Today’s the day we get to stacking models for Canadian real estate price targets. Last week I was asked by the Canada Mortgage and Housing Corporation (CMHC) to speak at their Housing Finance Symposium. These are pretty much my presentation notes, with more detail for non-technical audiences. In part 1, I discuss using a home price to income ratio model to show how unlikely it would be for real estate prices to catch up to incomes. In part 2, we use a median credit exhaustion model to see where people psychologically feel the risk/reward ratio of buying is too high. Today is the third and final part. We’ll combine both models to extract price targets, timelines, and recoveries.
Buy A House If You Want
Seriously. If you want to buy a house, buy whatever you want, whenever you want it. As long as you aren’t stretching yourself thin on payments. If you can afford to diversify your assets, you’ll probably be fine. I’ve said it before, and I’ll probably say it for as long as I deal with housing finance: There’s a cost of attached to what you want, that can’t be quantified.
Some people want to raise their kids in a house, others have a dream of just owning. If you buy something, and you lose a little money on it – consider that the cost of being you. Feel free to leave now if you’re the kind of person that watches HGTV, and dreams of your picture perfect housing future.
The rest of this piece is for people in housing finance, urban land banking, or that are curious how yesterday’s model translates into price targets. This read is for people like me, that have nearly zero emotional attachment to housing or money.
Layering Models For Precision, Confirmation, and Better Accuracy
Now, I can’t repeat this enough – a model isn’t 100% accurate. However, we can combine models to help build better precision and accuracy. You would be shocked at how close some financial models get. Today we’re going to combine the home price to income ratio, with the median credit exhaustion model. You don’t have to remember everything from the last two articles in the series, but you have to remember two things. One, home price to income ratios demonstrate support at 30% lower than peak, and prices take off again once hit. Two, median credit exhaustion models might be a solid indicator of when home prices begin growing again. Got it? Sweet.
Note: 2017 is partial, year-to-date. 2018 to 2021 are projections. I’m good with numbers, not a time traveler. Source: Better Dwelling.
Currently we’re at a home price to income ratio of 9.56, and if we assume the support is 30% lower, we get 6.69. Combining credit reduction (derived from the median credit exhaustion model), interest rate projections, and factoring the current rate of income growth – we plot the ratio. We would hit support in 2021. This is kind of interesting because it implies confirmation of a bottom. We now have two models that hit this number. Incomes would need to grow at a similar rate to the past 5 years, otherwise prices decline for much longer. This did not happen in the early 1990s, causing a little hiccup after prices bottomed, before they started climbing again.
On the upside, hitting the support in the same year could imply a quick bounce higher. Although as we move through the predicted plots, we would update the model with final numbers. Income stagnation, interest rates stalling or being slashed, etc. will have an impact on timelines. However, we’ve factored in everything we know at this point, and the bias is even a little higher to the upside.
Now remember, this is only two models combined. The more you can combine, the higher the rate of potential precision. Currently we use just over 200 different models to make our internal indicators. To be honest however, the dates are pretty similar, and the price targets are pretty close using just these two.
What This Means For Prices
Boring, get to the price targets – right? These two models when combined can be used to get a print on numbers. Today an average home in Toronto costs $780,000, and that’s just an average home – not a detached. By 2021, this model shows the cost of an average home in Toronto would be around $574,000 in 2017 dollars, about 26% lower. Now, those are 2017 dollars – not the sticker you would pay in 2021.
Let’s assume inflation hits the target every year for the next 4 years. In the unlikely event that it does, an average home would cost around $621,000 in 2021 dollars. That’s a sticker price that’s 20% lower than today’s price. If you’re wondering why I said “unlikely,” it’s because inflation has been about 25% below target for the past 5 years. An increase in interest rates, would make it even less likely for inflation to hit target. I’m an optimistic guy though, so let’s go with FrannyMo’s “firing on all cylinders” and assume we hit those targets.
Do OSFI B-20 Regulations Change This Model?
Yesterday, I’ve never received so many emails from portfolio managers asking the same question: Does the model factor in OSFI’s new B-20 regulation changes? No, it doesn’t. However, B-20 would have minimal impact on the model at this point. Our median credit exhaustion model isn’t based on the maximum credit available. Instead, it maps perceived consumer risk involved in taking out loans.
In my opinion, B-20 does reinforce that 2017 is the bottom of credit exhaustion. Credit isn’t just perceived to have high risk for home prices in Toronto at this point, but a cap has also been placed. This cap ensures that middle to lower income investors don’t max out their leverage. It sounds harsh, but let’s be honest – high income investors aren’t buying into this risk right now. Scotiabank CEO Brian Porter probably didn’t sell his house just six months after buying it because he thought it was the perfect time to invest. Ditto on why he ended up slashing the price ten months into ownership, to below his purchase price.
If you’re a millennial that read through anyway, close your mouth. Once again, modeling isn’t an exact science, it’s used to help mitigate large investor risk. Banksters and government employees are the ones interested in this. My goal here is to help millennials understand that you’re an active participant in this cycle – regardless of whether you go in blind knowing that. After all, money doesn’t care if you were unaware that you bought at peak.
You need to establish an appropriate personal risk model. Do not convince yourself every buy is an investment, especially if you did no research. Real estate almost always runs in a cycle. Figuring out where you are in the cycle, helps to determine if your buy is actually an investment, or a lifestyle choice.
If you’re from another part of Canada, sorry we only used Toronto for this presentation. Here’s a national read using the OECD House Price-To-Rent Index we ran earlier this year, showing that home prices could drop as much as 28% across the country.
Like this post? Like us on Facebook for the next one in your feed.
We encourage you to have a civil discussion. Note that reads "civil," which means don't act like jerks to each other. Still unclear? No name-calling, racism, or hate speech. Seriously, you're adults – act like it.
Any comments that violates these simple rules, will be removed promptly – along with your full comment history. Oh yeah, you'll also lose further commenting privileges. So if your comments disappear, it's not because the illuminati is screening you because they hate the truth, it's because you violated our simple rules.
What about external shocks such as a recession? Say NAFTA is dismantled…
Great article! Do you anticipate that detached homes will drop more in value than the condos and townhomes? Or do all homes tend to have a similar drop in value?
Typically there’s an acceptable gap between the two types of homes. If detached units drop, eventually condos will. Condos trail as a result of lower income buyers jumping into a trend, after it’s happened. Liquidity from smart money is usually picked up by less sophisticated investors.
“Liquidity from smart money is usually picked up by less sophisticated investors.”- can you explain what you mean? I’m a bit of a novice here.
Sure. When housing becomes a commodity, it stops acting like shelter and attracts investors. Investors treat housing the same way they treat all investments, and send it through the same cycles.
Phase 1, high-risk investors. These are people that buy when a market looks devastated. When people think there’s zero chance of a comeback, they begin scooping up houses. This is what US hedge funds did after 2010.
Part 2, industry. These are people that see what high risk investors did, and get a payday. This is where smart developers jump in.
Part 3, smart money. This is when you see home flippers jump in, and start fixing houses up. They’re not all that smart, but they’re scalping and making a profit. They brag to their friends. Stage 1 investors are selling to these people. This is kind of like when bank executives and pensions started lowering Canadian real estate holdings.
Part 4, wannabe smart money. These are people that see how “easy” it is for stage 3 investors, to make a living. It’s not, they just make it seem easy. This is where you get people that flip a couple of places for a salary type amount of money. They brag how easy investing is. Stage 2 investors are selling to these ones. This is where media talks about how easy it is.
Part 5, mass “investors.” These are the people that justify huge premiums, or think they’re going to be “locked” out. All stages before sell to these people. If homes were a stock, brokers would be jokingly calling them “bag holders.”
Stage 5 investors are the ones that indicate a top. When everyone starts saying something is a good investment, who are you selling to when you need a profit?
When these people lose a buttload of money, stage one starts again.
The end of this article links to a similar stage breakdown, where they use 4 stage cycles that they teach at Harvard business school. It’s a much better breakdown than I can do in a comment.
Nate that’s an interesting analysis, and we’re somewhere in the latter stages, but at the end of the day, nobody really knows where we are in these cycles with any certainty, including ‘smart money’. Otherwise, there would be many more Warren Buffet’s out there.
Indeed, we only see several (sometimes only 1 or 2) such cycles in a lifetime and therefore have as much or as little as 50/50 odds of getting it somewhat right. Then again, one has to be not only aware of how cycles work but have money to invest in the first place to take advantage of them.
I think essentially he means that the “smart money” is the educated investor that knows when to exit an asset holding (such as real estate). So the unsophisticated (ie. uneducated; someone who didn’t do research and bought on a mass trend) investor is buying the educated/sophisticated/”smart money” investors asset, without realizing they’re providing liquidity (cash) to the unsophisticated (chump) investor.
Or in other words. The cash a wise investor gets is from the stupid guy buying into the fad without knowing where the market is headed.
Oh apologies, had this tab open for awhile didn’t see the response.
Basically, it means that ponzi schemes always collapse when they run out of new money, but those on the bottom walk away with a fortune.
Perhaps ‘Liquidity for smart money is usually provided by less sophisticated investors, as the smart money abandons the market (liquidates)’ would be a better way to put it.
Condo townhouses and condo apartments will drop substantially more than detached houses since condo townhouses and condo apartments are still up substantially i price year over year whereas outside of the Toronto core detached houses are down year over year in most cities in the GTA.
Another razor sharp and precise analysis. I absolutely agree, buying a home isn’t always a profit maker, and people need to understand it’s okay to buy a home that doesn’t make money. This is the reason money managers want you to understand the concept of diversification.
Sometimes a lack of diversification works out. Sometimes it doesn’t. Ask yourself, can I afford to lose my only options in my old age? Most people, that’s a no. You don’t want to find out you’re 30% light on what you expected by retirement.
A persons age is relevant to buying versus renting. A young person can ride out a decline in property value because it will eventually rise and surpass its previous peak and the owner has earnings from a job to support them but a person approaching retirement who is relying on using the money from the sale of their home to enable retirement will either have to continue working, if the can, or be forced to reduce their expected retirement lifestyle or both.
why don’t banks lend money based on Income Ratios?
Because banks aren’t your friend, they’ll lend you the maximum you can carry. There’s little risk for Canadian banks, since all loses are socialized amongst taxpayers.
There’s a number of policies that came in the 1980s, because Canadians are too dumb to realize banks are private industry. One, there’s no risk since they’ll get a bailout – even on uninsured loans. Two, bankruptcy doesn’t get rid of housing debt in most provinces. Laws are built to make profits for banks, not ensure we make the correct financial decisions.
Banks spend tons of money to create and maintain the image that they are somehow looking out for your interests. Unfortunately it works on most.
Canadians eat this up too. The best thing banks have done in Canada, is convince Canadians they’re like government. Guess what? High ranking government is NOT allowed to say what they think, they’re restricted from making downside predictions.
[…] Part 3, combining models to build better forecasting timelines is now published. […]
All markets are not equal. We focus all of our attention on TOR/VAN, I’m concerned with the suburbs. If these overheated markets correct by 20-30% what happens as the money contracts inwards? I love this site but ‘canada’ is mentioned and then the articles go back to these two markets with little concern for the outer regions.
Can you look at the last crash and the impact? I know the demo/psycho has changed and people WANT to be in the city while the 80s crash was the opposite but there must be data showing what happens when the $$ contracts inwards. Please help us all understand this…many of our parents live in the suburbs and small communities…
The fringe areas around Vancouver will fall in price because of the new B20 OSFI rules and the fringe areas around Toronto will move up in price (so will all areas of the GTA) starting around March 2017 as interest rates fall starting around November this year in Canada.
Why would interest rates fall one month from now? That’s not happening.
As someone that saw your talk, I thank you for distributing this to other people. My employer actually paid to send me to this conference, to hear about the risk that Canadian homebuyers present to our organization. I hope you get to keynote next year when everything crumbles.
A crash requires one component beyond these conditions, that people don’t expect it. Despite the fact that you’re making enterprise level analysis available for mass consumers, many of them a) don’t care, or b) don’t believe you. The irony here is this is the trigger for a crash, that they think it’s unexpected. When it happens, I hope people look back and say “the banks paid the government to assemble a group of people to explain how much homeowners will lose, but I believed my Realtor’s weekend of education instead.”
Thanks for this three-part series, it’s is the most sensible analysis of the situation I have seen so far. As you mention in the article this forecast has a somewhat optimistic tilt. The assumption that the income growth will not be affected by this correction alone requires a lot of optimism to stand. There are other downward pressures that can easily materialize in this time frame. For example, it’s not a secret that participation of the foreign capital played a substantial role in development of this bubble and for some inexplicable reason some people believe that this participation will always be pushing prices upwards. I doubt that this participation has any correlation with income levels in Toronto…
Could you please do a similar analysis for Alberta market (Calgary and/or Edmonton)?
Me to, I would like an analysis for Calgary and Edmonton and Montreal as well.
In one word “roadkill” because of the masses leaving the entire province of Alberta. The fallout from the NAFTA and the gross overvaluations of detached houses compared to the prices of townhouses and apartments. Single detached houses in all cities in Alberta will plunge at least another 25 percent in price while townhouses and apartments fall to a lesser extent. Even apartments in Moncton and Saint John, New Brunswick cost quite a bit more than in all cities in Alberta except Fort McMurray.
I think you’re underestimating the impact of B-20, especially on that intangible, immeasureable thing called market psychology. If buyer’s can’t borrow enough for their “forever home”, or even a starter condo, listings will pile up, and open houses will be empty. Sellers will eventually get desparate, especially stage 5 “investors” and flippers, and prices will tumble. Buyers will continue to wait on the sidelines, not wanting to catch a falling knife…
Your prediction of a -30% bottom in 2021 may fit your model, but doesn’t measure the fear that can overtake a market, just as greed has overtaken price expectations these last few years.
I have lived through two major market corrections in the US; prices in California dropped an average of 30%, but in the outlying suburbs of the east SF Bay Area, such as Tracy and Stockton (think Maple Ridge and Mission), prices dropped by 50% and more. Condos were slaughtered.
People’s capacity for fear and greed, and how these emotions contribute to irrational decisions regarding real estate just can’t be quantified and inserted into a predictive model.
The B-20 rules will turn out to be nothing more than a head-fake. Investors and speculators will storm back into real estate as interest rates fall in Canada starting next month.
Please don’t take me wrong but if the world was working based on your models, the prices in such cities like London, Manhattan, New York, San Francisco, Singapore, Berlin, Paris and many others won’t be where they are today with prices. If you use average incomes in any of those cities to predict real estate prices, you won’t get anywhere. The fact is that today average income person cannot afford to live in a nice city. But there is enough very rich people “competing” to get a decent place in decent city. So try to revise your models to account how many rich people there out today who can pay whatever for a nice place to live in. Somehow Toronto is seen as one of those places. So good luck with your models that use average incomes as base of a theory. If Toronto indeed became one of those cities, then any model that uses income vs accumulated wealth/prestige just does not work.
You can always tell the clown that just bought a condo vs someone that understands markets. Let’s start with prices are falling in London, Manhattan, San Francisco, Singapore, and Paris.
Berlin’s housing is a third of the cost of Toronto. Germans have an extensive rental program, and a government that prioritized anti-ownership programs.
The Canadian government is currently planning a destigmatization of rental program, that will be launched soon. It’ll be focused on assisting loans for rental buildings.
On one side, we have models pointing towards falling prices, the World Bank, the bank CEOs, the IMF, the Governmemt of Canada, the CMHC, and the BoC saying home prices are in a bubble. Then, the government is going to make an active effort to stop homeownership.
On the other side, we have geniuses like you that say “world class” and “Next Manhattan” as an argument against it. In your circle jerk groups, you might feel like a genius. In the rest of the world, you sound like a crazy person.
But the point you’re missing is Poloz will always put the Canadian economy first before the housing market. When the Bank of Canada rate starts falling early next year money will flow back into real estate.
I kind of fail to see our rate falling as U.S raising their rate probably early next year (or even end of this year). Our loonie always catches up with the dollar somewhat (to avoid capital flight), doesn’t it?
That’s correct, Kurt, the BoC is practically in lockstep with the US when it comes to movement of interest rates. It’s been this way for decades. Besides, even if rates theoretically were to fall again, a few basis points is not enough to cause a flow back into real estate to any significant degree. It wasn’t interest rate hikes that caused the market to take a pause after the first quarter of this year. Government regulations put a chill on the market and the new stress test in January 2018 is bound to remove a lot of buyers from the pool regardless of any tweaks to interest rates.
If you compare what happened in 1989 in Toronto to what’s happening today some analysts were saying early this year that the peak of the bubble was still about two years away (2019) considering mortgage rates stay the same, income keep growing and debt affordability taken into consideration as well.
However shortly after that the Ontario housing plan was introduced and the mortgage rate is now up by about 0.5. So average prices fell from April and the bubble momentum got broken before maturity in 2019 as per some analysts.
So if those analysts were right about the 2019 peak assumption this means we’re back to maybe a half bubble territory and the prices are set to go up again in the next few years but now with the introduction of B-20 and the expectation of rising interest rate the price will flatten out instead of rising again which coincides with what some analysts from banks are predicting as flat prices for a while.
So all in all I find it a bit confusing with different views from different analysts and it’s getting more and more complicated to make a decision particularly with all the uncertainty in the market today.
If I have read all three articles correctly, this model does not pinpoint that the decline will begin this year. It tells you if/when the decline begins how much the prices will fall over how much time.
The credit exhaustion model does seem to indicate that the probability of the decline rises with falling index used by the model. So higher the price to income ratio compared to the model’s threshold the higher the chance of people’s sentiments tuning bearish and triggering correction down to the market’s support level.
Basically, what you are saying is that house prices are going to drop because we have run out of enough people with enough money to buy them at the current prices, and for various reasons houses still have to be sold.
So I am really late to the game here. I appreciate all that you’ve laid out and I am very much a novice in this area. Just have one question. In your other posts, you focus on Toronto, which makes sense. Is there anyway someone can take your model and apply it to their local area, if they have the appropriate data? Given that real estate is local and trends are regional (not always national) in nature, it would be great to see a break down by major urban areas in Canada.
Thanks again for putting this out there. Fascinating stuff.
Comments are closed.