When did Canada's recession end? A WCI poll

Canada doesn't have a counterpart to the NBER's Business Cycle Dating Committee, so we don't have a reference set of dates for previous business cycles. I'm pretty sure that most analysts would agree that the most recent expansion hit its peak in October of 2008, but picking a trough looks less obvious.

My understanding of the way the NBER committee works is that they look at a broad range of indicators and try to come up with a consensus. I've chosen 14 series and identified the months in which they hit their minimum values in 2009. If you have any preferred candidates I missed, let me know in the comments and I'll try to add them as well. I'm going to list them by month. Some series - such as the CLI - have troughs that lasted several months, and they will show up more than once.

February: Residential building permits, Nonresidential building permits

March: Sales of durable goods, TSX

April: Composite Leading Indicator, Sales of durable goods, Housing starts, Average work week - manufacturing, Hours worked

May: Composite Leading Indicator, GDP, Durable goods orders, Shipment-to-inventory ratio, Exports

June: Composite Leading Indicator, Retail trade, Imports

July: Employment, Retail trade

My vote would be May, but my preference isn't so strong as to reject a consensus for another month. Please enter your votes, data requests and thoughts in the comments.

Comments

[Nick Rowe]

July, because of employment?? But in an economy with a growing population, and growing labour force, I would maybe say a bit later than July, because it's only when I see total hours worked rising faster than (say) 2% on an annualised basis, that I would say we've passed bottom. And employment fell again after recovering.

Unemployment rate? It peaked, fell, then rose again.

Hours worked.

[Stephen Gordon]

Hours worked bottomed out in April; it's added to the list. Unemployment is still rising, and will until employment growth outpaces the rate of growth of the labour force.

[Neil]

If we're going to repetitively quote the definition of a recession as being 2 consecutive quarters of GDP decline, then why do we insist on muddying the waters with all these other indicators. Either the status of "recession" is linked to GDP, or not.

And if there really is so much grey area, then why do we insist on pinning an exact date to it?

[Justin Donelle]

I say it isn't over and it will only be over when we stop manipulating the money supply :)

[Nick Rowe]

Well, based on today's employment numbers, I'm now feeling a bit more optimistic it's over.

Neil: "recession" is a fuzzy concept. What one views as the most useful definition depends on one's theoretical perspective. I don't find the 2 quarters GDP decline an especially useful definition.

Justin: Then it will never be over!

[Mark]

I created an interesting and fun probit model for an Ontario recovery back in June. Tried to use variables that weren't subject to revision or data measurement error (employment is exception): the six-month per cent change in the TSX, year-over-year per cent change in LFS employment and housing starts, the six-month difference in the Scotia corporate bond index yield less long-term government spread and the one-year difference in real interest rates. It's similar to a model Macroadvisors uses for recession probabilities.

I got a 52% prob of trough in May and 98% trough in June. Not scientific, but I had fun :)

Used a similar equation for peak probability in July 2008 when recession was a big unknown in Ontario, got 21%. Updated again in October and got 100%.

[Stephen Gordon]

Cool!

[MattM]

Great news out of both the Canadian and US employment reports.

BLS (US) payroll -11k (relative to -100+k last month), unemployment from 10.2% to 10%, hours worked from 33 to 33.2

LFS (Canada) +79k, unemployment from 8.6% to 8.5%

I believe that hours worked and the payroll survey of employment comes out significantly later in Canada than in the US (not sure why)

[Stephen Gordon]

Hours worked is part of this release, but not part of the press release. I'm waiting for StatsCan to update their database, and then I'll post on it.

[Mark]

Y/Y hours worked were down, M/M they were up, but the data is not SA.

[MattM]

Errr...yeah, I would expect the hours to be down Y/Y. In the meantime we've had a recession.

[Justin Donelle]

Nick: Well, the boom will start back soon enough, just keep printing our way there. Once the money gets flowing a little more throughout the economy, then things will get back to normal. As long as people get their confidence back in the financial sector, then everything will be fine for another few years.

[MattM]

Yes, because booms and busts never happened when money was "real"....

[Justin Donelle]

Of course, they happened when bankers manipulated bank notes or when there was too much of an influx of gold circulating. In a world economy it would be difficult for this to happen, if bankers manipulated notes it would. That's why you force 100% reserve instead of fractional-reserve banking.

[Andrew F]

I think Nick's point is that we are always manipulating the currency, by virtue of the fact we set interest rates.

I think it's worth pointing out that 100% reserve would cause more problems than it would solve.

[Justin Donelle]

What problems? Having people suffer from bad decisions and learning, instead of letting the central bank back bad decisions and banks in the name of stabilizing the economy? The constant inflation makes most people worst off. It's difficult enough to retire when you have the bank devaluating your currency constant basis, when most of the increase in money doesn't make its way to everyone on an equal basis. The greatest inequality that will ever exist is the central bank, a negative inequality.

I agree with your argument of interest rates. If only the market set them, then we would have stability.

[westslope]

Mark:

I read that you have a yield curve component in your probit model. Is the model designed to date recession beginnings and endings or to forecast recessions several months out?

[Nick Rowe]

Justin: if the government is in the money business, it needs to manipulate/control/set something. Either the quantity of money, or the rate of interest at which it lends money, or the rate of inflation, or the exchange rate, or something. You can't just tell it to do nothing. And 100% reserves constrain commercial banks in how much money they create, but they don't constrain the government/central bank in how much money it creates.

The only consistent position for a libertarian to take on this issue is to argue that the government should get out of the money business altogether. Which some (e.g Hayek) do advocate.

[Justin Donelle]

Nick: That's what I was advocating, "The greatest inequality that will ever exist is the central bank." As well as when I argued that people should live with bad decisions and learn, meaning my argument for the abandonment of the central bank, followed by letting people learn and to demand honest 100% reserves, if a bank decides to inflate by dishonest fractional-reserve banking, it will fail and the citizens will demand strong accountability later on or those banks will be out of business. Sure, some people will lose their money and we'll have a minor recession due to the market's adjustment and the shedding of those bank notes, but things will recover faster and teach people to watch out for bank policies.

I also acknowledge that 100% reserves will constrain commercial banks and not the central bank, I never advocated restricting commercial policy and not the Central bank. The only way is to end the monopoly on money, as people like Hayek and Rothbard advocated. My comment on market determined interest rates is the subjective rate that individuals apply if they wish to lend money to another individual, whether a bank loan or a personal loan.

I'm not Tyler Cowen, who says he's a libertarian and then makes a 180 degree flip and then says right-libertarians need to live with big government and support bailout packages. That one always puzzled me.

[Stephen Gordon]

Thanks, Justin, for hijacking the comments. Greatly appreciated.

[Justin Donelle]

Sorry, Stephen.

I'll just write a blog about money and not fill yours will money issues instead of Canada's recession:)

[Pete Tucker]

The word here on the street in Alberta is that the recession is over, and welcome to the depression.
The housing crisis is similar to the states however I believe there is a delay factor following the crisis in the US. Calgary Realestate resembles the Las Vegas market, with much mortgage fraud, and bankrupt, vacant and abandoned condo projects. Most people here are over financed and way over the TDS ratios. Its just a matter of time before $#!~ hits the fan. Many builders here are panicking and having fire sales at greatly reduced prices but the market remains quite stagnant. A good percentage of construction worker have all been laid off.

[Nick Rowe]

Pete: but Teranet is showing a slight uptick in Calgary house prices http://www.housepriceindex.ca/Default.aspx , though it has been hard hit.

(What's a TDS ratio?)

[Nick Rowe]

Speaking of Teranet, its composite 6 index of house prices hit its minimum value in June.

[Pete Tucker]

TDS = Total Debt Service
What Does Total Debt Service Ratio - TDS Mean?
A debt service measure that financial lenders use as a rule of thumb to give a preliminary assessment of whether a potential borrower is already in too much debt. More specifically, this ratio shows the proportion of gross income that is already spent on housing-related and other similar payments.

[Pete Tucker]

Thanks for the note Nick. The methodology that http://www.housepriceindex.ca/Default.aspx uses is sales pair count or repeat sale methodology, which is selective in the data sampling and does not take into consideration many critical factors, including new home sales, or capital input for renovations. This can give a skewed, distorted and narrow scope of the total market that is being analyzed.

[Jim Sentance]

I've always felt that recessions should be gauged not by GDP but by GDP per capita. So once GDP starts growing faster than our population, its over.

[Jacques Simon]

Right now, the world economy is worth less than the value implied by the market value of its obligations. Real incomes haven’t grown in years. Manufacturing and, increasingly, service jobs are moving overseas. We are in the midst of a deflationary trend that is temporarily being masked by inventory restocking.

We won't see the recovery until 2012.

[Bailey]

I'll vote for June.

[David Heigham]

Let me predict that two years hence the cyclical turning point will turn out in the statistical wash-up to have been in the second quarter of 2009, and propbably late in the second quarter. The point of positive inflexion - when the trend of actual growth exceeds the trend of potential growth - will be being voted on about then.

[G Voss]

While we don't have a Canadian institution providing business cycle dates, the Economic Cycle Research Institute does provide dating for a number of countries, including Canada, using the NBER methodology. The chronologies can be found at http://www.businesscycle.com/resources/

My apologies if this has already been noted elsewhere.

[Stephen Gordon]

Huh. Interesting. They have the peak at January of 2008, which seems very odd to me.