In December 2021, China’s real estate market came to a sudden halt. In the run up, developers had been aggressively expanding their operations while also taking on massive amounts of debt.
Though the collapse was precipitated by the COVID-19 pandemic, it was exacerbated by reckless housing development.
The result was an unsustainable business model that collapsed when investors stopped paying off their debt totaling $300 billion USD (2% of China’s GDP).
More than any other sector in China, real estate companies control both the largest number of properties (30% of China’s GDP) and the biggest amount of debt ($300 billion USD; 2% of China’s GDP).
Recently, China’s property investors took note of the rapid pace of growth — especially in 2017, when housing prices increased by more than a third.
Using a system called “pre-sale,” massive real estate developers like Evergrande encouraged home buyers to pay off their mortgages before construction was finished.
Investors are still trying to digest the magnitude of the situation.
Think back to the 2008 financial crisis. Events like these have a lasting impact on markets, governments, and economic stability.
Now these investor payments (loans) will remain outstanding, and China’s major financial institutions are at risk. Those pre-sale purchasers are left with half-built houses and no timeline to completion.
On-the-ground accounts painted a less optimistic picture than what developers were promoting.
To look for physical changes to a landscape over time, we use non-coherent change detection (NCD) algorithms applied to synthetic aperture radar (SAR) data.
In the span of the last three years (2020-2022), the area of change in Xiangtan has decreased by 25.5% compared with the previous three years (2017-2019). This is proof that construction has stalled following the pandemic.
There are several ways Ursa Space can visualize these changes, including kernel density heatmaps using SAR data.
The kernel density heatmap below highlights higher values both as they cluster together and also based on their measured area in square meters. Brighter areas mean both clustering of change points and larger areas of change.
By running two instances of NCD for two different time ranges over Xiangtan, Hunan Province – first from 2017 to 2019 then from 2020 through November 2022 – we can see decreases or increases in rates of development:
The burst property bubble is not unique to China. Many countries have seen property market crashes in the past, and in today’s news cycle we see reports of housing slumps and deep freeze predictions on the US market.
However, the extent to which China’s problem snowballed is what made it particularly unique. When left unchecked, developers can promote optimistic outlooks which keep pre-sale loans thriving.
An alternative solution is to use SAR and NCD data on a regular basis to hold lenders and developers accountable as construction plays out.
If China had used these tools, warning signs of stalled construction would have been seen long before companies like Evergrande publicly acknowledged the problems and defaulted on debt.