I’ll be honest here: I had originally selected a different photo for today’s blog.
While searching with the “weather” theme, I found a beautiful photo of a giant lightning bolt striking down from the sky, with city lights below. But then I thought, “Maybe people won’t see the beauty that I see? Maybe they’ll see the dark, gloomy, depressing side of this photo instead?”
So I went with the bright and happy photo that you see above!
There are two sides to every coin, and where one person might smile, another might frown.
As I write this blog on Tuesday afternoon, I’m currently involved in a 31-offer melee on 10 Fennings Street in Queen West. My clients might have been inclined to frown, and say, “I can’t believe the amount of competition on this property, we have no chance.” But instead, they smiled, and said, “You told us that most offers will be ‘throw-aways,’ so what price do we need to offer to be successful?”
In the downtown condo market right now, half of the market participants are rejoicing while others are lamenting. The latter half are experiencing frustration, regret, and, oh yeah, hindsight.
As Monday’s blog post detailed, the Toronto condo market is on fire. Yes, fire. I’m not even going to say that it’s “hot” but rather it’s on fire, because it is.
My colleague came to me in November and asked me to price her King West condo. She said the clients “wanted” $700,000. I responded with the same cynical answer that I always provide in a situation like this: “That’s what they want, eh? Why not $750,000? Why not $800,000?”
Because it didn’t matter what the clients wanted. That property was worth about $665,000 in that market.
Today, she listed the property for sale at $699,900 and there have been 16 showings booked already, with multiple agents drafting offers.
I have a dozen other stories that I could tell, but I’ll stop short of that right now and get to my thesis for today.
What if we knew when condos were going to increase in price?
What if we knew when they would increase in price relative to other housing types?
A colleague of mine, who I mentioned in Monday’s blog post, had floated a theory that I wanted to test.
How big can the gap between houses and condos go before buyers start to flock to condos once again?
So I decided to set out with as much data as possible to try to find a correlation. And that can be problematic; when you’re trying to find a correlation, as you can often be guilty of confirmation bias.
Therefore, I’ve decided to conduct this experiment in real time!
Because hindsight is ever-present in this particular experiment, I am going to post write this blog post as I examine the data, live.
Together, we’ll be the judge of whether or not we can use past pricing to forecast future trends.
First, let’s plot the 416-detached housing price against the 416-condo price:
This looks like two jagged lines to you, and don’t worry, it looks the same to me.
It’s the gap between the two lines that interest me, specifically the relative difference, not the absolute.
In January of 2012, the average detached house price was $743,993, and the average 416-condo price was $343,835. That gap is $400,158.
In December of 2020, the average 416-detached price was $1,475,758, and the average 416-condo price was $625,828. That gap is $849,930.
The gap has grown substantially! Right?
But on a relative basis, the gap has merely grown from 116% to 136%.
That is to say that in January of 2012, it cost 116% more to purchase a detached house than a condo, and in December of 2020, it cost 136% more.
The gap has grown massively on an absolute basis, but not so much on a relative basis.
So where can we look at the change in the relative differences and try to spot a potential uptick in condo prices?
The highest this gap has ever been was back in February of 2017 at 205%.
By the end of 2017, detached prices had declined by 20.6%, but condo prices had increased by 3.4%. We know that 2017 was an odd year in the market and that April saw a large drop in prices and sales, but this “peak” gap between detached and condo prices could have served as a sign of things to come in the market.
By the end of 2018, detached prices had declined, on paper at least, by 27.2%, but condo prices were up 15.3%.
This is all hindsight, but if you were able to identify when the gap in pricing was overblown, you could certainly identify when the condo market was going to rise.
Now let’s say you were sitting down at Christmas looking at the state of the freehold market and the condo market, and you wanted to forecast whether 2021 could see a rise in condo prices.
How would it help to look back at this pricing gap?
Again, with the benefit of hindsight, we have to be careful.
These price gaps pale in comparison to the 205% we saw in February of 2017, but the trend we can see is that the gap between detached and condos is growing.
What’s to say this gap doesn’t grow in January, February, and beyond? Nothing, perhaps.
Now, let’s move on to 416-semi versus the 416 condo:
To be honest, the 416-semi line is quite similar to the 416-detached, it’s just less volatile.
This gap peaked in February of 2017 at 111%. But that’s the same peak as we saw in the detached market, so we either have consistency or we have a coincidence. And if we do have consistency, is there any data from which we could have made an accurate forecast?
This gap was at its lowest in August of 2018 when it was merely 52%.
This gap widened to as much as 78% but narrowed back to 54% the following August in 2019. In that time, semi-detached prices rose 7.3% and condo prices rose 5.8%.
Now here’s the kicker: since August of 2019, semi-detached prices have risen 21.3% and condo prices merely 1.1% on account of the massive decline in condo prices in the latter-half of 2020.
So if we replicate the same chart as we did above, will we draw any inferences?
As with the detached chart, we saw the gap widen significantly by December, but these gaps in 2020 are far more volatile. Up, down, up, down.
So perhaps the averages won’t help us?
Let’s use the HPI instead:
This time, the gap is actually its widest in 2016 – June, to be specific, at 168%.
For what it’s worth, the February 2017 number isn’t far behind at 165%, so even the HPI is consistent with the detached average and semi-detached averages above.
The curve is smoother, and maybe, just maybe, if we replicate the same chart as we did above, we’ll see an indicator toward the end of 2020:
Okay, so the HPI is smoother and thus the gap widens without any volatility.
However, we still have absolutely no idea how large that gap will be before the market turns around.
And that’s the problem here, folks.
What’s to say that the January gap isn’t 118%, and February doesn’t check in around 122%, before the condo market turns around?
But it DID turn around!
I’m living it as we speak!
We live and die by the average prices, and many of us know by heart that the average condo price peaked over $722,000 in February of last year and finished 2020 just over $625,000.
(that’s $722,675 and $625,828 to be exact, but my memory was pretty good…)
So if we’re in the spirit of making predictions, how about one from yours truly?
When the TRREB numbers are released next week, we will see that average condo price over $655,000. And by the end of February, it will be over $675,000. By April, we’ll be nearing $700,000.
Screenshot it, print it, record it, I’ll stand by it.
But at the risk of making this blog post about something else (because it’s not about predictions, but rather to use data for accurate forecasting), can any of us look at the charts or statistics above and accurately predict the future?
I’m all ears if you have suggestions.
I think back to my university days, looking at technical analysis of stock charts.
Martin Pring and John J. Murphy.
Moving averages, Bollinger bands, and oscillators.
Is there a way to use a rolling average with respect to that gap between the price of a house and the price of a condo, and when a particular month’s gap exceeds the 6-month moving average, we set off an alarm bell? Or a 12-month moving average?
There has to be a way.
There’s just too much data to not find an accurate forecasting tool.
Only, I’m not able to find it.
I look at the data above and I see nothing that could be used as an indicator of a particular market segment moving one way or another, relative to a different market segment.
The truth is, the best indicator I have is still my “feel” for the market. It’s not a predictive tool, but one that tells me exactly where we are in the market – something the statistics often ignore.
The market is literally changing by the day. I spent Tuesday night on the phone until 10:30pm, and one phone call was with a lady who is thinking of selling her condo, and a different agent gave her a strategy that I thought was absurd. But then I realized: he probably isn’t out there every day. He might not have been out there last week, let alone tonight. You have to be in this market up to your neck to understand what’s going on. So even if we did have a forecasting tool that was based on past data, I fear the market could still do a one-eighty before you could click =SUMIF in an Excel document.
Today was a wild day. As you can tell, I finished this post after a long day and even longer night, and the agents I’m chatting with over text right now are all licking their wounds.
Folks, I can’t explain what the market is like out there right now. Houses, condos, low-end, high end. Durham Region. Durham! It’s all nuts.
To my colleague who wanted to know if we can use past data to forecast the condo market, I’m sorry, I don’t think we can. We tried! But I don’t see it. Do you?
Happy to hear from the analysts on this one. Also happy to email you my spreadsheets if you want to play around with them.
See you Friday, folks!Back To Top Back To Comments