Canadian GDP — More Treat than Trick

• Canada’s economy continues to expand…

• ...as negativity remains too high...

• ...but the BoC’s forecasts will need some help…

• ...which keeps the upsizing debate still highly data dependent

• Markets shook off Canadian and US data

• US data generally suggests caution by the Fed…

• ...ahead of tomorrow’s payrolls

• Canadian GDP, m/m % change, SA, August
• Actual: 0.0
• Scotia: -0.1
• Consensus: 0.0
• Prior: 0.1 (revised down from 0.2)
• September GDP ‘flash’ guidance: +0.3

Canada’s economy ended Q3 on a firmer than expected note but the Bank of Canada’s growth forecasts for 2024H2 will still need a fair amount of help to come true. That’s by no means impossible, but the debate over continued easing with upsizing risk will remain focused on the heavy flow of data between now and the December 11^th BoC decision.

Before turning to the numbers, I often like to point to chart 1. You’d think that with a lot of the pessimism out there that Canada’s economy was tanking. Rubbish. Hogwash. It could be better, but about five percentage points worth of rate hikes has not caused recession, has not caused GDP to crater, and has kept a mild expansion on track. Growth is underperforming the supply side’s expansion which is partly behind disinflation and rate cuts. It could be better. It is not crisis mode and I continue to believe that the US election and external risks are set against an otherwise constructive backdrop for the household sector’s expansion over 2025–26.

Key is that GDP ‘flash’ guidance from Statistics Canada points to growth of 0.3% m/m that month. The August reading of 0% m/m SA matched earlier Statcan guidance that the month was “essentially unchanged” which was a tick stronger than I had forecast. The only reason August was a little stronger than I had forecast was unfortunately because the prior month’s growth rate of 0.2% m/m in July was revised down a tick, thus offering a lower jumping off point for August GDP; hence 0% instead of –0.1.

What this all translates into is tracking Q3 GDP growth of about 1% on the combination of solid September growth, no growth in August, and weaker than previously understood growth in July, but with strong caveats (chart 2). The BoC's 1.5% Q3 GDP growth forecast uses the different concept of expenditure-based GDP (versus the monthlies that are production-based). The differences can be through inventory and net trade contributions to growth that the expenditure side captures more readily. Expenditure-based GDP basically asks how higher income was achieved, including inventory addition/depletion and import leakage changes. In other words, we can't say definitely with incomplete data on trade and inventories that the BoC is too high.

Canadian Q4 GDP has 0.8% q/q SAAR growth baked-in. This concept estimates the momentum effect as the math hands-off to Q4 before we get any actual Q4 data. That suggests a weak running head start into Q4 GDP, but it could easily have been worse absent the September 'flash' upside.

Chart 3 shows the industry break down of what drove August GDP. Gains in about five sectors offset losses in about three others with the rest little changed. 

Statcan doesn’t provide a breakdown of the ‘flash’ estimate for September, but their verbal guidance points to “increases in finance and insurance, construction, and retail trade were partially offset by decreases in mining, quarrying, and oil and gas extraction.” No doubt some of the bounce was also off the prior month’s wildfires and strikes.

And so where's the BoC going from here on the path to the December 11th meeting? It's still highly data dependent, but we can very tentatively say that the BoC needs a bit of help from the data to get its 2% q/q SAAR forecast for Q4 and 1.5% for Q3. If they don't get that help, then they'll have more slack than they had expected into year-end which raise further concern about undershooting the 2% inflation target and could keep alive upsizing risk.

That said, there are two rounds of labour readings, one CPI report, more GDP data for Q3 and the preliminary October reading, the US election, the Fed etc etc all ahead of the December 11th BoC decision. So, hold off on the tin foil hats, it's going to come down to patient tracking of data and developments.

The other part of all of this is that we now have a benchmark against which to intensify tracking of September GDP growth. Chart 4 shows that the ‘flash’ guidance performs reasonably well over time with surprises in both directions and an average revision to the flash of +0.1% m/m since they started provided flashes and nothing if we exclude the first and most volatile couple of the years by starting the tracking in 2022.

US DATA

A round of US reports was constructive to growth tracking but underlying inflation tracking sends a cautious signal to the FOMC.

Here’s the run down before getting to Fed implications.

• Key was that core PCE inflation was up by 0.25% which barely rounded up to 0.3% on screens, while the prior month was revised up to 0.2% rounding up from 0.16% m/m SA. That’s still much better than the kind of numbers we were getting in the past, but the mild trend of late is indicating we should be very careful in declaring victory over inflation (chart 5). A key driver of some of this recent upward trend has been core service prices (chart 6).

• Furthermore, the Employment Cost Index was up by 0.8% q/q SA nonannualized (0.9% consensus, 1.0% Scotia) which, while a smidge lighter than expected, is not light. The trend is pointed decidedly lower, but the annualized pace of 3.2% q/q SAAR remains hotter than the FOMC’s 2% inflation target (chart 7).

• Alongside the prices and costs data lies the evidence on growth. Consumer spending was up 0.5% m/m SA nonannualized in September (0.4% consensus and Scotia) and was revised up to 0.3% m/m SA in August from 0.2%. We now know the monthly distribution of yesterday’s reported 3.7% q/q SAAR consumption gain in Q3 and what it helps with is tracking hand-off math into Q4. With no data yet for Q4, US real consumer spending already has a gain of 1.2 ppts q/q SAAR baked in. Bring on the holiday shopping season!
• Lastly, weekly initial jobless claims fell back to 216k last week from 228k the prior week. This signal of the labour market remains strong.

There was not much market reaction to the US data releases, basically because they just largely reaffirm quarter point cut expectations for the November 7th decision next Thursday (instead of Wednesday because of the US election on Tuesday).

Dual mandate variables keep upsizing at bay. September’s nonfarm payrolls were firm (+254k), even though distorted by a whoppingly high SA factor. September core PCE accelerated with upward revisions.

And yet Powell will probably repeat that he sees signs of rebalancing in aggregate demand and aggregate supply that merit a less restrictive stance and cautious forward guidance.

I don't think tomorrow's payrolls are likely to matter a whole lot depending in part on the composition and drivers.

1. The Fed would look through a transitory hit to job growth from hurricanes and strikes that restrain hiring and job acceptance. History shows the next 1–2 months of payrolls rebound firmly. And the Fed always explains that it looks through such disruptions as monpol can't do anything about storms that just happened and the negatives cancel out in the rebuilding efforts.

2. But if payrolls are really strong like ADP was and despite storms and strikes, well then, that's another matter. Watch the seasonals tomorrow as well since October's SA factor is also likely to be high, and so there are a lot of conflicting drivers of the payrolls call. SA factors are suffering from a recency bias in that they are estimated in a way that is skewed to the most recent 3 years which means early year job growth tends to be overstated, summer understated, and then overstated again now because of the timing of the pandemic's disruptions.

My estimate for tomorrow's payrolls was 95k. I have a little less confidence it will be that low after ADP that signalled a 10% chance nonfarm comes in around 100k or less, but not much less. I think hours worked are going to bear the brunt of the October shocks which should be a negative for the next personal income report a month from now.