Guest Post: Bubbles and Blunderbuss

(Note from Ben:  The following is a guest post from ‘John in Ottawa’, a frequent commenter on this site.  I have added a few of my own thoughts at the end of the post)

Recently, Canada’s Finance Minister Flaherty announced new regulations tightening CMHC insured mortgage and lending standards.

The consensus view is that Flaherty tightened lending standards to take the pressure off the Bank of Canada.  Carney has expressed considerable concern over a rising level of consumer debt.  The normal response by the Bank would be to increase interest rates to take some of the sizzle out of consumer lending.

I have long believed it is inappropriate for the Bank to use interest rates to control consumer spending and borrowing.  It is too blunt an instrument.  Were Carney to increase rates in order to slow down consumer borrowing, he would also slow down business borrowing at a time when the recovery is still quite weak.

It is easier, and much more appropriate, for the government to change regulations to control spending.

But is regulation also too blunt an instrument?  Is there a housing bubble that must be gently deflated?  Is the housing bubble responsible for too much consumer debt?

The first chart below is taken from the Teranet Canadian House Price Index going back to 1999, the first year for which there is data on all six cities covered by the index.  The index for all six cities is set to 100 as of June, 2005.

An economic bubble is loosely defined as a significant, usually exponential, departure from an underlying long term trend.  From the chart, there is a linear trend from 1999 through the present.  During the prior ten years from 1989 to 1999, house prices were relatively unchanged.  After June, 2005 only Vancouver and Calgary depart from the trend in a meaningful way.

The linear regression trend line shown is for Ottawa (chosen arbitrarily as representative) and shows a high confidence for a linear trend.

The second chart zooms in on the period from June, 2005 until October, 2010, the latest figures available.  Again, it is clear that only Vancouver and Calgary deviate from the trend.   All other Canadian cities are continuing to track a very linear trend.

Vancouver and Calgary may fit the criteria of a bubble.  They broke away from the trend of other major Canadian cities as well as their own trend for the preceding twenty years.  It is impossible to determine whether a bubble exists without also examining coexisting events, facts, and timeframes.

Over the past ten years, house prices have gone up faster than the underlying rate of inflation.  Some would argue that any asset appreciation that is out of line with inflation is a bubble, but that isn’t the general view.  Certainly, Canada’s house prices have not appreciated on anything near the order of the dot com bubble where prices went parabolic.

If we go all the way back to 1981, the furthest I can go for Canadian house data, house prices today on a national level are exactly in line with inflation over the past 29 years.  $100 dollars of purchasing power in 1981 requires $228 today, or in terms of the house price index, today’s national composite index of 137 discounted for inflation would be 63.5 in 1981. In other words, the inflation line in the charts above intercept current prices on the right of the graph and intercept 1981 prices on the left off the graph.  So it can be argued that over the past 10 years and more we have simply played catch up.

Even Vancouver and Calgary are not terribly out of line with long term inflation, although the absolute value of Vancouver houses is difficult to fathom.

What is it called when an asset class is priced significantly below the rate of inflation: a deep hole?

It is apparent from both Flaherty’s and Carney’s comments that they believe excessive consumer borrowing and debt is a direct result of withdrawing equity from over inflated house prices, and that this may lead to serious problems down the road when interest rates go up.

I postulate that the excessive borrowing against home equity is most likely to have occurred where house prices are the most inflated in relative terms – Vancouver and Calgary.  The evidence shows that there is no housing bubble in eastern Canada.  It follows that it is unlikely there is undue credit risk in eastern Canada due to Home Equity Line of Credit (HELOC).

I agree that the Bank of Canada should not punish business in order to reign-in consumers and I think changing federal regulations was clever and appropriate.  However, I also believe the government should not punish eastern Canada in order to reign-in Vancouver and Calgary.

Perhaps it is politically untenable to target one particular region of Canada in preference to another region, but it would certainly be technically feasible for the government to have tightened lending rules solely for those regions of Canada where there is undue risk.

Flaherty’s action this past week feels a lot like Dad took away the car keys because the boy down the street got caught speeding.

Ben’s two cents:

I have a great deal of respect for John.  Though I agree that the Canadian bubble is most pronounced in a few larger centres, I would not be so quick to accept that easter Canada is off the hook.  Nor am I entirely convinced that HELOC growth has been concentrated in the West.

I’d like to dig deeper into those questions.  In the meantime, here are my three major unanswered questions after reading John’s post:

1)  The inflation data is certainly interesting.  Since the Teranet index has nearly tripled the rate of inflation over the past decade, it suggests that we must have had almost two decades of negative real growth in Canadian real estate prior to 2000.  Can this be?  This strikes me as being highly suspect.  Something about that data is not quite sitting right.

2)  Inflation is certainly one factor in determining fundamental value of real estate.  However, I would caution that inflation alone can be misleading.  Cost-push inflation caused by increased input costs when those costs are driven up by more expensive imports from other nations would have the net effect of squeezing margins.  Though it would show up in the CPI, it has no true bearing on sustaining real estate prices as the ‘inflation’ can not not passed on in the form of wage increases.  I’m not suggesting this is necessarily the case, but I’m just highlighting the reality that inflation, if not paired with an accompanying increase in income, does nothing to support house prices.  Ultimately if inflation and housing both rise at 5% while incomes rise at 2%, you’ve still got a problem. 

I would be far more swayed by data showing that house prices have paced income gains over the same period.  The fact that we are approaching all-time highs in the price/income multiple across much of the country greatly concerns me.

3)  I’m curious why the Teranet index shows less of a growth in house prices than other house price indices? Other data sources (OECD for example) certainly show a parabolic rise in house prices.  As the Teranet index tracks paired sales, it could suggest that the average new house is getting larger and more ‘luxurious’.  Thus the bulk of the house price increases would be driven by new home prices which would only register in the Teranet index after their first sale.  This data is only presented as a percent change over the previous sale price. 

Could the effect of the McMasion generation be lost in the Teranet index but caught in other indices?  If so I see major implications with regards to demographics and the general trend in downsizing over the next decade. 



This entry was posted in Guest Posts, Real Estate and tagged , , , , , , , . Bookmark the permalink.

23 Responses to Guest Post: Bubbles and Blunderbuss

  1. LS says:

    If we go all the way back to 1981, the furthest I can go for Canadian house data

    This is a very suspect year to choose. In Victoria at least, 1981 was the peak of a huge bubble. Average inflation adjusted home price was $290k then and over 4 years it plunged 40%.

    So if you’re going to extrapolate from that year then you’re basically saying we’ve surpassed the bubble in the 80s and are really set for a big decline.

  2. Jordan says:

    As an elaboration on Ben’s point no. 3, I’ve often wondered if the Teranet index is corrected in some way for “turnover frequency”. For example, I would presume that the average condo unit is bought and sold more frequently than the average single-family home, so there are likely to be more “paired sales” of condo units than SFHs in a given time period. Does this lead to over-representation of the condo market in the index? What about new construction, and what about square footage?

    I’d put more stock in the Teranet numbers if the methodology used to calculate them were a little more transparent.

    • jesse says:

      Try going to the Teranet website. They actually publish their algorithm for you to look over. It helps to have a graduate degree in statistics to understand it, though. Outliers, such as major renos and rebuilds, are discarded.

      The same-sale methodologies like Teranet and the original Case-Shiller are viewed as the best measure of apples-to-apples tracking of the housing market over time. What a specific owner will care about is what his particular property is selling for over time. The Teranet index provides exactly that. If a McMansion is built, it is only compared only against itself and no others. All properties are then collated based on their individual price histories to produce the HPI.

  3. jesse says:

    “I postulate that the excessive borrowing against home equity is most likely to have occurred where house prices are the most inflated in relative terms”

    I would be careful to postulate that most of the debt has been taken on by those in western Canada. The debt levels Ben has been highlighting on this blog are averaged from across the country, where the majority of the population lives east of Manitoba. Surely a large % of the recent debt growth has occurred in these more populated regions.

    “However, I also believe the government should not punish eastern Canada in order to reign-in Vancouver and Calgary.”

    In my view the issue isn’t “punishing” households in one region of the country. The issue is whether the debt chains being offered, with government backing, is a good thing as a matter of policy. Bubbles may be occurring in certain regions — Calgary never recovered to its peak of several years ago BTW — but the risks of bubbles forming by continuing historically loose credit conditions through can easily start occurring elsewhere (or at least I cannot see any impediment to such). I would argue that certain areas within cities are feeling the effects of these conditions alongside Vancouver et al.

    As I have stated before, we can look at housing as an asset class. We have decent parallels with certain US jurisdictions on what acceptable returns are post-bubble, which should consider primarily future cash flows from operations (rents minus expenses) and not speculative capital appreciation above inflation. Contrasting parts of Canada and the US on this basis shows a distinct difference between their prices-to-rent ratios, enough to indicate speculation is at play, even in the lower-priced Canadian markets. Analysis of housing should be focused on earnings projections and yields, not price projections.

  4. Vince says:

    Two points. First, that is quite a steep trend for home prices, which over a longer time frame could be viewed as a bubble. A deviation off of that trend would be considered a parabolic type move, which should end the trend, as it did in Calgary. Second, the Canadian $ went up around 40-50% versus the US $ during this period, so the chart understates the magnitude of the real estate bubble in Canada. Higher Cdn$ means less exports and less money to pay for real estate. When real estate goes down it will be in US$ terms and in absolute terms.

  5. John in Ottawa says:

    It is nice to be able to give Ben a break today. It is hard to imagine how he can hold family and work together while putting so much time and effort into this blog.

    Excellent comments and questions today. I will use them to explore further into the issues addressed in my post. Clearly, there are many more questions and lines of inquiry to follow in getting to the bottom of the housing and debt issues in Canada along with their potential prescriptions and consequences.

    I’ll use this week to consider the feedback received and see what data I can dig up to shed more light. Regional data in particular is difficult to find. Hopefully, I can give Ben a break again next weekend.

  6. LS says:

    Sorry John, but the more I think about your post the more flawed it becomes. The decade of steep house price appreciation started in 2000. You only have data for that decade, and it is completely impossible to judge a bubble (or excessive appreciation, whatever you want to call it) by just looking at the bubble itself. If you can’t look at the longer trend then the data is useless for making that determination.

  7. Lumpen says:

    John, great post. You’ve got some interesting data and good analysis.

    However, I don’t agree with your definition of a bubble:

    significant, usually exponential, departure from an underlying long term trend

    Emerging markets, both in business and countries, exhibit exponential growth rates. Consider Internet companies – Yahoo, Ebay, Google, etc. – these had, or still have, enormous fundamental growth. Were / are they a bubble in their 100%+ growth phases?

    I would argue no, it’s not the rate of change of a fundamental metric that defines a bubble, but its valuation. People can disagree on whether Google should trade for 15x earnings or 25x earnings given its current earnings growth rate. However, I don’t read about anyone believing that it should trade for 50x.

    Expanding beyond equities, if an asset price cannot be justified by fundamentals, that’s where you’ve got an overvalued market. Is it necessarily a bubble? That’s more difficult to say. But I’d be interested on your take on the change in rents and cap rates for the 6 cities you mentioned. CMHC has historical data on average rents. Ex.:

    On a micro basis, I don’t see much in the way of rental stock that can provide a reasonable cash return when financed, even using today’s low rates. Again, not necessarily a bubble waiting to pop, but a strong indicator that future returns will be lower than those in the recent past.

    • jesse says:

      “Emerging markets, both in business and countries, exhibit exponential growth rates”

      That’s driven by income growth, not price growth. That is, a stock trades at higher earnings multiples because its future cash flows are expected to increase faster than a stock with low earnings multiples.

      For example, Hong Kong and Taiwan typically have higher price-rent ratios (real estate equivalent of price-earnings ratio with stocks). It’s not because real estate is more dear to own than rent, it’s because people expect incomes to outpace inflation for the foreseeable future. In Canada I see little evidence underlying cash flows from property are increasing more than inflation. Ergo the earnings multiple should be lower.

      • Lumpen says:

        Completely agree – rapid price growth with rapid fundamentals growth isn’t a bubble in itself (although can be if the impetus for the fundamentals growth is reversed).

        Looking at prices in a vacuum without the fundamentals can’t tell you if something is dirt cheap or wildly expensive, because assets are probably “fairly valued” about 5% of the time.

  8. mac says:

    Great post, John. Thanks.

    I wonder if house prices in Vancouver and Calgary were to flatten from here on in and the Eastern Canada “linear regression trend line” were to continue on its way up (in-non-bubble fashion as you assume), how long until those two lines intersect?

    If you are correct in your assumption that there may be no bubble in Eastern Canada, it makes me wonder why these Vancouver and Calgary deviated so much from the trend?

    As for tightening of policy for the benefit of two Western cities and to the possible detriment of Eastern Canada, I can tell you it never quite feels that way out here. My opinion before reading your post was that a bubble was forming in Toronto.

    Considering what happened to the business and the economy the last time Toronto’s RE bubble burst, it’s possible that steps are being taken to mitigate that happening again. And, again, considering your charts, those steps may be entirely effective for that region.

    Which leads me to wonder… if there is only a bubble in Vancouver and Calgary (with Calgary’s bubble already somewhat deflated)… what would be the effect of higher than normal foreclosure rates, underwater homeowners, etc. in two of Canada’s less populous cities on the CMHC, the Canadian taxpayer and the economy as a whole. It could be far less dire than any of us are thinking and certainly not comparable to what happened in the US.

    Please carry on with your line of thought as it is refreshing to consider another way this thing may unfold. Any work you do in answering my first question is much appreciated.

  9. John in Ottawa says:

    I’m still taking note of all the comments and will take them all into consideration as I proceed.

    A vexing problem remains, “What constitutes a bubble.” The definitions at WikiInvest and Wikipedia aren’t much help. Alan Greenspan and Ben Barnanke didn’t seem to know what a bubble is back in 2006.

    I would like to propose, for the purposes of this discussion, that a bubble (to extend the metaphore) is something that can “pop.” That is, it is an inflated asset for which the price can return rapidly to the un-inflated state, and more over, cause economic havoc as it deflates.

    After all, people lose money in all kinds of ways all the time. A so-called bubble that simply results in a few “greater fools” losing their shirts is about as interesting as discussing Aunt Sally’s donut recipe.

    I think our definition of a bubble should include a risk to the people and communities in the form of bankruptcy and default, and to the banks and/or the Canadian tax payor in general. In effect, the consequences of a bubble burst should look something akin to the banking devastation and subsequent tax payor bailouts in the US or Ireland.

    Thoughts, comments?

    • HHV says:

      John, the definition I use is:

      A speculative bubble is usually caused by exaggerated expectations of future growth, price appreciation, or other events that could cause an increase in asset values. This drives trading volumes higher, and as more investors rally around the heightened expectation, buyers outnumber sellers, pushing prices beyond what an objective analysis of intrinsic value would suggest.

      The bubble is not completed until prices fall back down to normalized levels; this usually involves a period of steep decline in price during which most investors panic and sell out of their investments.

      While each speculative bubble has its own driving factors and variables, most involve a combination of fundamental and psychological forces. In the beginning, attractive fundamentals may drive prices higher, but over time behavioral finance theories suggest that people invest so as to not “miss the boat” on high returns gained by others. When the artificially high prices inevitably fall, most short-term investors are shaken out of the market after which the market can return to being driven by fundamental metrics.

      In plain language, the questions you need to answer IMO are:

      1. Are people buying houses for a home or for an investment?
      2. If they’re buying them for an investment are they making money today (either saving on rent or have positive cash-flow from rents collected) or are they banking on them being worth significantly more in the near future (speculation)?

      If the answer to 1 is home, who cares?

      If the answer to 2 is pinned on the dreams of someone paying more for the home in the near future (within the term of the current mortgage), then it’s speculation pure and simple. If there are enough people acting in this manner a speculative bubble can form. Because we only have anecdotal data regarding the volume of “speculators” in our markets it’s highly likely we’ll ever be able to concretely answer this question unless the bubble bursts.

      Ben has already pointed to many current economic data sets that support a bubble thesis. He’s not the first blogger or real estate market pundit to do so. In 4 years of writing, I’ve never seen any strong economic writing argue against the existence of a bubble–all we ever see is a few choice quotes from industry-backed economists who suggest the Canadian situation is not a US-style bubble and therefore won’t have a US-style collapse.

      • John in Ottawa says:

        Thank you for the reply. Every bit helps to get us closer to an answer.

        I want to be careful to distinguish a bubble from a blister. All markets suffer from boom and bust cycles, but few of these cycles qualify as bubbles. Not, at least, in the sense of the Tulip bubble, or the South Sea bubble, or even the Dot Com bubble.

        This far, Canada’s aggregate housing market isn’t any more out of whack than it has been many times in the past. We’ve had corrections before.

        Certainly, Canada’s market hasn’t seen the six fold increase since 1986 that Australia has seen, or four fold increase since 1996 that Ireland experienced. In fact, Canada’s national composite index has not risen far enough in the past 10 years to reach the CS index that the US has fallen to.

        We may be suffering from bubble envy!

        Have Vancouver and Calgary deviated from the national trend and are they likely to experience a correction? Yes and probably. It seems to be in progress.

        Is Canada as a whole likely to experience a correction? Someday because we always do, but it isn’t at all clear when that may be, what it will look like, or if it will by any different from the two others I have been through since I bought my first house in 1976.

        What will be the systemic consequences if any? Well, at this point we are trying to determine the probability that there will be systemic consequences.

      • LS says:

        I think your definition of bubble is too extreme. Do we have the biggest real estate bubble in the world? Certainly not. That doesn’t mean were not highly overvalued. We can argue semantics about exactly what constitutes a bubble, but really that is beside the point. It doesn’t matter what you call it, we’re in line for a significant correction of some sort.

        but it isn’t at all clear when that may be, what it will look like, or if it will by any different from the two others I have been through since I bought my first house in 1976.

        Those were some pretty serious corrections. At least 1989 I would call a bubble. After the 1981 peak, prices dropped almost 25% nation wide. After the 1989 peak they dropped the same amount and it took 15 years to come back to that level.

        It won’t be as bad as the US, but even if it repeats that cycle it won’t be anything to laugh about. I don’t think anyone but the extreme fringe is predicting chaos and bank failures.

  10. Potato says:

    The linear vs exponential thing is a bit of a red herring: at growth rates of ~5%/year, with even a bit of noise, it’s very difficult to spot the difference between linear and exponential growth looking only at 5-10 years of data. Nonetheless, that’s all that’s needed to move a large component of a typical family’s budgetary expenses like housing from normal to unaffordable, and set the stage for a collapse/correction.

    I think our definition of a bubble should include a risk to the people and communities in the form of bankruptcy and default, and to the banks and/or the Canadian tax payor in general. In effect, the consequences of a bubble burst should look something akin to the banking devastation and subsequent tax payor bailouts in the US or Ireland.

    I’d agree with the first bit, and on that count I think housing in Toronto, Vancouver, and a few other cities is already well into bubble territory: a correction from these levels will leave current buyers with negative equity, unable to sell if they need to, some facing bankruptcy, many facing significantly tightened budgets due to their debt load taken on in the heydays.

    I don’t know if it needs to be quite as brutal as the US or Ireland experienced, though: the system survived the 1989 Toronto housing bubble mostly intact; likewise the tech bubble pop of 2000 didn’t lead to widespread chaos and bailouts, but few dispute it was a bubble.

    I agree with LS: It doesn’t matter what you call it, but trouble is brewing. It only takes a few dozen percent drop in real estate to really hurt families, and it doesn’t matter much whether you attach a bubble label to that or not.

  11. Pingback: Asset bubbles according to Jeremy Grantham: Applications to the Canadian real estate market | Financial Insights

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s