As a follow-up to Mohican's post from March 10, here are some files I've been working on recently.
I've tracked 3 sample properties-- a Vancouver westside bungalow (what I'd like to move up to), a Vancouver westside townhouse (what I currently own, bought during the lull in 2004), and a Richmond bungalow (my father-in-law lives in one). My interest in developing these charts was to discover how prices rise over time-- it seemed expensive when I bought my place, but not outrageous. Prices have, of course, risen since then.

First chart: a westside bungalow over 30 years. The mean curve rises at roughly 1.5x the rate of inflation. Note the thin dashed line-- that shows the property price as though it was rising at the mortgage rate. Prices have, over 30 years, never sustainably rised at the mortgage rate. Note that this house left the mean in 2003.

Second chart: a westside townhouse over 20-some years. The mean curve also rises at roughly 1.5x the rate of inflation. This property left the mean shortly after the Vancouver bungalow.

Third chart: a Richmond bungalow over 30 years. Interestingly, prices only rise at the rate of inflation here. Perhaps the lower "mean" line is due to the "desirability" of the westside? This property didn't leave the mean until 2005!

Fourth Chart: all three properties, all together. 5-year mortgage rates and inflation rates are also overlaid onto the chart. (I couldn't figure out how to get Excel to have a second scale for the percentages, so I had to wildly scale them up).
Observations:
-Westside properties tracked up at 1.5x inflation.
-Richmond property tracked up at 1.0x inflation.
-Long-term, prices tend to rise at *less* than mortgage rates, making housing a poor speculative investment.
-During booms, housing is probably a great speculative investment!
-Neither the Richmond bungalow nor the westside townhouse benefitted much from the Hong Kong migration-- perhaps those investors were looking for land with a better investment value?
With all three sample properties, it'll take 15 years for any of the means to catch up to current property values! (unless selling prices drop or inflation goes up)
I'm willing to bet that the three sample properties have a 20-33% (nominal) drop in store over the next five years. However, until housing starts fall off a cliff (which will happen maybe a year after prices go flat-- perhaps early 2008?), and until much of the Olympic contruction is finished (~2009), I don't think we'll see significant drops.
Right now, our local economy is artificially-supported by government spending projects (Olympics) and speculator-sponsored condo development. Once the Olympic construction is done, the government will have finished building an amazing amount of infrastructure, and they (the government and us taxpayers) won't be keen on more deficit-spending to build more infrastructure-- hwy 99, Olympic venues, Athlete's Village, the new Convention Centre, RAV, Green Line (Coquitlam light rail, I think), Port Mann's twin, the Golden Ears bridge (I can't remember its name-- the one that goes from Maple Ridge to Port Kells)...
All those projects ending at the same time. How many construction-related workers moved to Vancouver for those projects? How many of them bought houses or condos? How many bought a handful of presale condos because prices always go up? How many of them will sell their places once the government freezes new infrastructure development after the Olympics?
Note: these charts use Royal LePage pricing data up to Sept 2006, and inflation/mortgage rate data from Sauder. Where data was not available, it was interpolated or omitted.