The Joys of Economic Modeling

A physicist, an engineer and an economist go for a job interview. The interviewer, not in a mood to use lots of his gray cells, asks a simple question to each of the candidate. He asks, “What is the sum of 2 and 2?”
The physicist snaps immediately: exactly 4.
The engineer thinks for a while and says, 4 with may be 2 per cent margin of error.
Then comes the economist's turn. When asked the same question, he responds, “What do you want it to be?”

So goes the old joke about the nature of dismal science. Forecasting of any type is inherently risky, but when it is applied in the realm of social sciences it can be self serving.

Of course, this perambulatory text was required before trying to question some of the “research” published by the UBC school of Business.
There are generally very few problems with the models used in forecasting or in pricing based models. The models are generally borrowed from the domain of physical sciences and the underlying "math" of these models is solid. The problems are usually with the assumptions behind these models.
No wonder, it was the contribution of similar simplistic and infallible modeling that has left the US financial system almost bankrupt. Stupid is as stupid does. Especially when making important assumptions.
Most financial institutions priced their mortgage securities based on the assumption that real estate always goes up. And they gently added an appreciation factor of 4 or 5 per cent in all their calculations. Of course, prices go up, until they don’t.
And that’s precisely the mechanism you use to come out with ridiculous statements such as Vancouver being over priced by a mere 11 per cent and Edmonton actually being under priced by 8 per cent as mentioned in this study.
Vancouver, is the mother of all bubbles and we’ll see how far it’s going to fall when everything is said and done.
But this whole paper is just a glorified buy versus rent calculator designed to appeal to authority. There’s nothing interesting there- just a simple formula with unreliable data, with no references to data collection techniques or sample space. We don't know for example, how many data points were analyzed when collecting rent information. How many ads were actually verified for accuracy on craigslist or Kijiji? Did they actually call the advertisers? Did they negotiate the rates? One would expect some rigour or explanation on data collection when an academic paper is published.
But this study isn't meant for publication in a peer reviewed journal. It's designed to get the greatest of all fools who have somehow managed to so far resist the temptation to buy.
Rather than wasting 20 minutes or so on this paper, prospective buyers will be better served if they use this intuitive and simple buy versus rent calculator from NYT.

The Real Estate complex in this country knows that things are falling apart pretty much all across the Canada and they are trying their level best to contain the impending crisis. This "academic work" is too feeble an attempt and unlikely to convince anyone but severely delusional.