Test Scientifically the Determinants of Housing Price

Transcription

1 EYTalk Dec 12, 2016 Test Scientifically the Determinants of Housing Price Edward Yiu Legislative Council Member 1

2 Two Major Hypotheses Insuff Supply Housing Supply Real Interest Rate Ultra Low IR Land Supply High Inflation

3 No Evidence on the Insuff Supply Hypothesis There is no relationship between the proportion of private housing completions and property prices in terms of their correlation coefficient. (Bauhinia Foundation Research Centre, 2014) Bauhinia Foundation Research Centre (2014) Relationship between Proportion of Private Housing Completions, Amount of Private Housing Completions, Property Prices in Hong Kong,

4 50% Housing Supply v HPIyoy High High % yoy HPI % Completions (RVD) Low High % % % % -20% % -40% High Low Corr =

5 Land Supply and Vacancy High Low Low Low Medium High Wang, H., Zhang, C. and Dai, W. (2013) Rental Adjustment and Housing Prices: Evidence from Hong Kong s Residential Property Market, HKIMR Working Paper No. 01/

6 Asian Pacific Study Refutes Insuff Supply Hypothesis the coefficient of land supply index is [positive], which contradicts with the theoretical prediction that increases in land supply have a dampening effect on house prices in the long run. On a second thought, this may reflect an impact in the reverse directly, i.e. higher house prices provide an incentive for developers to build up new residential property projects (Glindro et al., 2007) Glindro, E.T., Subhanij T., Szeto, J. and Zhu, H. (2007) Are Asia-Pacific Housing Prices Too High For Comfort?, BIS paper

7 Science Requires Empirical Evidence to be science it must be testable and falsifiable. How do we test this theory of supply and demand curves? Demand curves can t be found. Neither can supply curves. They are purely hypothetical constructs with no way of being measured. It is not falsifiable. This is fatal for any theory if it is to be considered science. (Aguayo, 2014) Aguayo, R. 2004, The Metaknowledge Advantage: The Key to Success in the New Economy, Ch. 19 New Economic Paradigms, p.210 7

8 Wheaton Theory Contradicts Insuff Asset Market: Valuation Supply Hypothesis P=R/i Rent $ Property Market: Rent Determination Price $ Stock (sq. ft) Price affects Supply P = f(c) Asset Market Construction Construction (sq. ft) S=C- S Property Market: Stock Adjustment DiPasquale-Wheaton s (1992) Model

9 RIR Hypothesis Undeniably, real interest rates are both theoretically and practically important determinants of aggregate demand.... the direction of asset price movements will depend on whether the long-term real interest rate is expected to rise or fall (HKMA Bulletin 8/1999)

10 RIR Hypothesis Casual Evidence Snapshots confirm the hypothesis HIBOR % CPIA yoy % RIR HPI (yr end)

11 Asian Financial Crisis 1997 HK

12 Corr = -58%; RIR v HPIyoy

13 Econometric Model RIR effect Positive and Negative RIR Effects separated

14 Negative RIR causes HPI hikes -0.83% -2.61% Yiu, C.Y. (2009) Negative real interest rate and housing bubble implosion an empirical study in Hong Kong, Journal of Financial Management of Property and Construction, 14(3),

15 Global Financial Crisis 2008 HK

16 Predictive Power Yiu, C.Y. (2009/07/15) Negative Real Interest Rate and Housing Bubble Implosion - An Empirical Study in Hong Kong, SSRN 2009 CCL = 75; 2016 CCL = 144; ~100% up

17 Is RIR Theory Still Correct? Dip into Negative Zone since Nov 2008; Just Surface Up to Positive Zone in Oct 2016

18 Evidence in Other Economies Japan s Lost Decade 1980s HK Asian Financial Crisis 1997 US s Subprime Crisis 2007 Spain s Sovereign Debt Crisis 2010s

19 RIR v US Subprime Crisis 2006

20 RIR v Spain Housing Bubble -9% /cr09129.pdf

21 RIR v Spain Housing Price

22 Econometric Models

23 Empirical Results Variable Model 1 Model 2 Model 3 Dependent Variable yoy_hpi yoy_hpi dlog(hpi) constant *** * yoy_gdp 2.092*** 1.815*** --- yoy_une 0.061*** RIR *** RIR_Before *** *** RIR_After *** 0.004** dlog(gdp) *** dlog(une) **

24 .4 RIR v Singapore Housing Price PI_SIN_URA_YOY RIR_SIN_3M 24

25 Empirical Evidence Variable Model 1 Dependent Variable yoy_hpi constant yoy_gdp 0.375** yoy_une RIR **

26 RIR Theory P = = a a( 1+ g) ( 1+ r) ( 1+ r) T t -1 a( 1+ g) t ( ) t= 1 1+ r + ( 1+ g) ( ) 0 3 K 1+ r a = if T and r > g r - g a a ( 1+ g) ( ) 1+ r T -1 T Yiu, C.Y. and Hui, E.C.M. (2005). Capitalization Rate and Expected Return: A Cointegration Analysis in a Housing Market, Journal of Real Estate Education and Practice, 9(1),

27 RIR +ve Again 6.00 Real Interest Rate

28 HPIyoy v RIR

29 HPIyoy(+5m) v RIR Corr = -53%; y = x R² =

30 The End Enquiry: 30