Blessed by the Potato

This spring looks like real estate prices are coming off the boil. People seem to be wondering if prices could possibly go down despite years of bulls and the FOMO crowd saying that it only ever goes up.

There have also been more people wondering about housing bears. “Remember housing bears? Didn’t they once say that housing could possibly go down in the future? Is this what ‘down’ is?” And of course the big one “So how bad could this get?” As us long-time housing bears come out of hibernation this spring, perhaps it’s worth a long (and snarky) review of the thesis.

A Low Growl

There are lots of ways to determine what a house is “worth” or what it “should” be priced at. You can look at historical relationships between incomes and debt and house prices, but that’s all very high-level stuff, and hard to apply to one specific house that you want to live in. The argument that has always resonated best for me is that a house is primarily a place to live, and shelter is also one of life’s greatest expenses, so it’s worth looking at all your options. Renting your shelter or buying it are the alternatives available.

So we can examine the relative costs between renting and buying, and I made a spreadsheet and a whole series of posts to do that and others made rules-of-thumb and YouTube videos. And basically house prices had gotten so expensive relative to rents that it looked like you were better off renting, and the only way buying would possibly come out ahead is if already-expensive housing became even more expensive at a very high clip and even then you needed cheap leverage.

Well, we got skyrocketing prices and cheap leverage, so people who bought anyway look like they came out ahead (for now). But it didn’t have to turn out that way, and the question is always how much further can those trends go before something breaks? The bigger the price:rent ratio gets, the harder it is to fix it through anything other than a housing crash (assuming of course, that these fundamentals matter at all).

But if doing a price-to-rent comparison is so easy anyone with a physics degree and insomnia can do it, why did prices get so detached from fundamentals in the first place?

Models of the World

The market is far from monolithic: there are many players in it. Not just buyers, sellers, and agents, but buyers with different pricing models at work – people are playing the game with different rules. We can build a thought experiment with different agents (agents in the sense of actors in a model, not the people whose job it is to sling houses) using different strategies and approaches, and how prices might change.

A decade ago we talked about supply and demand, and how speculative demand works: in a textbook economics market, as the price increases, fewer buyers demand the thing the textbook is discussing. Even if they disagree on the assumptions and arrive at different estimates of fair value, people who use a rent-vs-buy framework will, all else being equal, start to drop out of the market as the price gets too high, and choose to rent instead. But with speculative demand, you don’t get that: high prices beget high price expectations, which helps bring in new buyers to replace (or even exceed) those who got priced out.

So that’s one of our groups of buyers and their pricing models: prices going up is good, because they’ll continue to go up, so pay whatever you need to today as you’ll only be richer in the future. This group is what we’ll call the FOMO-MOMO group: fear of missing out/momentum. They expect that as prices go up, they will continue to go up. And the faster they go up and the closer housing gets to completely unaffordable, the more they will pull demand forward and buy more because they are more and more afraid of missing out or getting priced out (e.g., those stories of people buying places for their children). Their pricing model is basically to look at the last comparable price, see how many people show up for the bidding war, and add $10k for every nose on top of the last price.

It can have big positive-feedback cycles, as prices going up makes them feel wealthier, and feel more certain in their worldview, and gives them easier access to credit. A subset of them buy more and more properties: as the first purchase works out and goes up, they build equity, which they can drain with a refinance to buy another place, and another, helping to propel the market upward while adding leverage and fragility to the structure.

The bear case is of course that this is insanity, but we’ll get to that.

The next class of buyers are the ones who just want a place to live, but who skip step zero (i.e., the rent-vs-buy analysis). So they are only looking to buy, and they’re just simply going to go into the market at some pre-determined time point (say on their 30th birthdays for the sake of modelling) and buy as much as they possibly can (AMAP). They don’t care too much about what the price of housing is doing – up, down, or sideways – they just borrow as much as they are able and buy as much house as possible. Their only limit is their ability to service the debt used to finance the purchase and things like a stress test that affect their buying power.

Then you have the “rational” buyers (R): those who may want to buy a house, but want to do so at a “fair” price, and will go rent or GTFO if the prices get too extreme. They may use a price:rent or price:income basis, like dinosaurs, but they’re looking to buy without becoming house poor.

So if we have a group of price-insensitive buyers, and even a small supply/demand imbalance, the FOMO-MOMO and AMAP crowd can rapidly drive prices to crazy levels, and drive the R’s completely out of the market – the demand is strongly inelastic from those buyers. In addition to all the speculation about how this was working in Toronto and Vancouver’s run-ups, we saw more natural experiments with this in the pandemic: prices in many small exurb towns doubled with very small increases in the number of price-insensitive buyers hitting the region.

These pricing models used by the currently dominant agents lead to a powerful positive-feedback loop. But what happens when the market finally changes?

Eschatology

As soon as prices stop going up, the momentum buyers may hit pause, creating a reinforcing cycle in the other direction: fewer buyers, putting fear into others, as fewer and fewer believe it’s a dip and start to fear prices won’t be up so much in five years’ time that any price today can be justified. Plus, how do you even set a price if your model is to multiply X by the number of noses in the bidding war and nobody else shows up to bid?

If the FOMO-MOMO crowd disappears, the next marginal buyer is the AMAP, who simply will borrow as much as they can for a house and pay that. With interest rates rising, their ability to borrow is more limited, so prices will have to come down a bit to match their maximum affordability.

So our mental model suggests that the first new equilibrium point would be wherever the new constraint on buying power puts prices.

Take a house that’s currently selling for say $1.5M in Toronto. An AMAP buyer that could afford that place at 2.5% (a mortgage payment of ~$5400/mo) may only be able to pay $1.3M at 4%, or $1.2M at 5%. Not a huge difference, which is what a lot of people are saying now – a slowdown or correction is in the works from interest rate increases, but far from the crash the bears have been crying about that would be needed to restore actual affordability. We’re still in the post-pandemic bump there, and not even negative year-over-year thanks to that massive run-up at the end of 2021/start of 2022.

The true bear case is if the fomo momo crowd steps back, and there aren’t enough price-insensitive AMAP buyers (or their purchasing power is significantly impaired, which may happen for example if they need a down payment from the bank of mom and dad, which is also facing higher financing costs on their HELOC, or interest rates rise even more, or inflation hits all the other line items in the GDS calculation at the same time that interest rates rise) to move the needle. The next marginal buyer is then the “rational” buyer.

And the problem is that the price that the “rational” buyer or investor buyer steps in is far, far below the current prices. That $1.5M (peak) house might rent for $3200/mo, a price:rent of 469X. To get back to say 250X, the price has to come down to $800k – a big air gap below current prices.

So that’s how a crash could play out from too-high prices and a shock that breaks the momentum and psychology. It’s certainly plausible, but is it at all likely?

And that’s the problem with this mental model: it produces some narratives to help explain how we got here, and some potential future paths, and where some reasonable stopping points for bottoms might be (when the monthly payment for AMAPs balances out, when the price:rent gets back to historic norms, etc.). But it doesn’t make any actual predictions about what will happen – there isn’t survey data out there about how many AMAPs are left, for example.

The Turn

Though the market does seems to be turning now, with rising interest rates putting fear into the hearts of the fomo momo crowd, and reducing how much “as much as possible” is for the AMAP crowd – the “every 50 bp reduces the maximum amount you can borrow by X,” etc. articles that you’ve no doubt seen all over.

But wait, there’s more! In the calculation for the debt service ratios is also a line for heating costs. And natural gas has roughly tripled in price over the last few months, which will take another decent chunk ($100/mo or so?) out of the maximum borrowing capacity of buyers.

So in that example of a $1.5M house that an AMAP could reach for at 2.5% going to $1.2M at 5%, add in an extra $100/mo hit on heating costs and now their maximum serviceable price is hit by another $20k.

I see lots of people still saying that there’s no way it could crash though, and there’s a lot of good reason to think that way. The government explicitly said they wouldn’t allow even a 10% correction to happen, and they’ve made plenty of (deliberate?) policy errors to get us into this over-priced mess in the first place. And the market has gone straight up in the face of bear logic for so long that people believe that there must have been a mistake in the logic somewhere. Every time we’ve had a good reason to think a correction was at hand (2012, 2017, 2020), the market has given bears a thrashing and gone right back to resuming its lunar trajectory.

And maybe they’re right. The fundamentals have gotten worse (better if you want high prices): we do also have a supply problem now on top of the demand surge, and we can see it in rents that have gone up above the rate of inflation (esp. for product that makes for a good [illegal] AirBnB). However, the increase in rents isn’t even in the same solar system as the increase in prices. So the floor has been raised… but it’s still a long way down.

And the demand is still high: for all the demand that has been pulled forward, there are plenty of Canadians and newcomers who want to own, they’re simply priced out. Maybe there are hordes of would-be buyers on the sidelines who will come rushing in to buy on any weakness, and this will only be a gully, not a crash. But at what price can those priced-out hordes come in? If they are themselves AMAPs who were priced out, then prices need to drop enough that even with higher rates and heating costs that they can service the debt again (I am very carefully avoiding the word “afford”), so we’d need to see prices fall by at least 25-30%.

Six Impossible Things Before Breakfast

Through the years, many people have acknowledged how expensive Canada’s housing markets have become, but say that a crash is impossible for reasons.

Back in the day when Alberta was even bubblier than Toronto, people said Toronto could crash, but no way Calgary would because its house prices were supported by the oil industry, and demand for oil wasn’t going anywhere. Just a few years later, the price of oil crashed, though miraculously Alberta experienced a soft landing rather than a crash. Record-low interest rates (which drove Vancouver and Toronto to new heights as they stabilized Alberta) likely helped a lot with that feat.

In Toronto, interest rates have been one of the key impossible things supporting the market: people can afford the payments to keep the market elevated as long as interest rates stay low. And far past the crisis days of 2008, emergency rates stayed with us, and even went lower and lower. It didn’t seem like the BoC was ever going to raise rates, and there are only so many years financial pundits can warn people that rates will eventually go back up before buyers tune the warning out and just go hog wild on debt. Even as inflation started to rage at the end of 2021 and beginning of 2022, the BoC stood its cowardly ground, and bulls became even more emboldened in their call that rates would never go up.

And besides, the argument went, “they” couldn’t raise rates: we’re too indebted as a society, higher rates would crash prices and cause a recession, and they would never do anything to endanger house prices.

Well, it looks like we’re in for some rising rates. [Ed. note: since the first draft of this post and the enormous amount of time it took me to hit publish, the BoC has raised rates 100 bp in two 50-bp steps after the first cowardly 25 bp hike in March] And if the argument was that rates couldn’t rise because house prices would fall… I guess house prices are going to fall then?

When markets get badly distorted, but the most obvious corrective force is “impossible”, well, impossible things seem to happen all the time to break theses in nasty ways. Many people seem to be learning the hard way that real estate prices can go down, and rates can go up.

The Big Misunderstanding

One other common sentiment out there is that “bears were wrong and so will always be wrong. Even with rising rates and a correction, house prices will never go back to 2011 levels so bears have lost! Don’t listen to bears!” And this sentiment is either an attempt to show a brave face as a bull, or a deep misunderstanding of the rent-vs-buy math that we’ve been talking about for that decade.

If you were in Toronto in 2011 and decided to rent when the price:rent was 300X, the bulls were (so far), right post hoc and had a faster increase in net worth… if they were sufficiently levered (and sell and go rent now to lock in that gain). But check that spreadsheet again: the base case for renting was that you would be ~$100k ahead (back then six figures was a lot of money and not just a kicker to win a bidding war) assuming historical ~3% increases in real estate. Bears don’t need prices to go back down to nominally what they were (though we do need a correction to pull ahead < a href="http://www.holypotato.net/?p=2516">post-hoc), as the difference in monthly cashflows and investments can close the gap. Indeed, at various points along the way the price:rent got so disjointed that you needed a 6%/yr increase in house prices forever just to keep pace with the renter. The last decade saw closer to 7%/yr in Toronto, so those who bought have done well for themselves… but how much longer can that rate of appreciation keep up?

Someone who passed on a $700k house in Toronto to rent and invest the difference does not need house prices to crash below $700k to come out ahead – they just need prices to come back to ~$1.2M by 2024 to win in the post hoc comparison. Yes, that will still require a correction from here, but the delta is not nearly as big as those who just want to say being early is the same as being wrong want to paint it as. If you did make the “disastrous” choice to rent, you’re not nearly as far behind as all that.

How Far Down?

A common question I see directed at bears now that people remember that we exist is how far would it have to drop before you’d buy? I’ve been quite happy renting here, and the LL doesn’t seem to be in any hurry to throw us out. Even though it would only take a ~30% correction from the peak to end up ahead (and prices are already down ~10%), I’m not sure that I would be buying there – the price:rent would still be pretty elevated at that point, and we have to be forward-looking with our decisions.

There is a kernel of truth to all that stuff people say about not perfectly timing the market and all that, so I wouldn’t be trying to pick the bottom once we get back into range of some semblance of a fair value. But even at 30% off, unless rents also spike, we’re still talking about very elevated price:rent levels around here. There’s no way to know if the R’s in our mental model will ever become the marginal buyer again, but for doomcasting, that would mean prices dropping something like 45% from the spring 2022 peak.