, , , , , , ,

The previous post noted that the Globe and Mail is writing a number of insightful articles about debt in Canada. I also noted that, especially for people who are concerned about real estate markets, the fact that the articles focus on individuals is misleading. If you are not one of those individuals then what you would really like to know is the extent of systemic risk.

Part 1 looked at the poor quality of the available data. This second part looks at some of the mechanisms which might link real estate markets, debt and systemic risk. It starts with two related questions:

  • If there were a problem in a real estate market, whose problem is it?
  • What is the mechanism which creates a worrisome link to another market?

Clear answers to these questions are critical, if only because house prices can fall at any time regardless of whether there is a bubble. To a buyer, this event is good news.  To the owner who experiences a capital loss, it is bad news. But, many owners hold onto their home for many years, and prices can recover to eliminate any paper losses. (The obvious exception is a senior who expects to use their wealth to fund their retirement, and there exist government programs to help.)

The obvious external effect of a crash, if it happened, would be that people’s wealth would fall and that, according to the permanent income hypothesis, consumption would also fall. The expected effect on retailers seem to be relatively small, especially if Canadians are not using the home as an ATM now.

Would the decrease in house prices decrease quantity supplied of housing (just as economic theory says!)? Yes. That, plus looking for evidence of excess demand, is why any discussion of price trends should be complemented by data on quantity supplied and quantity demanded.

The magnitude of this effect is in doubt. Construction is a largish share of GDP: about 7% of which residential construction is about one-third. The lags in the construction process which annoy so many people when they want to build something also discourage people from stopping a project once it has started. And even at the worst time, Canada’s population is growing and they need some place to live. So, again, the day after we learn that house prices in Canada peaked, construction sites will not start to lay off massive numbers.

The other big sector that might take a hit would be banks. To understand how the banking sector might spread the effect of a crash in a real estate market, it helps to look at the annual reports of the big banks.

(Actually, any student of business should get read the reports for a large company. Mostly, they are painful to read. They should teach the benefits of learning to write better . They should also reveal lots of practical details into what a company thinks about (which can help you during a job interview). Reading them thoughtfully, a student should think about “how did they calculate this number?”, “is the number bigger or smaller than I expected?” and “as an investor, is that the information what I would really like to know?”.)

Consider the 204 page annual report of Canada’s largest bank: the Royal Bank of Canada (RBC).   They talk a lot about

  • total value of residential mortgages: $215 billion plus $43 billion in HELOCs
  • average loan to value ratio: a little over 70 percent

Scary, but developing your number sense would give some perspective:

  • What are RBC’s total assets, owner equity, and income?
  • On average, and ignoring mortgage insurance, how far would the value of a home have to fall in order to imply that a bank loses money on the loan (assuming that it takes over the property and sells it at market value)?

(Answers: $940 billion, $49 billion; $9 billion; a little under 30 percent. But research has shown that many people continue to pay their mortgage even if it is more profitable to default. It would be easy to say that “The Sky is Falling!!” if the default rate were to rise, but number sense gives perspective. The effect of a rise in the rate from 0.5% to 0.51 percent differs from 0.5% to 2%, and even that bad case is less than one fifth of what the US experienced at its peak. The facts in Canada say, that the rate about 0.3% (i.e. 3 per 1000 residential mortgages) and has been falling since 2009. The annual report includes some sensitivity analysis in the section entitled Risk Management, but the language confuses me.)

There has been a lot of discussion of whether the implicit government guarantee on the mortgage insurance offered by CMHC (primarily, but not exclusively).  That could lead to the kinds of problems seen in other countries but that debate is much more complex: [1] [2] [3]. Sometimes, I think that commentators confuse “risk management” with “avoiding risk entirely”. It is impossible to avoid all risk, and insurance is intended to insure against bad outcomes.

To me, the evidence indicating that the potential for contagion from the real estate markets is not convincing. If house prices were to fall in Canada, there would be a lot of unhappy people. That unhappiness does not mean that house prices must fall next week or that that unhappiness would be any different than if the stock market fell next week. There is stronger evidence that poor logic plus bad, incomplete and non-existent information opens the door to people writing now about scary possibilities.