When the other shoe falls: the macro impact of a housing bust

The Canadian media didn’t pay much attention to this economic crisis before the stock markets plunged last fall, and that has continued to be a major focus. But this focus on the stockmarket and the mistaken perception that this crisis is all the fault of the U.S., is leading to the overly optimistic forecasts of recovery.  There’s been very little attention paid to the economic impact of a housing market bust in Canada , perhaps in the hopes it won’t happen, but the impacts are likely to be severe.

There is no doubt that the stock market drop has been dramatic: the TSX has lost about 50% of its value since last June and continues to stumble lower. With market cap of the TSX at $2 trillion a year ago, this drop is equivalent to a loss of about one trillion dollars. This more than the value of all wages and salaries received by everybody in Canada in a year, and equal to about $30,000 for every single Canadian. At this magnitude, the stock market crash should have a much more devastating effect in Canada, but it hasn’t yet

Why not?

According to Statistics Canada’s Survey of  Financial Security, less than 25% of Canadian families directly own either stocks or mutual funds outside of their pension funds, RRSPs or RRIFs, and much of this ownership is concentrated among the wealthiest.

That is a major reason why the Bank of Canada and others figure that changes in stock market values don’t tend to have a big impact on consumer spending: the “wealth effect” from changes in equity values in Canada is estimated to be small and statistically insignificant.

While only a minority of Canadian families own equities, more than 70% have assets in the form of RRSPs, RRIFs or employer pension plans.  More normal economic cycles usually wouldn’t have a big impact on consumer spending and the economy through these pension holdings either. Pension assets are generally held for the long term, pension funds are required to do a valuation every three years and are allowed five years to fund these deficiencies: usually long enough for stock markets to recover.

But, as Andrew Jackson’s recent excellent post has pointed out, even if equity markets recover by 50% over the next five years (growing at an ambitious rate of 8.5% per year), this will still leave $125 billion hole to restore employer pension plans over the next five years.

That amount could soak up all, or a major part of all expected nominal wage increases during this period unless other measures are taken and drag GDP growth down by about 1.25% a year.

That’s one shoe that has dropped in Canada.

The other shoe that is now dropping in Canada is housing prices.

Housing busts tend to have a bigger impact on the economy than stock market busts because housing wealth is shared much more broadly than stock market wealth. Over 60% of Canadians families owned their own home in 2005, including over 80% in the middle income quintile and up. And because housing values are also seen as more stable than stock prices, housing tends to have a much stronger wealth effect. The study by Lise Pichette at the Bank of Canada estimated a housing wealth effect of 5.7%: that is, every dollar of additional housing wealth increases consumer spending by 5.7 cents.  Other estimates of the housing wealth effect are even higher.  The Congressional Budget Office and federal reserve economists have studied this, as did Alan Greenspan.

This is why the IMF found back in 2003, that real estate busts tend to be twice as deep and last twice as long as stick market busts .

Dean Baker at the Centre for Economic Policy and Research has done some very good analysis of the macro-economic impact of the housing and stock market bust.

When I did some calculation of the impact of the housing boom and bust in Canada a year and a half ago, I estimated that the housing boom had created about $500 billion of extra “wealth” in Canada, over and above what the value of the stock of houses would have been if they had increased at the rate of inflation.

If this additional housing wealth were to evaporate from a drop in house prices down to an inflationary trend then, using the Bank of Canada’s wealth effect of 5.7%, consumer spending would be expected to drop by about $25 billion a year. That is equivalent to lower GDP growth of about 1.5% per year.

This doesn’t take account of lower level of activity in the construction industry because of lower housing starts. If we add these and the stock market effect through pensions, then the total drag on growth could easily surpass 3%.

My calculations from a year ago were before house prices started to decline in Canada and included a rough adjustment for quality because at that time we didn’t have a quality adjusted housing price index, similar to the Case-Shiller index in the U.S.

Now, very fortunately, the National Bank has teamed up with Teranet to produce a repeat-sales match house price index for Canada.

An analysis of this house price index shows that house prices have increased almost as much in Canada as they did in the U.S. Case Shiller, (increase of 96% from Jan 2000 to peak in Canada compared to an increase of 106% in the U.S.) but that we are peaking, and now declining, with about a two year lag compared to the U.S.

The index was developed partly in order to develop a forward market to hedge against housing price declines. Initial betting on the forward market showed expectations of a 20% drop in housing prices, and seven years for prices to recover.

Interestingly enough, this is close to the average 30% house price declines associated with housing busts reported by the IMF in 2003. (I’ve been expecting a house price correction in Canada of about 20-30%.).Even some bank economists are now finally acknowledging that house prices will drop significantly, although I think they are still underestimating how much they will fall.

Canadians had about $2 trillion in housing wealth, based on numbers from Statscan’s financial security survey. A 20% drop in this value translates to a decline in housing wealth of $400 billion. The wealth effect on that would work out to lower consumer spending of about $20 billion a year.

Now there is the chance that the housing wealth effect is not symmetrical.

I could only see two studies that have estimated for that, but they have come up with different results: in one case it is bigger on the downside, in the other case, the wealth effect from house price declines was much smaller. Even where the wealth effect was smaller, the authors estimated that the combined effect of this housing downturn (including the construction industry income and investment effects and financial market effects of the mortgage crisis) would reduce U.S. GNP growth by 3% each year until recovery.

The federal budget took steps to prevent a decline in the construction industry with its various temporary tax incentives and infrastructure spending. These may forestall some of the decline in the industry, but they won’t prevent an ultimate decline in housing prices.  And, with a slowdown in starts, eventually construction industry employment and GDP will drop closer to or below its longer term average of about 5% of the economy.

The bottom line is that the housing bust will also lead to a substantial drag on the economy going ahead. I expect that the combined effect of housing wealth effect and the slowdown in the housing/construction industry will reduce annual growth by 2%-3% a year in Canada for a number of years.

If we add in the stock market wealth effect through pensions as calculated by Andrew, then forecasts by bank economists and others of a short and sharp recession followed by a strong recovery in Canada are very hard to believe, even if we are optimistic about stimulus spending, the punch from lower interestrates and remedies to the financial crisis and credit crunch.

Update:

I found it interesting that the Bank of Canada has now just acknowledged that economic growth will be slower than they had forecasted a little more than a month ago as a result of “wealth effects on domestic demand” among other factors.  

The problem is that, unless there is whopping rebound of equity and housing prices, the impact of these wealth effects will be drawn out over a number of years.  As a result, we won’t see a strong and sustained recovery until then.   The “output gap” in the Bank of Canada’s economic models may close before then, but only because they will redraw the line of potential growth lower so the output gap is also smaller.