Scotia calculates the risk of sharp housing correction is ‘low’; Gaping holes in their logic

I want to spend a bit more time looking at the Scotia Economics report on Canadian real estate trends released yesterday.  The report can be found here.

In particular I want to zone in on page 4 which has as a title for this section, “Risk of a Sharp House Price Correction Remains Low”.  I’m frankly quite shocked at some of the inconsistencies and glaring holes in the logic of the authors.  Let’s first highlight the key quotes from this section:

Considering a variety of housing valuation metrics, Canadian home prices are overvalued relative to their ‘fundamental’ or ‘long-term’ trend. By one common measure — price-to-income — valuations are approaching record highs. As of the third quarter of 2010, average home prices (measured by the national average MLS price) were a near-record 5.7 times average annual disposable income (measured by disposable income per worker). This compares to an average ratio of 4.1 over the 1980-2010 period.

The steady rise in Canadian home prices over the past decade has been underpinned by favourable demand fundamentals — a strong economy, record employment rates, low borrowing costs, and rising immigration and homeownership rates.  It also reflects some catch-up from the underperforming domestic housing markets of the 1990s.

At the same time, it is not particularly surprising that house prices have overshot in recent years given persistently tight resale housing supply and unprecedented low interest rates.

I’ll chime in here just to say that up to this point there is little to disagree with.  Indeed we have experienced strong economic growth for much of the past decade.  However, I would suggest that during that time, and particularly since about 2005, our economy has become increasingly reliant on consumer spending, itself based on credit expansion, to sustain the economic growth.  As I’ve noted many times before, Canada arguably has seen credit excesses in the past half decade that certainly suggest that if we’re not experiencing a credit bubble, it’s the next closest thing.  We’d be remiss to observe the unparalleled rise in consumer debt in recent years and not understand how that debt acts to buoy the broader economy, a theme I explore often on this blog.

The bottom line is that the economy looks fantastic at the tail end of a credit binge.  Unemployment is low, economic growth is high, and people project that into the indefinite future.  But then a funny thing happens:  People have to repay the debt.  That’s money that would have been available to sustain consumption (65% of our total GDP), but instead is now being diverted back towards debt repayment.  The impact on economic growth should be obvious.  At any rate, back to Scotia.

It is widely expected that the housing boom of the past decade will be followed by a period of softness.

While a ‘period of softness’ is well within the real of possibility, experience tells us that it is very unlikely.  I’ve explored this in a previous post, so I won’t elaborate too much here.

While housing affordability has become more strained after years of rising prices, it does not follow that the ratio of home prices relative to income must return to its long-term average. This ratio has been trending higher for several decades, corresponding with a rising aggregate household debt-to-income ratio.

Several factors help account for the rising long-term trend.  Canadians today are more comfortable carrying a larger debt load (the biggest share of which is mortgage debt) relative to their annual income than in the past due in part due to financial product innovation and rising household net worth.

That last paragraph is frankly ridiculous.  The fact that people are comfortable with their debt loads has no bearing on their actual ability to sustain those debt levels.  In fact, according to recent surveys, 1 in 10 Canadians could not handle an unexpected expense of $500.  Even more alarming is that according to a stress test scenario by the Bank of Canada, it would take only a 0.5% increase in interest rates for 1.1 million Canadian households to become at risk of defaulting on their consumer credit or mortgage-related debt.  Now that strikes me as extremely high, but let’s not miss the point:  People are maxed out and are increasingly at the mercy of interest rates.  That they are ‘comfortable’ in their current position has no bearing on how this will all play out.

However, Scotia is right that house price to income ratios have expanded primarily as a result of falling interest rates and financial innovations.  Let’s cut through the banker talk and acknowledge that ‘financial innovations’ equals ‘extended amortizations and lower down payments on CMHC-insured mortgages‘, a relatively new and untested experiment in CMHC’s nearly 70 year history.

Some economists have calculated that between 40% and 70% of all house price increases since 2004 are directly attributable to expansion in allowable amortization.  Though we may disagree with those specific numbers, certainly longer amortizations have had an impact on house price to income ratios.

Furthermore, it’s awfully difficult to miss the fact that the price/income ratio is negatively correlated with interest rates.  That is, as interest rates rise, the price/income ratio falls.  So it’s no wonder that we’ve seen a consistent expansion in the price/income ratio since 1980 when we look at how interest rates in Canada have moved in that time frame:

The more important question is will these same dynamics support a further expansion in price/income ratios going forward.  The answer is clear.  We are entering a period of time that is fundamentally different from the last decade.  Rather than extending amortization lengths, the government has moved to shorten them….twice now since 2008, with the new rules set to take effect in just over two weeks.  Furthermore, interest rates are still hovering near historic lows.  While the speed at which they will rise going forward may be up for debate, certainly the direction is not.  Back to the Scotia report…

Consistently low nominal interest rates, urban intensification, and shifting lifestyle and investment considerations may also have contributed to a structural — and sustainable — increase in the (house price to income) ratio.

Taking into account the long-term rising trend in house prices relative to income, the current overvaluation in average Canadian home prices is probably around 10%. While based on a fairly crude analysis of one housing metric, this is consistent with several other estimates, including recent research by the IMF. A 10% overvaluation is fairly typical toward the end of a strong housing expansionary phase.

We expect the adjustment to restore long-run affordability will be relatively orderly given a low share of high-risk mortgages, only gradually rising interest rates, and continued population and income growth. Assuming flat nominal home prices (i.e. modest real price declines) and annual income growth of 2% per worker, the average price-to-income ratio would return to its long-term trend by mid-decade.

I don’t know where to begin with these last paragraphs.  Some of the statements are so absurd that it boggles my mind that this could be written by a bank-employed economist.  When I see such glaringly false statements, I’m reminded of two posts I wrote dealing with the ability of bank-employed economists to be completely forthright in their assessment of the housing market:

Why the big banks have to lie through their teeth

Are the big banks really concerned about the consumer or are there ulterior motives?

Note first of all that the 10% overvaluation estimate is based on the rising trend in house prices relative to income.  Note also that the authors refer to this as a sustainable increase.

Now it’s hard to know if they mean to imply that the rate of increase is sustainable or that it is sustainable at its current level.  If it’s the former, it’s a flat-out ridiculous statement.  If we followed the rate of growth out another 20 years we would find that house prices would be almost 7.5 times income.  At current interest rates of roughly 3% and assuming the home was purchased with 5% down on a 30 year am, it would consume 40% of income just for the principal and interest payments alone.

But if interest rates normalize closer to 6% (which they certainly will over the next two decades), it would consume nearly 60% of income.  Add in taxes, insurance, and maintenance and you’ll be pushing 70% of income just to carry a home.  Sure it’s theoretically possible, but in practice the economy would wither as virtually all money would be spent on the very basic necessities leaving little for discretionary spending.

If in fact the authors are implying that the price/income ratio is sustainable at its current level, I would also beg to differ.  The full weight of demographics, changing consumer expectations, rising interest rates, and tightening lending standards have yet to be felt.  As they are, it’s all but assured that price/income ratios will come under pressure.

But what of the notion that house prices could stagnate while incomes catch up?  As I’ve already discussed, asset prices that stray markedly from their long-term measures of fundamental value seldom correct by going sideways.  In fact, so rare is this occurrence that it has led some academics who study real estate asset bubbles to assert that, ”soft landing, i.e., an upgoing phase followed by a plateau, has rarely (if ever) been observed.”

Finally, I do have to point out a massive error in calculations contained in the report.  The report asserts that price/income ratios can return to their long-term trend line (4.1) from their current level (5.7) by the middle of the decade if house prices stagnate while incomes grow at a 2% clip.  This is ridiculous.

The average home sold on the MLS in Canada in January 2011 was just shy of $350,000.   At the current price/income ratio of 5.7, it assumes a salary of $61,400.  To return to the long-term average of 4.1, income would have to equal $85,300.  Yet to move from 61K to 85K at a 2% annual increase would take 17 years!

We can give the authors a very gracious benefit of doubt and assume they meant real (inflation adjusted) wage increases of 2%, though if they did mean this they would very likely have said so.  Assuming that the Bank of Canada is able to meet their approximate 2% annual inflation target, it implies a nominal wage increase of 4% annually.  Yet even at that clip it would take nearly 9 years for the ratio to normalize.  The fact that this basic math somehow slipped past the authors casts the whole report in a hazy light.

While the headline of this section of the report suggests a low probability of a sharp house price correction, the support for this assertion is largely absent.  Indeed there are glaring holes in their logic and equally glaring errors in their calculations to the point that it calls the credibility of the report into question.

Cheers

Ben