Blessed by the Potato

Wayfare sent me an interesting, thought-provoking article in the Toronto star “Sizzlers and Fizzlers”. I shouldn’t say article, it really doesn’t have much writing to it, no analysis or commentary. It’s an interesting factoid. The PDF with what juicy information there is is here. The headline number is that in the last 40 years, Toronto real estate was up 1250% (since 1969), compared to the TSX being up 1244%. This implies that real estate was just as good if not a better investment over the long term than the stock market (once you factor in carrying costs, stocks look like the better place to keep your money, but then you can really leverage the hell out of real estate).

That really put the tingles to my scientist senses* (like Spidey senses, but for nerds… though Peter Parker was a bit of dweeb himself), since this does not jive with just about everything else I’ve read. It makes me want to look deeper into the methods to see how these conclusions came about; it sets off my Saganist “extrodinary claims require extraordinary proof” alarm.

* – Humorous aside, we were talking today in the lab about curious results, including one study that exposed subjects to a 13-hour and 18-hour exposure period every day. Explaining where they found their 31-hour days would have made for a much more interesting paper.

Just about everything else I’ve read, which is more general and not specific to Toronto, puts the compound inflation-adjusted rate of return for housing at about 1% or so (so the nominal rate should be 3-5%). So to see the 6.9% figure in this article (the same as stocks) was a little baffling. Is Toronto really that special? Was all that other stuff about long term trends I read wrong?

Unfortunately, the article does not give anywhere near enough information to say for sure. There are some inconsistencies though. First off, their numbers for the stock market look to be about right, in terms of the nominal prices over the years, but don’t look to include the return from dividends, which should improve the stock market figures by a fair margin. I also have to wonder if the choice of starting year in 1969 could affect things, since the 70’s were a rough decade for stocks — would the picture look different starting from 1959 or 1979? (I don’t actually know, but it makes me wonder) Some areas (and Toronto’s overall average) have not, in inflation-adjusted terms, come back to their 1989 peaks yet, and my dad was saying not too long ago that the housing boom had just recently brought their house back up to that point. Naturally, some areas have been hotter than others. Surprisingly, Willowdale (C-14) was one of the very coldest areas in the study, which just doesn’t agree with the fact that it’s such a desirable neighbourhood. Was it just simply already a wealthy, desirable neighbourhood in 1968, so when the data started in 1969 there wasn’t a whole lot of up left to go? I doubt that, since it used to be, from the perspective of a kid growing up there, a “normal middle class” neighbourhood, and is now full of, well, rich people and condos. The average home price there according to the PDF, $390,520, is almost exactly the same as the overall Toronto average at $390,839 (but it was a good deal above average at the 1989 peak, hence the underperformance). However, a quick trip to MLS will show that the lowest non-condominum asking price is $420k, so how the average got to be $390k in a neighbourhood full of million dollar homes is completely beyond me. The lowest price for a condo is in the $160k range, so it’s possible that all the new condo development in the area is really bringing down the average. But what does that mean for the overall study then? North York’s condo buildup started just a little bit ahead of a lot of other areas, so is the fact that what must be a large number of these units are now being flipped and bringing the average of “resale” homes down disproportionately hurting that area? Likewise, are the people who recently sold their SFHs in the “gentrifying” areas like Leslieville to developers for gagillions of dollars bringing the average way, way up because the condos built on those sites haven’t had a chance to be flipped yet? Since they’re using average values and not medians, are a few $25 million dollar homes in a few areas really skewing the results? Do these results in any way have bearing on the reality of owning a home in Toronto over the past 40 years? Do they offer any insight for what might be in store in the future?? Disturbingly, some of these questions suggest that, for example Willowdale, might have performed even better than the factoid’s already generous rate of return. Or, that the really great returns will come from places that are presently farmland, and not areas that currently look good. Or, that it was a really good idea to own real estate in the 70’s and 80’s, and a really bad one in the 90’s, with the 00’s being a bit of a wash, so far.

The study seems to have enough holes that I’m not going to reject my earlier information that suggests that housing, in general, underperforms the stock market. It is possible that Toronto was a special exception in the last 40 years compared to the overall averages of US (and more rarely for data discussion, Canadian) real estate. After all, in the last 40 years Toronto went from being a backwater city up in, you know, Canada, to being a world-class metropolitan city. Toronto had a huge boom in building and real estate in the 70’s: that’s when almost all of the recognizable downtown was built up: First Canada Place, Royal Bank Plaza, the CN Tower, etc. Two-thirds of those impressive 40-year real estate returns were due to gains in the 70’s and 80’s (yes, it’s cherry-picking at a market peak, but in the second 20 years of this study, housing only returned 2% nominal straight-line returns per year, with negative growth when adjusting for inflation). In just the last 10 years the suburbs have expanded so much some people have started calling them the exurbs, another period of massive growth for Toronto, and also what turned out to be another real estate bubble in the States. So while the headline 40-year overall nominal return number is attention-grabbing, especially next to the very similar number for the nominal TSX return, I’m not convinced that a house is the best place to put my money.