House Price Indexes at Statistics Canada Practices and

House Price Indexes at Statistics Canada: Practices and Research Serra Erdur & Marc Prud’homme Prices Division Presented for the Ottawa Group October 11 th, 2007

Outline l l l Alternative measures of shelter (part 2) Hedonics project (part 2) The “New” NHPI USER-COST ALTERNATIVE OOH RESULTS HEDONICS & RP NEW NHPI 2

Part 1 l Alternative Measures of Shelter m m User cost No shelter USER-COST ALTERNATIVE OOH RESULTS HEDONICS & RP NEW NHPI 3

User-Cost Formula r : mortgage interest rate M : mortgage debt i : 5 -year bond rate D : depreciation RC : recurring costs K : capital gains USER-COST ALTERNATIVE OOH RESULTS HEDONICS & RP NEW NHPI 4

Alternative Measures of Owned Accommodation Components USER-COST ALTERNATIVE OOH RESULTS HEDONICS & RP NEW NHPI 5

Canada Price Index Series for Owned Accommodation, 2000 Q 1=100 USER-COST ALTERNATIVE OOH RESULTS HEDONICS & RP NEW NHPI 6

Canada Price Index Series for All Items, 2000 Q 1=100 113 -116 107 -109 USER-COST ALTERNATIVE OOH RESULTS HEDONICS & RP NEW NHPI 7

Annual Growth Rates for Alternative Concepts, 2000 -2005 Money Outlays Concept Official Concept No Own. Acc User Cost Case 2 User Cost Case 3 Rental Equivalence Concept Owned Accommodation 2. 2% -2. 1% -1. 3% Shelter 2. 3% 1. 1% -0. 6% All items 1. 9% 1. 0% Commodity group USER-COST Net Purchase Concept Excluding equity payments Including equity payments Based on purchases Based on downpayment and mortgage payments 1. 4% 1. 8% 1. 0% 3. 3% 2. 3% -0. 1% 1. 8% 2. 0% 1. 6% 2. 9% 2. 3% 1. 1% 1. 8% 1. 7% 2. 1% 1. 9% ALTERNATIVE OOH RESULTS HEDONICS & RP NEW NHPI 8

Part 2 l Hedonics and repeat sales USER-COST ALTERNATIVE OOH RESULTS HEDONICS & RP NEW NHPI 9

Hedonic and RS Price Indices vs. NHPI for Ottawa, 1996 -2007/05 USER-COST ALTERNATIVE OOH RESULTS HEDONICS & RP NEW NHPI 10

Hedonic and RS Price Indices vs. NHPI for Ottawa, 1996 -2005: Annual Growth Rates WRS Nonfilter ed 3&30 yrs filter ed Characteristic filtered 5. 9% 5. 7% 5. 9% USER-COST Laspeyres Adj. Period Yearly Time Hedonic Monthly Dummy Hedonic Average Median NHPI 4. 9% 5. 2% 5. 5% 5. 8% 6. 2% 4. 5% ALTERNATIVE OOH RESULTS HEDONICS & RP NEW NHPI 11

Land / Structure Split: Land Price Index for Ottawa, 1996=100 USER-COST ALTERNATIVE OOH RESULTS HEDONICS & RP NEW NHPI 12

Part 3 l The “New” NHPI USER-COST ALTERNATIVE OOH RESULTS HEDONICS & RP NEW NHPI 13

The samples USER-COST ALTERNATIVE OOH RESULTS HEDONICS & RP NEW NHPI 14

Sample decay: Calgary USER-COST ALTERNATIVE OOH RESULTS HEDONICS & RP NEW NHPI 15

Sample decay: Toronto USER-COST ALTERNATIVE OOH RESULTS HEDONICS & RP NEW NHPI 16

New NHPI Results: Calgary USER-COST ALTERNATIVE OOH RESULTS HEDONICS & RP NEW NHPI 17

New NHPI Results: Calgary USER-COST ALTERNATIVE OOH RESULTS HEDONICS & RP NEW NHPI 18

New NHPI Results: Toronto USER-COST ALTERNATIVE OOH RESULTS HEDONICS & RP NEW NHPI 19

New NHPI Results: Toronto USER-COST ALTERNATIVE OOH RESULTS HEDONICS & RP NEW NHPI 20

The Workshop Marion Steel - Guelph University: “I would choose the hedonic replacement, partly because it represents the least disruption from the current procedure. Points to consider in the future include investigating the issue of whether house models should be weighted by number of units sold. ” l l Tsur Somerville - UBC Centre for Urban Economics and Real Estate: “Any variant of a hedonic index will be an important improvement over the existing method. ” USER-COST ALTERNATIVE OOH RESULTS HEDONICS & RP NEW NHPI 21

The Workshop l Paul Anglin - Guelph University “I think that there should be an emphasis on using a method that is easy to compute for two reasons: to make it easy to communicate to an outside audience and to produce robust estimates. ” l François Des. Rosiers - Université Laval: “Turning to hedonics for revisiting the NHPI methodology is the best decision Statistics Canada could take. Considering though that a drawback of using this approach is the relatively small sample size available - between 130 (Toronto) and 38 (Calgary) new house sales -, I would suggest to compute the index for a given month “t” by pooling the samples of the three preceding months while resorting to some kind of moving average. ” USER-COST ALTERNATIVE OOH RESULTS HEDONICS & RP NEW NHPI 22

THANK YOU USER-COST ALTERNATIVE OOH RESULTS HEDONICS & RP NEW NHPI 23