tag:blogger.com,1999:blog-31427364.post591287630909474075..comments2024-03-26T03:52:23.395-07:00Comments on Housing Analysis: How to do a quality-adjusted housing indexmohicanhttp://www.blogger.com/profile/06094213357140749289noreply@blogger.comBlogger5125tag:blogger.com,1999:blog-31427364.post-29908585151689082622009-02-03T20:25:00.000-08:002009-02-03T20:25:00.000-08:00Good work VHB.It seems normal to me to have some s...Good work VHB.<BR/><BR/>It seems normal to me to have some statistical fluctuations within the benchmark from month to month. Over longer periods of time these fluctuation work themselves out either by having larger or smaller price changes in previous or future months.mohicanhttps://www.blogger.com/profile/06094213357140749289noreply@blogger.comtag:blogger.com,1999:blog-31427364.post-14527548782738292252009-02-03T19:13:00.000-08:002009-02-03T19:13:00.000-08:00If you look at the regions with the most sales and...If you look at the regions with the most sales and ignore areas with few sales (hello Port Moody with 2 sales and detached +170k) you get a feeling that the numbers are mixed with no clear direction:<BR/><BR/>all detached <BR/>van west (46 sales): down 30k<BR/>van east (42 sales): up 13k<BR/>maple ridge (38 sales): up 14k<BR/>richmond (30 sales): down 37k<BR/>burnaby (29 sales): up 11k<BR/><BR/>So I don't see much significance in the benchmark "increase".Pandahttps://www.blogger.com/profile/07255920606087666835noreply@blogger.comtag:blogger.com,1999:blog-31427364.post-82101778550146109412009-02-03T16:45:00.000-08:002009-02-03T16:45:00.000-08:00There were still 700+ sales, more than enough to k...There were still 700+ sales, more than enough to keep the same confidence intervals. In fact REBGV gives a decent indication of this with the "price range" column in their data. For all of Vancouver this number hasn't moved all that much for January.<BR/><BR/>Likely there was either a true levelling off or, as VHB implies, a systematic skew in the data.jessehttps://www.blogger.com/profile/02155122147972263497noreply@blogger.comtag:blogger.com,1999:blog-31427364.post-17253430676621715442009-02-03T15:18:00.000-08:002009-02-03T15:18:00.000-08:00It also may be that the accuracy of any statistica...It also may be that the accuracy of any statistical calculation decreases with sample size. Since the sample size is number of sales, which is reducing, expect this number to be less than relevantRoad Ragerhttps://www.blogger.com/profile/07971511506963939609noreply@blogger.comtag:blogger.com,1999:blog-31427364.post-5955467134740730162009-02-03T14:25:00.000-08:002009-02-03T14:25:00.000-08:00Thanks for this VHB. A few notes about the HPI. Fi...Thanks for this VHB. A few notes about the HPI. <BR/><BR/>First, we have seen in San Diego that different tiers (low, mid, and high priced properties) declined at different rates on a PSF basis. Until now Vancouver has bucked that trend.<BR/><BR/>Second the HPI does not account for quality adjustments. For example if all properties sold had major high quality renos or construction, the price PSF would be higher. This is not captured in the HPI, based upon the definitions from the website. I agree with you what they are describing is hedonic regression but only apparently for more tangible variables like lot size, structure size, and number of rooms, not necessarily quality.<BR/><BR/>Note the Case Shiller index tries to eliminate major renos by excluding properties with obviously extraneous appreciations from previous sale. These are deemed to indicate a new structure or massive renovation. They can be excluded because the volume of "non-renovated" units can make up a more contiguous and statistically significant data set.<BR/><BR/>Still, some buyers are finding value at these prices. Not many of them. And that's the important takeaway.jessehttps://www.blogger.com/profile/02155122147972263497noreply@blogger.com