Thursday, February 16, 2012

Housing Prices and the Credit Impulse: A Singaporean Analysis

According to latest reports on the Singapore property market, the number of private homes sold soared to a 14-month high in January 2012, indicating that housing demand remained buoyant, contrary to expectations of a slowdown in housing market activity due to the implementation of ABSD (additional buyer stamp duty) in December 2011. Indeed, ABSD appears to have had a muted effect on reining in demand for mass market properties, and recent launches such as the Watertown and the Hillier selling extremely well, and approx 90% of sales going to Singaporeans.

As to be expected, none of the articles published and none of the experts quoted have made any detailed analysis of the state of credit expansion as a facilitator or driver of housing demand and housing prices. Economist Steve Keen and other analysts have argued (see references below) that the real explanation for rising house prices is rising credit. To be more precise, what drives the change in house prices is the acceleration of mortgage debt. The acceleration of debt has been dubbed "the credit impulse" by Michael Biggs et al, and more recently termed the "credit accelerator" by Keen.

In this post, I attempt an amateur empirical analysis of the interaction between Singapore credit activity and housing activity, inspired by the methods espoused by Keen. (disclaimer: I am not a professional economist so I make no claims to the expertness or accuracy of this analysis and welcome all constructive criticism).

The chart below shows a plot of housing credit (Source: MSB 1.5A Banks: Loans and Advances of DBUs to Non-Bank Customers by Industry, Consumer Loans — Housing and Bridging Loans via MAS) vs. the Condominium Price Index (Source: REALIS).

(Note on MAS statistics: Nov 98 numbers are an outlier due to DBS' acquisition of POSBank and the subsequent inclusion of POSBank as part of the banking system.)

This chart by itself doesn't necessarily tell us anything, instead what we are interested is the relationship between the acceleration of credit and the change in housing prices, which is plot on the chart below.

(Data points from Q4'98 and Q1'99 removed as outliers due to POSB & DBS merger)

Correlation analysis reveals a correlation of 0.123 with no timing delay between housing index and credit acceleration, and a correlation of 0.185 assuming the credit accelerator leads housing price index by 3 months. This is much smaller than Steve Keen's calculated correlation of 0.58 for Australia, and is possibly explained by the following factors:
  • Singapore's private housing sector has, until recently, had no restrictions on purchases by foreigners, whereas Australia has had stringent capital controls in place regarding foreign purchases of property in Singapore. This makes Singapore property prices much more exposed to non-local monetary conditions.
  • Singapore's economy as a whole is also much more significantly export-oriented and exposed to external economic shocks than is Australia, as such events like the Asian Financial Crisis and the Global Financial Crisis have resulted in several outliers to the data that may have distorted the correlation analysis.
Going forward, I suspect that Singapore's condo prices are going to be significantly more sensitive to local credit conditions than in the past, because the AB SD has effectively shut out much of foreign buying from the local market. I look forward to analyzing the relationship between credit impulse and housing prices in the future and expect the correlation to increase.

Nevertheless, the positive correlation shown in the data above does suggest that housing credit is a material contributor to housing price growth in Singapore, albeit there are also many other non-local-credit related factors affecting housing prices in Singapore.

Steve Keen's Debtwatch (May 2nd, 2011) "House Prices and the Credit Impulse"
Biggs, Michael; Thomas Mayer and Andreas Pick. 2010. "Credit and Economic Recovery: Demystifying Phoenix Miracles." SSRN eLibrary.
MAS: Monthly Statistical Bulletin

No comments: