Table 1 Summary Statistics of Key Variables.

From: Investors matter when prices are dispersed: the effects of investor activities on housing price dispersion

 

Mean/Prop.

SD

Min.

Max.

Panel A. property-level transaction data (N = 376,548)

 Total Price (in million HKD)

3.50

2.57

0.25

28.50

 Unite Price (in 10,000 HKD)

6.84

3.80

0.78

30.49

 Saleable Area (Squared meters)

49.90

15.80

16.54

223.62

 Bay Window Area (Square meters)

0.24

0.82

0.00

10.13

 Building age

15.44

9.40

−1.67

49.92

 Floor

18.14

12.45

0.00

81.00

 Duplex

0.00

0.04

0.00

1.00

 Roof

0.01

0.10

0.00

1.00

 Balcony

0.08

0.27

0.00

1.00

 Parking

0.01

0.11

0.00

1.00

 Garden

0.00

0.03

0.00

1.00

 Bay Window Included

0.05

0.22

0.00

1.00

 labeled as investment property

0.48

0.50

0.00

1.00

 labeled as flip (subgroup of investment property: holding tenure <2 years)

0.05

0.23

0.00

1.00

Panel B. housing estate-level panel data (N = 13,297)

 Price dispersion (standard deviation method)

0.06

0.04

0.00

0.79

 Investor entry turnover (%)

0.47

0.63

0.00

10.54

 Flipper entry turnover (%)

0.24

0.38

0.00

4.73

 Investors’ investment properties turnover (%)

0.45

0.43

0.00

5.05

 Investors’ flipped properties turnover (%)

0.05

0.11

0.00

1.76

 Average neighborhood investment activities (within 1 mile radius) (%)

49.41

13.51

0.00

100.00

 Average neighborhood investment activities (within 2 mile radius) (%)

49.45

10.03

0.00

100.00

 Trading volume

24.22

40.48

2

669

 Liquidity

0.01

0.01

0.00

0.12

 Housing estate-specific house price index

0.00

0.77

−1.47

2.32

 Private residential market yields

3.84

1.05

2.15

6.40

 Supply (Completions divided by stock)

0.02

0.01

0.00

0.06

 Estate size (Number of units in the estate)

2,806

2,335

892

20,503

  1. aPanel A presents property-level transaction data used in Eq. (1) to derive price residuals from the combined model. It includes 376,548 observations, the sample size for Eq. (1), from which housing price dispersion at the estate level is calculated using the standard deviation of price residuals. Panel B represents the estate-year-quarter panel, tracking variable changes across different estates over time. It contains 13,297 observations, corresponding to the regression sample in Eq. (3). The summary statistics for Panel B are based on this sample.
  2. bBuilding age is calculated as the transaction year minus the completion year. Negative values indicate presales.
  3. cThe housing estate-specific house price index is extracted from the coefficients on interaction fixed effects between housing estate and year-quarter, and the values are normalized to 1.
  4. dHousing supply = Completions/stock of residential properties; This variable varies with different classes of Saleable Floor Area (SFA). Data Source: Hong Kong SAR Government Rating and Valuation Department. https://www.rvd.gov.hk/mobile/en/property_market_statistics/completions.html;
  5. eData Source of property market yields in the private residential market: Hong Kong SAR Government Rating and Valuation Department. https://www.rvd.gov.hk/mobile/en/property_market_statistics/index.html.
  6. fHousing price dispersion (the standard error of price residuals in Eq. (1)) is only computed when the trading volume exceeds two, hence the minimum trading volume is greater than zero.
Back to article page