The supply side of housing (number of units per quarter; log annual difference; 3Q2002-4Q2004):
In the top chart, the top dotted line is the average from 2002-2008 (about 49.2k per quarter) while the bottom dotted line is the average since (about 24.4k per quarter) . In the bottom chart, the corresponding averages are 5.7% and 2.2%.
So, a big part of the reason why house price inflation increased over the past decade or so is due to the shortfall in supply.
Technical Notes:
- Data from various reports published by the National Property Information Centre
- Note that the above data series excludes service apartments to maintain consistency (2016 NAPIC data does not include service apartments)
Posts
Do we have any idea of the distribution of the supply? High-end vs mid- and low-end?
ReplyDelete@Roger
ReplyDeleteNAPIC only publishes by either by average price or by quantity of type (terrace, condo etc) unfortunately. The transaction data is available, but it does not give you any idea of outstanding supply.
And all this data is a pain to work with, because its mostly in pdf format.
However, the Property market report has some data (value and no of units) on unsold completed properties. The following are the average prices for Selangor:
1H15: RM537k
2H15: RM688k
1H16: RM874k
I think we can infer that most of the unsold supply is at the high end.
Johore has the biggest current overhang (20% of Malaysian total):
1H15: RM329k
2H15: RM605k
1H16: RM653k
Depending on how one classified 'Residential', nowadays, developers loves to build 'Serviced Apartments' such as SOHO on Commercial land. These units are used as residential units but classified as commercial. More often than not, the utility bills are under commercial, thus, does the chart include these SOHO or exclude?
ReplyDelete@James
DeleteNAPIC actually removed SOHO and service apartments from the data, for exactly the reasons you describe.
However, the back series is still available, up to 2015, and it shows a 50% increase from 2012 (from about 50k total units to 75k units). Nevertheless, both together are about 1.5% of the total stock, and less than 10% of the total completions during that period. It doesn't invalidate the overall narrative, and adding the SOHO/SA numbers back in doesn't make a significant difference. We're still looking at a substantial drop in construction, relative to the previous decade.