Sheung Shui is a residential neighborhood over an hour away from Hong Kong island, so remote that no one I know would make the commute more often than weekly. This part of the New Territories is far from most of what makes Hong Kong attractive, but its main advantage is in being only one train stop away from either of the two main border checkpoints with the land of Renminbi: Lo Wu and Lok Ma Chau. Past the second of these crossings, and only a few minutes away, is the Futian District of Shenzhen, a heavily developed residential area very well located and connected to the commercial activity of Shenzhen.
So given not only how close these two places are, but also how Sheung Shui seems to be a relative backwater to Hong Kong while Futian seems to be a prime central location in Shenzhen, how is it that a square foot in a new building on this side of the border still seems to average more than double the price of a square foot on the side where all the development and business seems to be happening?
My comparison is extremely quick and dirty, looking at only two factors across 50 properties on each side, and feel that even though the two samples are not exactly apples-apples, the differences are not as many or as difficult as I would have expected:
- Centaline’s CentaMap is an excellent tool for sorting data either on listings or recent transactions on properties in Hong Kong, but I have not been able to get it to work as well for Shenzhen. The underlying Centadata in Shenzhen shows recent rentals, but does not seem to show transactions.
- For Shenzhen I found data points more easily by looking at for-rent listings on Centaline’s CentaNet – this means I am comparing the “offer” side of the Futian market with the “last traded” prices in Sheung Shui, which should bias prices in the mainland’s favor.
- Many listings show prices for both sale and rental of the same property. This makes both Hong Kong and Shenzhen very easy markets to compare rental yields.
A brief 2-photo aside on the HK-SZ border: It is difficult to exaggerate the differences between the Hong Kong Special Administrative Region side and the People’s Republic side of the border, whether viewed from an aerial map, felt in the crowds crossing the bridge, or smelled in the train stations on either side. Despite their continued claims of convergence, and the unparalleled growth I have seen in Shenzhen over the past 15 years, there is still something that seems “developed” on this side that just doesn’t seem anywhere near north of the border.
This development pattern might be similar to others along similar borders with such a high GDP gradient – this being one of my favorites of the US-Mexico border:
Back to the property gradient between Hong Kong and Shenzhen, using the 50 data points on each side, and roughly came up with the following table and bullet observations:
|Futian, Shenzhen||Sheung Shui, Hong Kong|
|Average Price of a new 600′ unit:||HK$2.3mio||HK$4.6mio|
|Average Price per SqFt of a new 600′ unit:||HK$3,800||HK$7,600|
|Average Annual Depreciation per SqFt (cross-sectional)||HK$30||HK$84|
|Average size premium per SqFT||+HK$0.77||-HK$1.97|
The charts below further show how these differences may be enough to be statistically significant, even if not apples-apples, with some of the following caveats:
- Futian had apartments far newer (not surprisingly) than the generally 20+ year old ones in Sheung Shui, but also had a far broader selection of larger apartments (surprisingly for me).
- Among the smaller Sheung Shui apartments, price per square foot seems to rise as units get smaller (indicating a floor on absolute apartment prices on the low end), while Futian shows the more typical HK pattern of larger apartments costing more per square foot. If I had more data, I would expect a U-shaped curve of price per square foot as a function of absolute size, with a bottom around 500-600 sqft, and a spike at the very high end.
- The twice as valuable HK buildings also seem to depreciate at about double the rate per square foot as they age, but with more data I would be able to show this usually isn’t linear.
One common explanation for the difference in property prices is the premium placed on the Hong Kong legal system and its better reputation for preserving property rights than the mainland’s. If this were true, we would expect rental yields to be higher on the Shenzhen side of the border than on the Hong Kong side, but in my sample rental yields in Sheung Shui averaged 3.5% while Futian’s were only around 2.5%. Despite the Hong Kong property market’s reputation as a destination for wealthy Chinese keen to stash they money offshore, more of this money still seems to be driving up prices of PRC apartments relative to the demands for people to actually live in them.
I would very much like to better quantify the value property investors and users put in being on either side of this and many more borders in Asia, both economically and legally, but as in any property market the challenge will be in getting more good data. Meanwhile, the relatively low yields and difficult legal system tell me Shenzhen is a great market to study academically, but not one I will be rushing into for value investments any time soon.