Goldman with a note that blows up its own ‘long commodities’ recommendation:
Q: What has led to the current stresses in China’s property market?
A: The authorities imposed new regulations over the past year that significantly constrained property-related borrowing, particularly by more levered developers.
There are 4226 words left in this subscriber-only article.
Get your first month for $1
China’s property sector has long been a major engine of growth for the economy. As the country urbanized in recent decades and household incomes improved, housing construction boomed, growing at a double-digit annual pace over the 2000s and even faster in the immediate aftermath of the global financial crisis, when policymakers eased fiscal and credit policies to offset downside pressures from abroad.
However, this growth increasingly came at the cost of higher leverage. Household mortgage debt increased from about 10% of GDP before the global financial crisis to 33% now. And property developer debt has increased sharply as well. Our property team estimate total liabilities for the property industry increased from RMB 12.5tn in 2009 to RMB 85.7tn in 2020.
As part of their efforts to manage potential risks in the economy, Chinese regulators instituted the “three red lines” (3RL) regulation last year, summoning a dozen of the largest developers to report monthly on three metrics that would need to be met to increase debt: 1) a total liability to total asset ratio (excl. presales deposits) below 70%, 2) a net gearing ratio below 100%, and 3) cash to short-term debt ratio above 100%. For most developers under GS analysts’ coverage, at least one of these metrics was binding as of the initial announcement, implying a constraint on borrowing and a need to slow land acquisition significantly (from ~20% yoy growth overall in 2017-19 to -10% yoy in 2021-22). Developers have made progress in meeting these targets but most are still crossing at least one “red line”.
At the end of 2020, the authorities also set constraints on individual banks’ property lending, specifically 1) a maximum share for property-related financing in total loans (varying from 22.5% for regional banks to 40% for large banks), and 2) a further constraint for mortgages specifically (ranging from 17.5% to 32.5% of the total loan book). Given some grace period to bring numbers into compliance, our banks analysts estimate this regulation will slow mortgage growth to the high single digits from 12-16% in recent years. (Efforts to constrain shadow banking activities, including trust lending, represent yet another constraint.)
As one of the fastest-growing and heavily-borrowing developers in China more than quadrupling revenues over the five years before the pandemicEvergrande has felt some of the most intense pressure from the new borrowing constraints. At the time of our property analysts’ initial assessment a year ago, Evergrande missed all three “red lines”. Activity has slowed sharply in recent months, the value of its shares has fallen, and prices for offshore-listed bonds fell significantly over the summer and now hover in the 20s (implying that market participants have essentially priced in a default). Credit spreads for developers, particularly lower-rated firms, have risen very sharply (Exhibit 1). The high level of uncertainty about how in-progress projects will be completed has reduced sales sharply in recent weeks, compounding the financing pressures on developers, and leading to intense market focus on how the government will limit further spillovers to the broader property sector and the overall Chinese economy.
Q: How would a potential default of Evergrande affect China’s financial system, and what would be the risks to overall financial stability?
A: We think the risk of a major financial crisis on the back of the Evergrande situation is low for three reasons.
- First, it appears that banks’ exposure to Evergrande is not very large relative to the size of system loans or capital. Our credit strategy team estimates roughly $60bn bank and trust loans borrowed by Evergrande, and media reports suggest about $33bn bank loans. These are small compared to Chinese banks’ total loan book of $27tn. This is not to say that no financial institutions would face challenges. Evergrande is the largest owner of Shengjing Bank, which had RMB 1tn total assets as of H1 2021 and reportedly had significant exposure to Evergrande.
- Second, the Chinese banking system is in a much healthier place compared to a few years ago thanks to deleveraging and de-risking efforts. As our banks analysts have explained, shadow banking and wealth management products (WMP) are likely the most vulnerable assets for banks in response to property market risks. But WMP balance has declined 11% since 2016 and property financing accounts for only 8.6% of the WMP balance now.
- Third, unlike the US subprime crisis which featured loose lending standards and even mortgages with negative amortization, Chinese homebuyers need to make at least a 30% down payment for first-home purchases and much more for second-home purchases (In the past some cities reduced down payment requirements to 25% or 20% for first-home purchases during local housing downturns to boost demand and stabilize house price). Additionally, mortgages are recourse loans in China, which implies that banks can go after borrowers’ other assets in the event of mortgage defaults. These differences reduce the risk of a vicious cycle where house price declines trigger mortgage defaults and foreclosure sales which in turn lead to further house price declines.
Although there remain significant uncertainties, the nature of the Chinese financial system and mortgage markets, the fact that the ongoing “three red lines” and restrictions to property lending are initiated by the government, and the broader deleveraging and de-risking push of the past five years, suggest the risk of the Evergrande situation causing a broader financial crisis in China is relatively low as long as policymakers moderate pressure to avoid liquidity problems across the entire industry. However, there are meaningful risks to Chinese economic growth, as discussed below.
Q: How large is the contribution of real estate to China’s GDP?
A: There is no consensus on this and precise estimates depend on one’s definition (e.g. residential only, or broader property sector), whether the focus is on new construction flows or the entire housing stock, and whether the upstream activities linked to construction (e.g. cement) are included.
According to the National Bureau of Statistics (NBS), the construction industry’s share of GDP is 7%. Property construction in turn accounts for roughly 70% of total construction (with the rest being infrastructure), translating into ~5% of GDP (=7%*70%). In addition, not all properties are residential (e.g., commercial real estate), so the direct contribution of building new residential housing to GDP is even smaller than 5%. Real estate services are another 7% of GDP. But much of real estate services are building management and services associated with existing rental housing, neither of which should be significantly impacted by new property sales. Therefore, although the construction and sales of commodity housing is very important to the broader economy, the direct and indirect contribution of residential housing construction to GDP appears to be in the teens rather than the 20%+ numbers cited in the media and other studies.
Q: How did you estimate the impact of Evergrande and China property downturn on GDP?
A: To quantify the impact of property sector slowdown on GDP growth, we considered impacts via real activity and financial conditions separately. We recently estimated a drag of 1.4% on 2022 GDP in our base case scenario (land sales and housing starts down 15% and property sales and house prices down 5%), with an impact up to 4.1% of GDP in a severe scenario where these housing declines are doubled.
On real activities, direct effects on GDP growth include:
- Subdued construction activities given declines in volume of property new starts and weaker real estate services due to lower property sales.
- Land sales revenue is also an important source of financing channel for local governments, so reduction in land sales revenue also implies less local fiscal spending.
- Another direct impact is through the wealth effect on private consumption, though the household consumption response to property prices has been relatively moderate in China.
Indirect effects of property downturn on real activity include the negative spillover to upstream products and services. Based on the World Input-Output Table, sectors that could face the most significant headwinds are non-metal minerals (cement and glass), wood products, mining sector and metals manufacturing. (See here for an example of analysis using the Input-Output table to identify sectors most affected by construction activity.) A slowdown in the property sector also weighs on financial services. Downstream moving-in related products such as furniture and home appliances may decelerate along with slower property completions. (See here for discussions on the link between housing completions and the consumption demand for moving-in related goods.)
Financial conditions also play an important role in affecting growth. In the case of equity market selloff, credit spreads widening, and/or deceleration in credit growth, our Financial Conditions Index (FCI) tightens and generates a negative impulse to growth. The peak of the drag typically takes place 2-3 quarters after the FCI tightening. (See here for more details on how we construct our China FCI measure and empirical evidence on the impact of FCI on growth.)
Q: How much would the property market downturn affect consumption spending in China?
A: We expect the effect to be small, given ambiguous evidence on housing wealth effects. We take into consideration two mechanisms in assessing the effect of a property market slowdown on consumption: (1) the wealth effect and (2) the demand for moving-in related goods. In our base case where property prices decline 5% but completions increase 10% next year, the negative impact from the property market to consumption operates mainly through the wealth effect. But in China, evidence on the link between housing and household consumption has been less than straightforward at times. On the one hand, the homeownership rate is high (90% based on 2011 China Household Finance Survey) and the wealth implications of higher house prices can be large. On the other hand, some empirical literature previously found Chinese households tend to reduce consumption and increase saving when facing rising house prices, mainly due to a strong incentive to invest in properties amid strict borrowing constraints. The investment motive may have contributed to more muted consumption response to increases in property prices (Exhibit 2). It is also consistent with our previous estimate of a small equity wealth effect and the finding that household consumption is more sensitive to cash flow changes than to changes in wealth. Overall, we do not expect the property market downturn to become a large drag to consumption, barring the bearish scenario with considerable spillovers from Evergrande and a significant slowdown in the broader economy.
Q: How could the Evergrande stresses and a China property downturn affect economies outside China?
A: China property is a big sector, but its import intensity is relatively low, so the real activity (import) spillovers from China residential construction to most economies should be fairly modest.
To assess the effects of a China property sector downturn on other economies, we can distinguish between spillovers via the real economy (generally trade) and via financial conditions. Starting with the real economy, a weaker property sector and Chinese economy would reduce Chinas imports from other countries, transmitting a part of its negative growth impulse abroad. Using input-output tables to trace the foreign exposure to China’s residential construction sector suggests ~$250bn of foreign value-added split roughly equally into three main groups:
- Commodities (with knock-on effects to commodity prices). Weaker demand in commodity-intensive sectors of the economy will reduce both the volume and (given Chinas outsized role in international commodities markets) likely the price of its commodity imports.
- Manufactured goods. Much of the manufactured goods attributable to construction end demand are also commodity-intensive, such as wood and metal products. Although there are also modest imports of construction-related equipment and in related sectors home furnishings, import intensity in these sectors is relatively low. (Beyond the construction sector itself, any negative effects on household income and employment from slower construction activity could weaken consumer goods imports on the margin.)
- Services. A large part of China’s overall services imports are really Chinese residents’ travel and tourism, which are already at extremely low levels because of Covid-related quarantines. But the construction sector does utilize foreign business services such as international transportation of goods, financial services, and others.
In previous work, we quantified potential spillovers from a slowdown in Chinese growth via imports to the rest of the region. We found that a one-percentage-point growth slowdown in China could slow growth as much as 0.2-0.3pp a number of smaller open economies in the region via real economy channels (Exhibit 3).4 But the residential construction sector is actually less-import-intensive than the Chinese economy as a whole; even a regional commodity-intensive exporter like Australia has only slightly over 1% of its GDP attributable to exports to China’s residential construction sector, on our estimates. The bottom line is that in our base-case scenario of a modest contraction in Chinese housing activity next year, damage to regional economic activity (especially in larger economies outside China) should be limited.
Q: What about financial spillovers to markets outside China?
A: There are several potential financial channels through which a property downturn could affect markets (and thereby financial conditions and growth) elsewhere, with equities likely the most important.
- Credit. China property firms have approximately $210bn of USD debt offshore (of which ~$20bn is from Evergrande), held by a range of investors. While potentially significant from a returns perspective for individual funds or investors, we believe a large part of this would likely be held by long-only investors and would not expect losses here to have major macro implications. Foreign banks lending to Chinese firms varies greatly by jurisdiction; a recent study by researchers at the US Federal Reserve found very low US exposures to borrowers in China and Hong Kong, but highlighted banks in the UK and Singapore as having more significant exposures (however, it should be noted this study looked at overall exposures, not those specific to residential property).
- Equities. Although our equity strategists classify China property as the largest asset class globally, and see it impacting ~15% of listed company profits in China either directly or indirectly, spillovers to global equity markets appear to have been modest so far. Global equities are down about 3% from their peak in early Sept (MSCI World) on concerns about China’s growth and perhaps also to some degree on incrementally more hawkish news on Fed tapering timetable. The moves that are likely most attributable to China property concerns are those involving mining/commodity firms exposed to China-centric commodities, e.g. iron ore (prices have plunged to less than half their peak level this summer, Exhibit 4).
- FX. If weaker China growth results in a weaker CNY, then on net this would tighten financial conditions elsewhere (as currency depreciation is zero-sum). We do not expect a large currency depreciation in the current episode, as current account fundamentals are strong with a surplus of over 2% of GDP and significant structural bond inflows related to index inclusion (China’s inclusion in the FTSE WGBI index begins next month). Chinese policymakers also appear to have better control of the capital account than they did during the 2015-16 episode.
The net effect of a significant slowdown in China is generally to tighten financial conditions elsewhere. Month-to-date we have seen our GS Financial Conditions Indexes (FCIs) tighten 24bp in the US and 7bp in the euro area mostly driven by lower equity prices. Japan’s FCI has tightened 3bp on a stronger yen, though Korea’s and Australia’s FCIs have actually eased (with weaker FX more than offsetting stronger equities). The bottom line is that so far, these changes are quite modest outside the US (and in that case, may be attributable to a moderately more hawkish shift in Fed tapering views in recent weeks).
Q: Why do you see some urgency for the government to communicate a plan?
A: In our view, the current level of uncertainty is contributing to unsustainably high funding costs for developers, very low levels of sales activity, and a halt in activity at many of Evergrandes current projects.
Our relatively sanguine view on the risks Evergrande may impose on China’s financial stability is predicated on the baseline assumption of limited contagion effects. This in turn requires clear communication from the government on the plan to contain any spillovers to broader property market via sales uncertainty, developer financing, or other channels. There are three reasons behind our thinking. First, unlike non-durable goods consumption, confidence plays an important role in households’ decision to purchase homes. When developers’ ability to deliver apartments is in question, presales of properties (an important funding source for developers) may fall (Exhibit 5). Second, housing markets tend to exhibit significant momentum. Once expectations of future price movements are set, it takes months, if not years, to reverse. Third, with local Covid outbreaks resurfacing from time to time, consumption and service industries are unlikely to rebound sharply. At the same time, exports appear to have stopped rising in volume terms in recent months. Production inputs, from semiconductors to certain raw materials, face severe shortages and constrain production. With this backdrop, a sharp deceleration in the property market could exacerbate and amplify the downward pressure on economic growth and labor market. For these reasons, we think it would beneficial for the government to communicate a plan soon to circuit-break the worries surrounding Evergrande from spreading to the broader property market.
Q: How do you think the Evergrande situation will ultimately be resolved?
A: Given the size and complexity of Evergrande, we do not know how the situation will be resolved.
In the case of resolving a potential default, one possibility could be to find third parties, or ‘white knights’, to take over the onshore property business as part of a debt and equity restructuring. This could be done nationally or at the provincial or city level, potentially involving a consortium, and could include government-linked entities. The latest media reports point to some involvement of local governments and SOEs, although the situation is fluid and uncertainty is great. Our credit strategy team’s analysis of previous China SOE default cases suggests the outcomes are by no means consistent across episodes. With 800 property projects in over 200 cities, any resolution for Evergrande is also likely to be significantly different from the government’s approach with regional banks (e.g., Baoshang) examined by our banks analyst.