The reports headline conclusion: a third of Van houses are falling into Chinese hands. That compares with the 5% estimate adopted by both the BC government and the provincial realtors association, plus a survey of members done by the local real estate board. In other words, what Routledge discovered would be a six-fold jump over the best estimate of everybody else.
So, how did he do it? Where did this data come from?
He made it up. Routledge admits it was a back of the envelope conclusion that he drew from 100% non-Canadian data that could have built-in biases that overstate (or understate) the amount of capital inflows. So, the methodology, whatever it was, probably sucked.
But, wait. We actually do know the methodology. Routledge took total sales figures for purchases by non-US residents from the American National Association of Realtors, which showed a big jump last year in Chinese action. He then found a completely unrelated survey done by the UK-based Financial Times which (in a multiple choice format) queried 77 high net worth Chinese guys who had purchased real estate abroad. Yes, 77 people. Admittedly not a statistically significant sample size, Routledge admits. No kidding.
Anyway, he took the results of the Times survey then applied the ratios of buyers in Vancouver (there were nine of them) and multiplied the US numbers by this factor to come up with a total probable dollar investment value for the YVR market assuming any of this was representative of the larger wholes.
That led him to this: One could hypothesize that for every four high net worth buyers from China who purchase a U.S. residence, one purchases a residence in Toronto; and, for every three high net worth investors from China who purchase a U.S. residence, one purchases a residence in Vancouver. Of course, given the fact Routledge had zero Canadian stats to go on, he could only deal in probable dollar values. That means, even if he were correct, there is no way of knowing how many homes in Vancouver went into Chinese nationals hands. It certainly was not 33%.