Tuesday, February 28, 2012

Vancouver Housing Bubble: Part 2

Two weeks ago, we looked at the possibility of a Canadian housing bubble, and saw that Vancouver looked particularly bubble-icious in terms of new inventory. This is often a good indicator, as builders in a free, capitalist society should only overbuild if they are getting a strong price signal to do so. But of course, it is possible that builders are just irrationally overbuilding for some reason (e.g. government stimulus, tax credits, expectations of price increases etc.), so it does make sense to look at more metrics than just new home inventory before concluding that there is a price bubble.

The ratio of house prices to income levels is another useful metric for judging how expensive houses are relative to a historical standard. Canada-wide, this ratio is elevated, but perhaps not dramatically so. Recall from last week, however, that there is considerable variability with respect to housing data within the country. It, therefore, makes sense to look at this ratio for specific regions in order to identify trouble spots.

Considering Vancouver has the highest house prices in the country, it makes sense to start there. Consider the house price to income ratio* in Vancouver over the last 35 years:

That is a scary chart. For this ratio to be back in line with its levels of just 10 years ago, prices will have to fall by 50% or incomes rise by 100%...which of those is most likely? The former. Though this ratio can continue to rise for a while, with cheap money funding further asset purchases, it can't go on forever. At some point, it seems very likely that some heavy pain will be felt in this region.

* Income used is "average family income" as per the Statistics Canada CANSIM database. House prices are "detached residential average sales prices" as per REBGV


another value investor said...

I am bearish on Vancouver RE as well and am waiting for a better time to get into the market. I can see two decent arguments against using price/income (P/I) ratio as a valid indicator:

1. Purchases by foreigners. They tend to have low (local) income and thus distort the P/I.

2. Densification in the city. More people are living in smaller, multi-family dwellings now compared to the previous generation. Of course these places tend to be cheaper on an absolute basis (but definitely not $/SF!).

Do you have any comments regarding these two points?

Saj Karsan said...


I agree P/I may not tell the whole story, which is why I would recommend looking at some other metrics as well (e.g. price to rent, new home inventory etc.).

Regarding your points:

1) Even if this is distorted by a few buyers, it should hardly change the region's average income so significantly as to totally skew the graph

2) You can test for this directly, by comparing condo prices (instead of house prices) to rents or incomes to see if something has changed dramatically in the last few years