• Soft trend retail sales figures...
  • ...have been due to the rotation toward services
 
  • Canadian retail sales, m/m % value change, SA, June:
  • Actual: 0.1
  • Scotia: 0.0
  • Consensus: 0.0
  • Prior: 0.1 (revised from 0.2)
  • July preliminary guidance: 0.4

Canadian retail sales put in a soft performance that drove a decline in shorter-term Government of Canada bond yields and a depreciation in the Canadian dollar on a morning when the bias was already headed in both of those directions given overnight developments.

I think there are key points that counsel caution in terms of taking the softness as a literal reflection of what’s happening to the Canadian consumer. Noting these points will be followed by recapping the key observations in the data.

RETAIL OFFERS A DISTORTED IMPRESSION OF CONSUMER STRENGTHS

Chart 1 indicates that retail sales volumes have been treading water for a protracted period of time. 

Chart 1: Canadian Retail Sales Levels

This has occurred during a time when total consumption has posted solid trend growth. The gap is explained by the rotation toward services spending that is not captured in retail sales one bit in Canada given how Statcan defines them.

Chart 2 effectively illustrates the point. Retail sales volumes have been listless for a long time now, but the volume of total consumer spending has been rising at a decent clip throughout the past couple of years. The gap is largely explained by the rotation toward services spending.

Chart 2: Retail Sales Understate Consumption in Canada

This is vividly shown in chart 3. Folks found something else to do with their time and money as they distanced themselves from the pandemic. 

Chart 3: Household Consumption Expenditure Shifting Towards Services

A caution is that we're mixing apples and oranges in terms of concepts here so just treat the point as a general one. We are comparing retail sales volumes of goods sold versus overall consumption value-added in GDP accounting and they are not directly comparable, but they both control for price effects and I'm comfortable with the general point being made here.

This wedge between the two types of evidence is because Canadian retail sales exclude restaurants, bars, hotels, airlines, other modes of travel, ticket purchases (concerts, sports etc) and financial transactions and basically anything that is service oriented. They also exclude online purchases through foreign e-commerce sites (eg. Amazon.com, but Amazon.ca is included). In essence, retail sales don’t capture any of the great rotation toward services spending. That’s a uniquely Canadian thing. The US, by contrast, seeks to include at least some of this type of spending by including, for example, spending at eating and drinking establishments.

Statcan produces separate figures for the dollar value of spending at food service providers and drinking places (mostly bars and restaurants in plain English...) but the data lags and only goes up to May with June due in the coming days and July data a long way off. Chart 4 shows how this type of spending has been strongly advancing but again, this is not included in retail.

Chart 4: Spending on Food Services & Bars Increasing

We see this macroeconomic trend being reinforced by earnings reports as companies serving the nondurable consumption sector have underperformed companies like airlines. Within these sectors there is also wide latitude for varying performance by individual companies.

DETAILS—NO BREADTH BEYOND AUTOS

Nominal sales have been trending around little change of late. They were revised a tick lower to a 0.1% m/m gain in May. Then June’s advance guidance that sales were roughly unchanged was on the mark with sales up only 0.1% m/m. Advance guidance for July points to a 0.4% lift.

The volume of sales fell by 0.2% m/m in June. We don’t get anything from Statcan about July other than nominal sales guidance, but the 0.4% gain probably means that volumes were little changed last month in light of CPI.

That means that retail sales volumes fell 3.2% q/q at a seasonally adjusted and annualized rate in Q2 after a 5.2% Q1 gain with -0.7% baked in for Q3 based on Q2 and assuming July's volumes are little changed. Chart 5.

Chart 5: Canadian Real Retail Sales Growth

There was fairly high breadth to June's softness by sector in both dollar value (chart 6) and volume terms (chart 7). Vehicles and parts were up 2.5% m/m SA led by new auto sales and 'other' which is ATVs etc. Sales ex-autos fell 0.8% m/m. Most of the rest of the report was soft and so this is basically an autos report. What is being spent on autos and services leaves little else to consider but that does not for one second mean that consumers are not spending.

Chart 6: Weighted Retailer Contributions to June Nominal Retail Sales; Chart 7: Weighted Retailer Contributions to June Real Retail Sales