Home sales slowed by over 40 per cent in July but does that mean prices are headed for a massive correction? Robert Both, Macro Strategist, TD Securities, shares his housing valuation model with Sara D’Elia and says while home prices are overvalued, the gap is probably smaller than you thought.
In Toronto, July home sales sunk by over 40%, but my guest says not to confuse home sales with home prices. I want to start with your chart, Robert. Your model looks at home price gap since the 1980s. The middle line is what you call a fairly valued market. The blue line is home values in Toronto relative to a quote unquote "fair price." But to be clear, this isn't dollar number or an average price. And the gray line represents home prices in Canada relative to the fair price nationally. The heart of the data here, Robert, is really fair price, so how do you go about calculating that number?
So to come up with an estimate for fair price, we built a model using a set of seven economic indicators, most notably household income, borrowing rates, and household wealth, as well as labor force participation. And to isolate the impacts or the drivers of house price appreciation from regular inflation, we based our analysis off of a measure of real house prices, which is the average selling price deflated by CPI. Now, you can use this to compare it to actual home prices. And whenever there is a gap, which we call a valuation gap, it would signal an overvalued or an undervalued market.
An important takeaway here is that it looks like from the last almost 20 years, we've been overvalued for the most part. So is this something new?
Well, most traditional metrics of housing market valuation, such as price to income or price to rent ratios have also suggested an overvalued market for a number of years. Persistent valuation gaps are a little more rare in our model. However, there are some important points that I'd like to highlight. So in mid-2016, we had the largest overvaluation in the national housing market in the history of our model, and that occurred at a point where the Vancouver housing market was approaching its top, while at the same time, the real economy was suffering from the fallout of the Fort McMurray wildfires. Meanwhile in 1994, we had the largest undervaluation in our model's history, and that occurred in Toronto, where the local housing market was still suffering from the collapse five years earlier at a time when the economy was growing at nearly 5% year over year.
I think it's shocking when we look at where your model is today because it doesn't look like the gap is that far off this middle line, but we've been hearing reports that Toronto housing can drop as much as 20% to 30%. So what is your model telling you?
So because our model is based off economic fundamentals, we don't see that large of an overvaluation because the economy is actually performing quite well. The model is most prominently driven by household incomes, and the outlook doesn't really support any large correction in equilibrium prices. Now, the most likely thing that would drive this would be an income shock, which is another way to say a recession. However, our base cases just for prices to rent remain more or less stable over the coming quarters.
Now, and that number being around 4%, which is quite different, because if you own a million dollar home in Toronto, which is quite common, that's a $40,000 decline versus where the estimates set $200,000 to $300,000. So are things not as bad as we thought?
That's exactly right. $40,000 might still sound like a lot of money to many Canadians, but for the average Toronto homeowner, they've seen their house rise in value by far more over the previous 12 months. So overall, it's not as big of a overvaluation as some numbers might suggest.
Thank you very much.
Thank you for having me.