debt ratio in 2020 in fact followed three years of slight decline, and resulted from both the sharp slowdown in GDP growth and the faster rise in the debt stock (+10% in 2020 vs. +7.2% per year on average between 2016 and 2019). The debt increase was mainly driven by the public sector.

Based on CNBC estimates¹², the debt of state-owned enterprises (including financing vehicles) represented around 70% of total corporate debt, i.e. 114% of GDP at end-2020 vs. 106% at end-2019¹³.

This means that financing vehicles were responsible for around 40% of that total. Accordingly, public-sector debt as a whole (government + non-financial corporates) totalled around 160% of GDP in 2020¹⁴.

Increasing interconnections between sovereign risk and credit risk

In 2020, public finances deteriorated while total SOE debt increased again. In addition to these dynamics, there is a growing interconnection between the financial performance of local governments and credit risks. On the one hand, the excessive debt of SOEs represents a contingent risk for the government. On the other, the fragility of some local governments is starting to lead to both tougher financing conditions and higher default risks for their enterprises. This could also affect the performance of financial institutions, particularly regional commercial banks that are the main creditors of local governments.

Payment difficulties experienced by state-owned enterprises have recently increased due to the combined effect of the deterioration in their financial performance and the weakening of state guarantees. In addition, credit conditions have become tighter since the fourth quarter of 2020.

The weak financial health of the SOE sector is not a new problem in China – it has been caused by poor governance, low profitability and excessive debt. The Covid-19 shock on activity and corporate profits has made the situation worse. SOEs’ capacity to service their debt has deteriorated substantially, especially since new credits in 2020 went more to firms that already had the heaviest debt before the Covid-19 crisis¹⁵. State-owned enterprises belonging to local governments (particularly in the transport and real estate sectors) are estimated to be among the least able to service their debt; in the first quarter of 2021, around 10% of local SOEs had an interest coverage ratio (ICR) of less than 1, according to World Bank estimates.

At the same time, whereas SOE debt had long benefitted from (explicit or implicit) state guarantee (either by the central or local governments), this unconditional support has started to erode. This has resulted firstly from the authorities’ reform efforts aimed at cleaning up practices in the financial sector and among SOEs, and at reducing moral hazard. However, the weakening of guarantees provided by local governments is also the result of their deteriorating public finances. Some local governments are simply no longer able to support their firms when required.

As a consequence, there has been a sharp increase in defaults among state-owned enterprises in the last year. So far, defaults have concerned bond debt more than bank loans (on which defaults are also less visible). In the local bond market, the total number of defaults among SOEs rose from 43 in 2017-2019 (with debt in default totalling RMB 41 billion) to 80 in 2020 (with debt in default totalling RMB 98 billion). The amount of debt falling into default was around RMB 38 billion in the first quarter of 2021 alone. See Chart 11.

The total amount of corporate debt in default remains limited (1% of all corporate bonds outstanding in 2020), but the rising frequency of default events clearly shows both the deterioration in financial positions and a change in behaviour in the Chinese market. Whereas most bond defaults initially affected private-sector enterprises (the first default happened in 2014), state-owned enterprises have accounted for most defaults since 2020, including some large firms. Meanwhile, although no financing vehicles have defaulted on their bond payments so far, payment difficulties have started to appear in relation to debts owed to shadow banking institutions.

The rise in default risk and the concerns of creditors have pushed up borrowing rates in the bond markets, the distinction between stateowned enterprises and private-sector enterprises has become less clear, and the market is pricing in less of an implied government guarantee, particularly in provinces that have the weakest finances. According to World Bank calculations, between the start of 2020 and mid-2021, the surplus risk premium applied to private-sector corporate bonds fell by around 20bp compared with bonds of SOEs owned by the central government and by 40-50bp compared with bonds of SOEs owned by local governments¹⁶.

Rising defaults among state-owned enterprises and the deterioration in local governments’ finances have, in turn, contributed to tougher credit conditions in the most fragile regions. As a matter of fact, the proportion of new credits taken by corporates and households in the most indebted provinces fell sharply in the second half of 2020¹⁷.

****

The increasing interdependence between local governments, their enterprises and regional banks is therefore creating negative dynamics in credit risk – thereby weakening the financial sector – and in public finances. These dynamics are likely to continue, and defaults by stateowned enterprises could multiply in the next few months. Efforts to clean up practices in the financial markets represent a positive development that should improve the allocation of capital in the medium term. In the short term, however, the challenge for the authorities is to keep events of default under control, in order to stop the contagion effects spreading to the financing conditions of other economic agents and to prevent any risk of instability in the financial system (such as a confidence crisis and a sudden adjustment of market rates, leading to further defaults). This means that the Chinese state is likely to continue supporting the most sensitive and strategically important firms. At the same time, the authorities are expected to continue reforms aimed at reducing the debt of state-owned enterprises and local governments, because making public finances more sustainable will be necessary to help them to realise their medium-term development strategy.

¹ The budget law adopted in 2014 aimed among other things to: i) improve reporting by government entities and introduce the annual publication of the “four budgets”, ii) streamline and improve the planning of local governments’ spending, and allow them to raise debt directly in the bond markets, and iii) reduce the use of financing vehicles and separate their activities from local governments’ budgets. Major progress has been achieved on the first point, and the local government bond market has grown quickly since 2014. However, the third objective has not been achieved, since financing vehicles have continued to proliferate.

² The VAT reform, initiated through a pilot programme in 2012, was extended across China in 2016: the “business tax” that existed alongside VAT and applied to certain industries was replaced by a VAT applicable to all goods and services (with various rates depending on the sector).

³ The main tax cuts were applied in 2018 (for an estimated total reduction of 1.5% of GDP) and 2019 (2% of GDP). 4 C. Wong, National University of Singapore, East Asian Institute: China’s post-Covid goldilocks budget – How big should it be? (18 March 2021).

⁴ C. Wong, National University of Singapore, East Asian Institute: China’s post-Covid goldilocks budget – How big should it be? (18 March 2021).

⁵ New investments provided for by the stimulus plan were aimed particularly at areas such as transport infrastructure, environment, water and healthcare, urban/rural development, industrial parks and “new” sectors (high tech, internet networks, 5G, artificial intelligence).

⁶ More generally, the Chinese economy has limited foreign-currency debt, estimated at 10% of GDP in 2020 and consisting mainly of debt owed by banks and non-financial companies.

⁷ Real GDP growth averaged 6.7% per year between 2015 and 2019 and slowed to 2.3% in 2020. We expect 8.2% in 2021 and then 5.4% per year on average between 2022 and 2025. Nominal GDP growth averaged 9% per year between 2015 and 2019 and was 3% in 2020. It is projected to accelerate to 10.2% in 2021 and then average 7.5% per year between 2022 and 2025.

⁸ Some of the first bonds issued by LGs replaced the portion of their financing vehicles’ debt that the new budget law forced them to recognise. This swap programme totalled RMB 15,400 billion, equal to two thirds of the debt of financing vehicles at end-2014 (around 20% of GDP).

⁹ IMF (2 December 2020): Article IV Staff report.

¹⁰ The frontier between financing vehicles and other SOEs is sometimes blurred. The purpose of financing vehicles is to support fiscal policy, whereas the operations of other SOEs are strictly commercial.

¹¹ Data from the CNBC (Center for National Balance Sheet of China) – NIFD (National Institution for Finance & Development).

¹² CNBC – NFID (1 April 2020): China’s leverage ratio likely to increase in 2020. 2019 Report.

¹³ Measured as a % of assets, the debt of SOEs rose in 2020 for the first time since 2017, reaching about 65% vs. 63.9% at end-2019.

¹⁴ These figures seem to be at the lower end of the likely range. According to other available estimates, SOEs account for 67%-85% of the total debt of Chinese corporates. See: World Bank (China economic update, July 2020 & June 2021), OECD (State-owned firms behind China’s corporate debt, 7 February 2019) and IFI (Global debt monitor database, September 2021).

¹⁵ IMF (April 2021): Global Financial Stability Report.

¹⁶ In the second half of 2020, the average spread was around 300bp for private-sector companies, 90bp for local government-owned companies and 70bp for central government-owned companies.

¹⁷ IMF (April 2021): Global Financial Stability Report.