May 20, 2022 | 12:55
Bringing Down the House—Corrections Past and Present
Bringing Down the House—Corrections Past and Present
Canadian home prices surged more than 50% from the early pandemic lows, but now they’re getting seriously tested. We’ve long maintained that demographic and supply-side fundamentals have driven price gains, even in the early stages of COVID-19 alongside some economic adjustments. But, as we warned early last year, more recent price behaviour has been driven by excess demand, market psychology and froth. Now, the Bank of Canada is on the scene and the market is softening sharply in some areas. So, when we speak of a housing correction it’s not a question of if, but where, how much and for how long?
Where Is the Risk?
Real home prices in Canada have historically grown at about 3% per year dating back to the early 1980s (Chart 1), roughly reflecting inflation, real wage growth and gradually falling interest rates. In the current episode, even as inflation has accelerated to multi-decade highs, real home prices have surged by more than a third in the span of two years, clearly stretching beyond that long-run baseline growth trend. Historically, we have seen some dramatic deviations from trend, all of which were eventually corrected in some fashion. In one extreme example, real home prices surged 38% above trend in the late-1980s, and took 15 years to recover. By the early-2000s, however, housing was exceptionally cheap, which set the stage for the long bull market that followed. Notably, for most of the 2010s, housing was doing roughly what it should have been doing (recall the false bubble calls we were constantly arguing against), until BoC rate cuts in 2015 led to some froth in B.C. and Ontario into 2017—that quickly corrected with rate hikes. As of Q1, when we believe the market peaked, real Canadian home prices sat 38% above trend, the widest deviation in the past 40 years.
By this measure, the most froth has accumulated in the suburbs and exurbs of Toronto. While Toronto prices have risen 41% above trend, exurbs (using markets 1-2 hours outside Toronto) have run ahead by more than 70%. Cottage country has seen a similar surge. Some regions, however, barely look frothy at all. Alberta is a good example where, after five years of declining prices leading into COVID, the market is still just catching up to its long-term baseline. Other markets, such as Vancouver, Montreal and Atlantic Canada are frothy, but not to the extent of Southern Ontario.
How Much Risk?
Interest rates are almost always at the root of housing market excess and corrections, and this episode is no different. After leaving policy too loose for too long, psychology and affordability have already been tested by just 75 bps of Bank of Canada tightening, and we expect another 125 bps by year-end. That effectively means that the market will go from being priced at mortgage rates of roughly 1.5%, to somewhere in the 3.75%-to-4.5% range, depending how bond yields evolve (Chart 2). The difficulty is that home prices were already stretched versus income and interest rate levels at those ultra-low rates, so the upward move will be that much tougher to swallow (Chart 3). Indeed, if we assume steady income growth and stable home prices, a move to 4% mortgage rates will stretch our valuation metric beyond that seen in the late-1980s. The calculus suggests that a 10-20% decline in prices would be needed, all else equal, just to maintain current (stretched) affordability. Mileage will vary by market and region, as noted above.
We can also look at this from the perspective of an investor. In Toronto, for example, cap rates on multifamily properties fell to just above 3% by the turn of the year, or roughly 150 bps above 10-year GoC yields. Historically, that spread has averaged about 250 bps above 10-year GoCs over the past two decades and, with the move in yields to 3% in recent weeks, a repricing is underway. To restore a more normal relationship, cap rates would probably have to rise to the 4%-to-5% range, allowing some leeway for scarcity and higher rent inflation. Depending on inflation and rent assumptions, that could carve around 20% off the underlying asset price.
The length of this correction depends crucially on how the economy handles tighter monetary policy and when inflation breaks. Specifically, can central banks engineer a soft-ish landing that allows the cycle to run on? If not, what is now an asset-price correction could evolve into more prolonged economy-driven weakness, but that is not our base-case view. That said, history suggests that localized price corrections in Canada usually take 2-to-3 years to bottom, and 4-to-5 years to fully recover (Chart 4). Interestingly, with the exception of oil-driven markets in Alberta, interest rates were the trigger for all major corrections since the 1980s.
What Will Stop It?
We believe underlying fundamentals remain in place to eventually support the market. Demographics are a legitimate source of demand, with elevated international immigration targets and the millennial cohort (peak age is now about 32) supporting both the rental and home buying market. There will be buyers waiting; but prices need to make sense in a higher-rate world.
Building costs are also a major factor, and the escalating cost of completing a new housing project should, in theory, put a floor under resale prices at some point, perhaps sooner than later. While resale prices have run ahead of residential building costs over the past two years, the latter were up 23% y/y in Q1 (Chart 5), according to StatCan, with little sign of slowing amid high input costs and labour shortages.
Finally, unlike the U.S. situation in 2008/09, the structure of Canada’s mortgage market should contain the spillover. Full-recourse mortgages (only Alberta has non-recourse, with some caveats) have historically limited defaults during even the deepest price corrections. And, stress testing has insulated the financial system from higher interest rates, at least from a capacity-to-pay perspective. For example, even while borrowers were taking on mortgages at 1.5% through the pandemic, they were stress tested in the 4.75%-to-5.25% range (Chart 6). This won’t save house prices from falling (they are driven more by the rate you actually pay than a backroom calculation), but it provides good insurance that payments will keep being made, especially with the labour market so tight.
Bottom Line: Froth is coming out of home prices, just as it is across a number of other asset classes that were boosted by excessively stimulative policy. While a much cooler housing market will weigh on economic growth, we believe that this will be largely an asset-price phenomenon, with underlying fundamentals eventually setting a floor, and the financial system well protected.