The phenomenal sales growth that Chinese developers achieved in recent years has relied partly on joint-venture projects. Used as a means to participate in large-ticket land auctions and to climb sales rankings, these collaborations have quickly become the norm in China. S&P Global Ratings believes these structures can convolute an issuer's results, and may obscure the key metrics we use to determine ratings.

Over the past three years, developers' average attributable sales ratio has dropped to 69% from 78%. The ratio is the portion of sales attributable to the company after adjusting for their shareholding in jointly controlled entities (JCEs). The ratio's decline shows that an increased part of sales is generated by entities with multiple partners--in other words, not all the sales belong to the rated developer.

Meanwhile, the unconsolidated portion of their contracted sales climbed to 35% on average in 2019, compared with 23% in 2017. This shows that use of JCEs has put an expanding portion of developers' sales off balance sheet (see chart 1).

Chart 1

In simple terms, a rising portion of a developer's contracted sales--derived through JCEs--is failing to materialize as consolidated revenue.

We rely on developers to voluntarily disclose much of the important details about their JCE involvement. The details are unaudited and unverified. This compounds our difficulty assessing in assessing developers' true credit metrics.

We see genuine benefits derived from using JCEs. Through such partnerships, developers can reduce project risk and spread out their financial burdens, while lowering land costs as competition is reduced during land auctions. JCEs can support developers' credit profile, contributing to growth using reasonable leverage.

However, we believe the operational and financial risks of JCEs are rising. These risks can be summarized in three main categories.

• Off-balance-sheet debt and leverage can be significant, depending on how developers provide financial guarantees to unconsolidated JCE debt. The debt may weigh on developers' credit profiles once it is proportionally consolidated into see-through metrics.

• There is potential for hidden debt. This is a particular risk for JCEs that involve minority stakes held by financial (non-developer) investors, whose positions may be disguised debt with de facto recourse to the developer.

• Mismanaged JCEs can depress developers' revenue recognition. A portion of contracted sales reported may not ever show up in their financials as revenue. This could be due to execution inefficiencies, or poorly applied consolidation treatments.

These risks may affect ratings. We have already taken negative rating actions on three entities--actions that referenced their use of the structures (see table 1).

Table 1

Our findings showed that 61% of rated developers derived more than 30% of contracted sales from unconsolidated JCEs in 2019. This was just about double the portion of Chinese developers that used JCEs this way, in 2017.

While this uptrend applied to basically all the rated developers, the growth was more significant for developers rated 'BB', followed by 'B' rated firms--in other words, entities whose growth aspirations are generally matched by their willingness to use aggressive structures (see chart 2).

Chart 2

How We Assess These Structures

While JCE use is increasingly relevant to analyzing a developer's exposures, disclosure can be patchy.

To understand the contribution of JCEs to our rated developers' debt, income and sales, we assess see-through leverage through proportionally consolidating the debt and earnings of these off-balance-sheet JCEs onto developers' reported figures.

To arrive at our adjusted debt, we add back attributable JCE debt and subtract guarantees provided to avoid double counting. We also add back attributable EBITDA contributions from JCEs, but take out dividends received from JCEs to arrive at our adjusted EBITDA.

We take this approach because dividend payments in most cases do not match cash flows up-streamed from JCEs. When JCEs pass cash to the partners, they do not always wait and go through the dividend channel (see chart 4).

In making these calculations, we compare the see-through metrics alongside the consolidated ones and pay particular attention to the potential effects of JCEs. Outlook triggers on the developers may also include metrics derived using both a consolidated and see-through treatments.

In general, we have observed that the size of off-balance sheet debt has not significantly deviated from the financial guarantees developers have provided to JCEs. As guarantees are audited and clearly disclosed, we believe these provide a proxy to gauge the magnitude of off-balance sheet debt (see chart 3).

Chart 3

Chart 4

We see a low risk of hidden debt being disguised as equity for these off-balance-sheet projects, given the partners in such entities are typically developers. We would expect such parties to co-develop these projects as equity partners. From the information provided by developers we rate, we estimate that roughly 80% of JCEs have other developers as partners.

Not All Off-Balance Sheet Debt Is Created Equal

In a well-managed JCE, the see-through leverage of developers is generally lower than the consolidated level. This is mainly because developers typically fund land expenditure themselves.

They borrow at the consolidated level and inject capital into their JCE. The JCE then pledges the land to secure construction financing, and develops the project with the funds. We have observed developers' consolidated debt-to-EBITDA ratios as being one to two times greater than the see-through debt-to-EBITDA ratios.

There are cases where consolidated leverage presents a worse picture than would a holistic, see-through profile. These developers typically generate high income from JCEs, and--in extreme cases--the attributable revenue from unconsolidated JCEs may be larger than the consolidated revenue.

A prime example would be where a developer funds most of its JCEs via capital injections rather than raising project debt. This could happen when JCE partners have notably different funding costs. The partner with the lower funding cost may be unwilling to borrow more expensive project debt. Under such situations, we put more emphasis on the see-through ratio.

To complicate matters further, a time lag may appear in off-balance sheet leverage. Developers may have incurred high debt at the JCE's project inception, but the profit contribution would likely be insignificant until later.

The mismatch will likely improve over a two- to three-year period as these projects bear fruit. That said, we normally take a more conservative view on project completion versus company estimates, due to the limited dealings these developers may have with JCEs.

There are other factors we must consider. For example, the size of the guarantees on JCE debt, the stage of development of these projects, and whether developers are able to consolidate these JCEs at the end of the project cycle may all alter the picture of their off-balance-sheet exposure.

Some Developers' Consolidated Leverage Is Understated

For those developers that do not provide financial guarantees to their JCEs (that is, their debt position is not visible through any public data) we find that see-through treatment makes an issuer's credit profile look less robust.

This is typical of state-owned enterprises (SOEs), given some have policies to refrain from offering such guarantees.

In such cases, their off-balance-sheet debt would not be captured as guarantees, nor in their consolidated metrics. Their see-through ratio of debt to EBITDA could be substantially higher once the JCE debt is included, affecting their credit profile.

Even though there is no direct recourse to the SOE without guarantees, we believe the state-owned developers would absorb these debts if their JCEs ran into financial distress. These projects may carry the developers' brand, and banks may view their support of the debt as implicit. Any nonpayment may dent the reputation of the SOE.

Some only consolidate their JCEs toward the end of the project cycle, which is yet another important complication. This is more likely to be the case when partners are little known passive investors. Such investors typically have no incentive to consolidate their JCE investments. This leaves the decision to consolidate or not consolidate the investments in the hands of the rated developer.

We notice that the proportion of guarantees provided by these developers is consistently high compared with their reported debt. This tells us they are heavily involved in unconsolidated JCEs. At the same time, we often see only a minimal share of the JCE's profit since they are opting to consolidate them once the projects finish.

When that happens, developers' leverage could be understated as revenue and earnings would be fully recognized due to full consolidation, while project level debt would have been largely repaid with sales proceeds by then.

Chart 5

While use of see-through ratios may change how we view the credit profiles of rated developers that heavily use JCEs, most developers' see-through leverage is largely similar to their consolidated levels. We believe that--barring exceptional cases discussed--leverage levels should converge as the projects move through the development cycle.

The Problem Doesn't Go Away Even When JCEs Are Consolidated

Although some JCEs are consolidated and their debts are booked on developers' balance sheets, risks may still emerge. This includes capital that was once classified as equity, resurfacing as debt.

Unlike off-balance-sheet JCEs where most partners are property developers, we understand anecdotally that about half of the partners involved in consolidated JCEs are not developers. Sometimes up to 70%-80% of the partners in JCEs are financial investors without controlling positions.

This matters because some of the financial partners may structure their investment to be de facto debt. Rated developers may be obliged to repay such investors with fixed returns.

In our view, JCEs involving other property developers, employee investment schemes, and minority shareholders arising from urban redevelopment projects or acquired projects are usually genuine equity partners. This may also be the case when well-regarded financial institutions are investing, which do not seek fixed returns from a project, and serve as real equity partners.

We are more skeptical when the financial partners are lesser known institutions, or are corporate (non-developer) or individual investors. When we come across such partners, we generally require more information to perform a deeper dive on those projects' return mechanisms. Developers only provide such information on a voluntary basis. They have on rare occasions acknowledged that they later reclassified the positions as debt.

The rise in non-controlling interests, and the increase in off-balance sheet JCEs are two sides of the same coin. When more partnerships were formed over the past years, the number of off-balance-sheet and consolidated JCEs both increased.

Some developers noted that their revenue growth would significantly lag behind their contracted sales if they kept most of these projects off balance sheet. As a result, more started to consolidate JCE projects. The percentage of non-controlling interests in JCEs reported on the balance sheets of rated developers, divided by their total equity, grew to 30% in 2019 from 16% in 2016.

Chart 6

More Sales, But Where Is The Revenue?

The widening gap between the growth in developers' total contracted sales and their revenue recognition is gaining more attention. In 2019, about two-thirds of rated developers' revenue fell below our expectations (see chart 7).

Chart 7

The rise of JCEs involves execution risk. Developers may need to rely on multiple partners for sales, construction, and delivery. The risk is higher if the partner is a weaker developer. However, even if the partners are all reputable, sizable outfits, risks remain. One or more partners may not prioritize joint projects given their small economic interest. These projects may take longer to finish than solely owned ventures, delaying revenue recognition. We saw this issue repeatedly in the 2019 results.

Such revenue slippage is exacerbated by the risk that the consolidation ratio (portion of total contracted sales that will be consolidated onto a developer's profit and loss statement) may turn out lower than the guidance provided.

As noted, some developers typically buy out their partners at the end of the project. However, we have seen cases where developers fail to acquire and consolidate at the project's conclusion. This may throw off developers' original revenue expectations, at least on a consolidated basis.

We believe we can estimate developers' consolidation ratios (see chart 8). This is derived from developer's reported contract liability, which represents proceeds accepted from homebuyers. If the amount of contract liability is small compared with their level of sales or revenue, this could be a sign that revenue disappointments may lie ahead.

Chart 8

The more our estimated consolidation ratio is below guidance, the higher the risk that revenue may fall below company guidance, and leverage may be greater than anticipated, in our view. This issue is especially apparent for 'B'-rated developers, implying again there could be more revenue letdowns in future years (see chart 9).

Chart 9

The JCE structure also raises questions about the genuine level of cash flow attributable to a rated entity. If developers continue to expand their use of JCEs with a declining attributable ratio, even if they consolidate such projects, developers' true cash generation may undershoot reported contracted sales or revenue. Some part of the reported sales or revenue may belong to other stakeholders. This may eventually translate into weaker earnings, higher debt, and a greater need to borrow.

JCEs Are Here To Stay--We Need To Reckon With Them

Although we use see-through analysis to assess developers' off-balance-sheet leverage, and qualitatively examine their non-controlling interests, our approach heavily relies on voluntary disclosure. As investor awareness grows, we believe developers will gradually become more willing to improve public disclosure and increase transparency to all stakeholders.

Developers undoubtedly will continue to rely on JCEs for growth, especially as sector growth eases and economic headwinds gather in the wake of the pandemic. Through these projects, they can lighten their land expenditure and improve their geographic and product diversity.

Competition is intense and developers, even market leaders, sometimes have no choice but to collaborate to get a stake in prime projects. But poorly managed JCEs may do more harm than good. If projects turn unprofitable, cracks could show from exposure to contingent liabilities and damage their financials and brand image alike.