Evolving Financial Stress in China’s Property Development Sector


Photo: zhihao – Getty Images

Abstract

Financial stress in China’s property development sector has attracted significant attention because it may have systemic consequences for financial stability in the broader Chinese economy. Though China Evergrande Group, one of the country’s largest and most leveraged property developers, has received a considerable share of this attention, risks in the sector were building for some time prior to Evergrande’s default in 2021. This article reviews contributing factors to the sector’s financial fragility and explores the characteristics of the financial stress faced by major developers. It also considers some likely consequences of this fragility for the Chinese property development sector and beyond.

Introduction

Real estate investment associated with rapid urbanisation and households’ demand for property as a financial asset has been an important source of economic growth in China in recent decades. Faced with substantial demand over a relatively short period of time, many Chinese real estate developers were able to run a risky and highly profitable business model, which relied on high leverage and rapid project turnover. However, this model also left developers particularly vulnerable to short-term funding stress – a risk that was realised in 2021 with the widely publicised default of China Evergrande Group. As large parts of China’s property development sector have come under severe financial stress over the past year or so, concerns about risks to China’s financial stability and the economic outlook have also increased. The wave of defaults and subsequent restructurings that are ongoing in the sector are likely to permanently change the structure of China’s property development sector and reduce its capacity to engage in the rapid, large-scale development seen in recent decades.

Urbanisation and residential investment in China

The real estate sector has been a major contributor to China’s economic growth, spurred by mass urbanisation over the past several decades. In the late 1970s, when a series of economic reforms and many market-oriented policies began to be introduced, less than one-fifth of the Chinese population lived in urban areas; by 2020, this share had grown to more than three-fifths (United Nations 2021). Because significant investment in urban real estate was required to facilitate this process, dwelling investment as a share of GDP in China increased rapidly over the past few decades (Graph 1).

Graph 1

Graph 1: Line graph of residential gross fixed capital formation (GFCF) as a proportion of GDP for China, Korea, Japan, Australia and United States since 1960. Chinese GFCF since 2010 has been higher than the other countries' historical GFCF.

The growth of the Chinese real estate sector gathered pace in the 1990s as the privatisation process of the housing stock, which had been largely state-owned until that point, began. These reforms unleashed a considerable increase in housing demand as residents were allowed to purchase the homes they had previously been renting from the state and also increase their housing consumption (Wang 2011). Demand was further spurred by households acquiring property as a financial asset; other asset classes like equities or bonds had historically been inaccessible or simply offered lower returns, partly because of artificially low interest rates (Adams et al 2021). Cultural factors also contributed to this demand, with home ownership often seen as a necessity for marriage (Glaeser et al 2017). Reflecting the characteristics of this underlying demand, Chinese home ownership rates are now among the highest in the world (Graph 2). Second-home ownership is also very common, with more than 20 per cent of urban households owning multiple homes (Huang, Yi and Clark 2020).

Graph 2

Graph 2: A horizontal bar graph showing home ownership rates of China and selected OECD countries as of 2014-2015. The graph shows that China has the highest home ownership rate, around 90 per cent, with most other countries around 50 to 70 per cent.

This high level of demand for housing enabled real estate development more broadly (including commercial real estate) to become a significant driver of the Chinese economy, directly affecting one-quarter of GDP in 2018 (Rogoff and Yang 2021). As a result, any concerns about the health of the Chinese real estate sector have substantial implications for Chinese economic growth, as well as for major trading partners, including Australia, which benefit from China’s demand for iron ore and coal exports (Kemp, Suthakar and Williams 2020).

China’s real estate development funding model and its risks

Significant real estate demand has benefited the Chinese property development sector and enabled some Chinese developers to become the largest in the world. The basic business model for most of these developers can be characterised in four steps:

  1. Obtain funding to purchase land from local governments.
  2. Presell properties to be developed to home buyers.
  3. Begin construction on the project.
  4. Use income from the project to secure funding to begin another project.

This reliance on rapid expansion was historically quite profitable, with the median major developer earning around a 7 per cent return on assets at the beginning of the 2010s, and some developers earning returns exceeding 10 per cent (Graph 3).

Graph 3

Graph 3: A two panel line graph of return on assets and interest coverage ratios for major listed property developers since 2010. The left panel shows that the median return on assets has fallen somewhat over the decade, as have the 90th and 10th percentile. The right panel shows the median interest coverage ratio has fallen a little, as have the 10th percentile, while the 90th percentile fell until 2016 and then remained stable.

However, profitability in the property sector has declined over time as China’s housing stock has developed, the urbanisation process has begun to slow, and financial markets have matured further and become more accessible for households. Profitability has also declined as authorities have cracked down on housing speculation and sought to rebalance the economy away from reliance on the sector. Indeed, in 2021 a large share of listed property developing companies did not make any profits at all, with around 10 per cent of companies reporting negative returns on assets. Likewise, as profitability has declined and indebtedness has increased, the ability of some of the weaker developers to cover their interest payments out of earnings has also steadily declined. By 2020, around 10 per cent of major listed developers had an interest coverage ratio of less than one – meaning they were not earning enough to service their debt, even before accounting for any maturing loans.

As profitability declined since 2010, developers increasingly innovated in their financing structures and business models, which increased financial risks in the sector. Issuance volumes of US dollar bonds increased through the 2010s, and these bonds grew to become a notable portion of major listed developers’ interest-bearing debt (Graph 4). US dollar bonds were popular among overseas investors who were attracted by high-yielding bonds, while Chinese developers preferred offshore issuance due to the favourable bond pricing resulting from this demand and the limited need to provide collateral. Private real estate developers’ US dollar bond issuance has been found to correlate with actual investment outlay, rather than just relative interest rates, implying that US dollar bond issuance was needed by these developers to fund activity (Ding, Huang and Zhou 2019).

Graph 4

Graph 4: A two panel graph. The left panel is a single stacked bar showing different types of liabilities as a share of major listed developers' total liabilities at the end of 2020. It shows CNY loans and bonds, presales, and accounts and notes payable as the largest proportions of liabilities. The right panel shows a stacked density chart of interest bearing debt since 2010 owed by developers that have issued US dollar bonds. The panel shows that overall interest bearing debt has increased significantly and US dollar bonds outstanding have increased a lot since 2010.

In general, property developers face less rollover risk from bond financing than traditional bank loans because of the lower frequency of refinancing and the larger value of bond issuance. However, declining profitability, rising leverage and the way in which bond issuance is structured can make developers vulnerable to changes in investor sentiment. To enable access to offshore equity and debt financing, many private developers shifted to a complex corporate structure whereby a ‘holding company’ is registered in Hong Kong (often listed on the Hong Kong stock exchange). This holding company retains a controlling equity stake in the onshore (mainland China) ‘operating company’, which holds all the real estate assets. The holding company would then have additional subsidiary corporate vehicles registered in other jurisdictions, such as the Cayman Islands, which would issue the US dollar-denominated bonds. Because the offshore bondholders are creditors in only the offshore holding company, which itself has only an equity claim in the operating company’s onshore assets, offshore creditors are ‘structurally subordinate’ to onshore creditors – that is, onshore creditors have priority over offshore creditors in the event of bankruptcy. Despite developers’ attempts to enhance their offshore credit quality with legal devices like ‘keepwell deeds’ (the enforceability of which is in question), offshore developer bonds are naturally riskier than onshore credit and are subject to sudden changes in offshore creditors’ confidence (Tudor-Ackroyd 2020). Developer funding also relied on non-interest-bearing debt, like accounts and commercial notes payable to suppliers, and funds from presales.

Over the 2010s, many developers expanded their business models well outside of real estate. However, rather than diversifying to improve cash flow stability, a number of developers invested in relatively speculative or unprofitable ventures, including football teams. Evergrande Group was particularly well known for this strategy, having invested in ventures as diverse as electric vehicles, a streaming service and theme parks.

Along with these changes in the composition of their financing and business models, developers were also operating with increasingly high leverage ratios and gearing ratios. This trend prompted a series of regulatory responses, which culminated in a crackdown on developer leverage, known as the ‘three red lines’ (Kemp, Suthakar and Williams 2020).

Tighter regulation of the property sector

By 2020, Chinese regulators had become increasingly concerned about growing leverage in the real estate development sector. Authorities implemented new regulations on the property sector in August that year, introducing the so-called ‘three red lines’ policy, which would apply first to 30 major developers before being rolled out to the rest of the sector (Wang et al 2020). The thresholds – or ‘red lines’ – were:

  1. The debt-to-asset ratio (not including presales) was not to exceed 70 per cent.
  2. The net debt to shareholders’ equity was not to exceed 100 per cent.
  3. The cash to short-term debt was not to fall below 100 per cent.

The penalty for non-compliance would be strict limits on developers’ allowable annual debt growth, depending on the number of red lines crossed. These thresholds were only a little stricter than the ratios actually reported by the median major developer for 2020, while developers in the 90th percentile (or 10th percentile for cash ratios) were well in breach of these ratios (Graph 5). One conclusion from this exercise was that a large number of major developers would need to undergo significant deleveraging in 2021.

Graph 5

Graph 5: A three panel line graph showing developers' leverage ratios, gearing ratios and cash ratios. Each panel shows the median, 10th and 90th percentile of one of these measures, and the threshold which would have amounted to a breach of the relevant 'red line'. Each panel shows that the median in 2020 was close to the relevant 'red line' level, while the 90th percentile (or 10th for cash ratios) developer was well in breach.

Although the three red lines policy was aimed at improving the financial stability of the real estate development sector in the long run, these regulations made the financial position of the riskiest developers even more precarious in the short term as they attempted to quickly deleverage (Graph 6).

Graph 6

Graph 6: A two panel vertical bar graph. The left panel shows the change in developers' total liabilities in 2021, categorised by how many 'red lines' they breached in 2020. Total liabilities increased for developers that breached no, or one, red line and decreased for those that breached three. The right panel shows the change in developers' earnings in 2021, also categorised by 'red lines' breached. All developers' earnings fell, but substantially more for those that breached two or three 'red lines'.

Evergrande Group’s spiral of financial stress

The level of financial stress faced by property developers was apparent by mid-2021 when two major Chinese developers, China Fortune Land and Sichuan Languang, defaulted on US dollar bonds (Bloomberg 2021a; Bloomberg 2021b). However, it was Evergrande Group that ignited more serious concerns about the industry later that year. Evergrande was China’s largest developer by total assets at the end of 2020, and was known for its high leverage and wide range of business ventures. Evergrande had been increasingly profitable until 2018; however, its earnings fell considerably over the next couple of years, even while its leverage increased, the interest coverage ratio fell and the maturity of its liabilities shortened.

As it became apparent over 2020 that Evergrande’s fundamentals had deteriorated, the volatility of the company’s bond and equity prices increased and became highly sensitive to negative news about the company (Graph 7). At the end of May 2021, Evergrande’s share and bond prices fell rapidly, triggered by news that regulators were examining transactions between Evergrande and Shengjing Bank, in which Evergrande held a major stake. The equity and bond price falls steepened in July on a string of negative news – a court in Jiangsu province froze Evergrande’s bank deposits, sales in some Evergrande developments were suspended, and Hong Kong banks suspended new mortgages for Evergrande developments in Hong Kong (Yu 2021; Hale 2021; Reuters 2021; Bloomberg 2021c). Despite Evergrande’s size and prominence, financial markets initially did not treat these issues as systemic or sector-wide.

Graph 7

Graph 7: A two-panel line graph where the top panel shows Evergrande's stock price since 2019 and the bottom shows the price of an Evergrande US dollar bond. Both panels show volatility in 2020, and a sharp fall in mid 2021.

Contagion in the development sector and beyond

Between July and September 2021, bond prices for some other highly leveraged private property developers started falling as markets became increasingly concerned about developers’ ability to repay their debts. These concerns were exacerbated by the warnings of imminent default by other developers around this time, including Fantasia Holdings. As concerns about the financial stability of property developers spread, a large number of local governments imposed further restrictions on the withdrawals of presales funds from escrow accounts by all developers (Yu and Jim 2022). Although these measures were aimed at ensuring the delivery of local projects by enforcing existing rules on the usage of presales funds, the change in enforcement effectively cut off a large number of developers from having access to these funds, which is an important source of developer funding, adding further to their funding stress (Yu and Jim 2022).

While financial markets initially viewed these developments as a liquidity crisis for a few developers, in late 2021 concerns started to shift towards solvency. Around this time, a significant gap opened up between the equity prices of state-owned developers and private developers (Graph 8). Rather than pricing developer bonds on the basis of the parent company’s liquidity – their cash and short-term assets – financial markets became focused on their solvency, including whether they would be likely to receive government support. The evolution of this shift in sentiment can be seen in rolling regressions of major listed developers’ US dollar-denominated bond prices on their financial characteristics as of 2020 – their cash holdings, current assets, change in net profits since 2019, and total liabilities – and whether they were a state-owned enterprise (see Appendix A). The results suggest that the largest contributing factors for bond price changes (holding all other variables constant) in the early stages of the crisis were the developer’s cash holdings and current assets (Graph 9) – that is, developers’ liquidity and ability to meet their short-term debt repayments. However, by late-September 2021, the focus had shifted to solvency, with the largest contributing factors being the developer’s total liabilities and whether they were a state-owned enterprise, indicating that markets became focused on simply whether a developer had explicit state support.

Graph 8

Graph 8: A line graph showing equity prices for state-owned and private property developers. The two lines diverge around September 2021, with the line for the state-owned developers stabilising while the privately owned developers' equity prices continue to fall.

Graph 9

Graph 9: A 2021-2022 line graph with shaded areas. The line shows the proporition of developer US dollar bonds that have a yield greater than 50 per cent. The line increases substantially around September-October 2021. The shaded areas show the largest contributing coefficients from the regression described in Appendix A at a given point in time. Until August 2021 the largest coefficient is cash, then current assets from August until October, then total liabilities until December, and whether the developer is a state-owned enterprise for the rest of the graph.

These developments made it more difficult for private developers to obtain additional financing, further increasing their likelihood of defaulting. Consequently, the vast majority of major listed private property developers, and more than half of all major listed developers (weighted by 2020 assets) have now either already defaulted or been under severe financial stress (Graph 10).

Graph 10

Graph 10: A horizontal bar graph showing developers' assets in 2020 by whether they are private or state-owned, and whether they have defaulted on a bond or have been under stress (defined as experiencing yields greater than 50 per cent). The graph shows that around a third of private developers have defaulted and most of the remaining developers are under stress. Only a small proportion of state-owned developers have been under stress and none have defaulted.

Developers continue facing difficulties in financing themselves as funding markets have dried up and demand for presales has softened. Over the past six months, developers under stress (or that have already defaulted on another bond) have had considerable difficulties issuing any bonds, leaving them with significant financing gaps as sizeable offshore bonds mature. By contrast, state-owned developers have largely been able to maintain access to onshore bond markets (where they have historically obtained a large share of their bond funding) (Graph 11). These recent developments suggest that the property development industry is likely to become smaller, more concentrated and more state-led than during the high growth boom of recent decades, as many defaulting developers are likely to either go through debt restructuring, be acquired by healthier developers or go bankrupt.

Graph 11

Graph 11: A two-panel time-series line and bar graph since 2021. Both panels show issuance and maturity of bonds as bars, split by whether they are onshore or offshore bonds, and net financing as a line. The top panel shows this for stressed and defaulted developers and the bottom panel shows this for other developers. Stressed and defaulted developers have had negative net financing for some time, driving by offshore debt maturities, while other developers have had mostly positive net financing, driven by onshore debt issuance.

Restructuring in the property development sector is likely to be a significant drag on China’s economic growth and place additional strains on local government finances for some time. The restructuring process, even when successful, is typically lengthy, during which a developer’s operations slow considerably. Developers undergoing restructuring are likely to demand less land for new properties even as they continue to sell properties close to completion. This slowdown is already evident through slower property sales in 2021 for defaulted developers. Recent evidence also suggests that local governments, which derive a large part of their revenue from land sales to developers, have sought to replace the reduced demand with purchases by local government financing vehicles (Bloomberg 2022a) – a kind of state-owned investment vehicle, which often engage in infrastructure development. As such, these land purchases are likely to add to ongoing concerns about the viability of these vehicles and the financial stability risks they pose (Holmes and Lancaster 2019; RBA 2022).

Graph 12

Graph 12: A two-panel vertical grouped bar graph. The left panel shows property sales split by year (from 2018 to 2021) and by whether a developer is defaulted, stressed or other. The right panel shows the same but for land purchases. The left panel shows that property sales have increased over time for stressed and other developers, but dipped in 2021 for defaulted developers. The right panel shows that land purchases have decreased over time for all developers, especially stressed and defaulted developers.

Along with direct effects on employment and activity in the property development sector and its suppliers, and the consequences to banks of holding non-performing property developer loans on their balance sheets, developer stress may also reduce confidence among home buyers, leading to a decline in housing prices. In July 2022, reports circulated that some home buyers who had not yet received properties they were presold had stopped making mortgage repayments on the properties (Bloomberg 2022b). Maintaining buyer confidence in the housing market even as major property developers remain under financial stress remains a key challenge for the authorities.

Conclusion

China’s property development sector came under severe financial stress in 2021. This followed the accumulation of risks in the sector over a number of years – the result of long-term economic and demographic trends that led to high demand for housing, and the highly leveraged business models that developers adopted to expand rapidly in response to this demand. After the authorities implemented regulation aimed at reducing leverage in the sector, a series of high-profile property developer bond defaults followed. These defaults exacerbated pressure on healthier developers as markets became increasingly concerned about developers’ solvency risk, rather than only liquidity risk, and the likelihood of state support. The sector is likely to consolidate and become more state-led as a consequence. The ongoing funding stress and difficulty for developers in delivering presold projects poses substantial risks to the real estate market as a whole, and has increased the risks surrounding local government financing.

Appendix A

Data

This analysis uses data on US dollar bonds that were outstanding between the start of 2021 and June 2022 issued by property developers listed in Hong Kong, Shenzhen or Shanghai, along with balance sheet information from the developers.

Bond price data are from Refinitiv, compiled by matching the International Securities Identification Number on bonds from issuers classified as property developers by the Wind Financial Terminal. Developers’ balance sheet data are from their 2020 annual reports. An indicator for whether a firm is a state-owned enterprise is derived for listed developers that had greater than CNY60 billion in assets in 2020 from a combination of the Wind Financial Terminal and developer websites; mixed-ownership firms are classified as state-owned enterprises where the central or a local government has a controlling voting stake. This yielded a sample of 63 developers with a total of 321 bonds outstanding over the period.

Model

The following regression model was used to quantify the effect of balance sheet characteristics and state-ownership on bond yields:

pricejki =β0+β1ln(Cash2020)k+β2ln(CurrentLiab2020)k +β3ln(CurrentAssets2020)k +β4ln(ΔNetProfit20202019)k +β5ln(TotalLiab2020)k+β6SOEk+ϵjki

Where:

  • j is the date, k is the developer and i is the bond
  • the variable price is the price of the bond
  • Cash is the firm’s cash and cash equivalent holdings
  • CurrentLiab and TotalLiab are current and total liabilities
  • CurrentAssets is the firm’s current assets
  • ΔNetProfit20202019 is the percentage change in the firm’s net income between 2019 and 2020
  • SOE is an indicator variable for being a state-owned enterprise.

Logs were taken for each variable other than the indicator variable so that the coefficients are interpretable as the effect for a 100 per cent change in a variable. Standard errors were clustered at the level of the developer to account for multiple bonds issued by the same developer.

The regression results can be found in Table A1.

To produce the coefficient estimates used in Graph 9, regressions were run over a rolling 60-day window throughout the sample, producing coefficient estimates on data from the previous 60 days.

Table A1: Developer Bond Price Regression Results
  Coefficient
(standard error in parentheses)
ln(Cash2020) 6.300
(9.610)
ln(CurrentLiab2020)jk 1.036
(22.523)
ln(CurrentAssets2020)jk 16.451
(26.338)
ln(ΔNetProfit20202019)jk 4.480
(1.905)
ln(TotalLiab2020)jk −20.159
(30.009)
SOE 28.951
(7.596)
Constant −27.878
(77.958)
R2 0.1551

Sources: Wind; Refinitiv; RBA

References

Adams N, D Jacobs, S Kenny, R Serena and M Sutton (2021), ‘China’s Evolving Financial System and Its Global Importance’, RBA Bulletin, September.

Bloomberg (2021a), ‘China Fortune Land Becomes First Victim of “Three Red Lines”’, Bloomberg News, 2 March.

Bloomberg (2021b), ‘Sichuan Languang Defaults on US$139 million Bond Payment as Debt Woes Spread among Chinese Developers’, South China Morning Post, 13 July.

Bloomberg (2021c), ‘Four of Hong Kong’s Top Banks Halt Some Evergrande Mortgages’, Bloomberg News, 21 July.

Bloomberg (2022a), ‘Property Crisis Means Chinese Cities Sell Land to Themselves’, Bloomberg News, 26 January.

Bloomberg (2022b), ‘Mortgage Boycott Shows How Deep China’s Property Crisis Goes’, Bloomberg News, 19 July.

Chongqing Municipal People’s Government (2021), ‘关于房地产预售资金监管,重庆出新规 [Regarding the supervision of real estate pre-sale funds, Chongqing has issued new regulations]’, 22 September.

Ding D, Y Huang and Y Zhou (2019), ‘China’s Offshore Corporate Dollar Bonds’, in A Schipke, M Rodlauer and L Zhang (eds), The Future of China’s Bond Market, International Monetary Fund, Washington DC, pp 357–379.

Glaeser E, W Huang, Y Ma and A Shleifer (2017), ‘A Real Estate Boom with Chinese Characteristics’, Journal of Economic Perspectives, 31(1), pp 93–116.

Hale T (2021), ‘China Evergrande Shares Tumble After Court Freezes $20m Bank Deposits’, Financial Times, 19 July.

Holmes A and D Lancaster (2019), ‘China’s Local Government Bond Market’, RBA Bulletin, June.

Huang Y, D Yi and W Clark (2020), ‘Multiple Home Ownership in Chinese Cities: An Institutional and Cultural Perspective’, Cities, 97, Art 102518.

Jim C (2020), ‘China Evergrande Shares, Bonds Suffer Sell-off on Cash-crunch Concern’, Reuters, 25 September.

Kemp J, A Suthakar and T Williams (2020), ‘China’s Residential Property Sector’, RBA Bulletin, June.

People’s Bank of China (2018), ‘中国金融稳定报告 [China Financial Stability Report]’, November.

RBA (Reserve Bank of Australia) (2022), ‘The Global Financial Environment’, Financial Stablity Review, April.

Reuters (2021), ‘Sales in Two China Evergrande Projects Suspended by Chinese Authorities’, Reuters, 20 July.

Rogoff K and Y Yang (2021), ‘Has China’s Housing Production Peaked?’, China and World Economy, 29(1), pp 1–31.

Tianjin Municipal People’s Government (2021), ‘预售资金监管办法的通知 [Notice of pre-sale funds supervision measures]’, Media Release No 11, 30 July.

Tudor-Ackroyd A (2020), ‘Bankruptcy Case Casts Doubt on US$96 billion of Offshore Chinese Bonds Claims, Puts Pressure on Financial Products Skirting Capital Controls’, South China Morning Post, 17 May.

United Nations (2021), ‘Urban Population (% of Total Population) – China’, World Urbanization Prospects: 2018 Revision, World Bank site. Available at <https://data.worldbank.org/indicator/SP.URB.TOTL.IN.ZS?locations=CN>.

Wang S (2011), ‘State Misallocation and Housing Prices: Theory and Evidence from China’, American Economic Review, 101, pp 2081–2107.

Wang J, M Niu, T Shen and G Cross (2020), ‘Developers Face New Debt Limits as Property Crackdown Continues’, Caixin Global, 2 September.

Yu X (2021), ‘China’s Evergrande Moves to Calm Nervous Investors’, The Wall Street Journal, 7 June.

Yu X and C Jim (2022), ‘China Gives Property Firms Easier Access to Escrow Funds – Report’, Reuters, 11 February.

Zhang X (2020), ‘论商品房预售资金的法律监管 [On the Legal Supervision of Pre-sale Funds of Commercial Housing]’, Housing and Real Estate, pp 13–14.

Source: Reserve Bank of Australia [2021]