In another thread (below) I forecasted consumer spending (PCE, personal consumption expenditures) from retail sales (ARS).
ARS is released mid-month and PCE (which I believe will be closer to the actual input for consumer spending in GDP) is not released until month end.
So the idea is to use retail sales to get a slightly earlier (2 weeks, lol) read on consumer spending in GDP.
Retail sales (for month of May) will be released Tuesday, June 17.
Economists’ consensus forecast is -0.7% for May month over month, with the decline due to a slowdown in auto sales relative to April (as best I can tell).
May Advanced Retail Sales (ARS) were -0.9% versus April, weaker than expected (-0.7%).
Using the regression model noted above, forecast PCE (consumer spending used in GDP) projects to a forecast of about 0.2% PCE (nominal), with range of about 0.0% to +0.4%.
CPI inflation for May was 0.1%. So adjusting the above forecast for PCE for inflation gives about +0.1% (-0.1% to +0.3%) real PCE for May.
Summarizing for Q2, real PCE is
+0.1% Apr actual
+0.1% (-0.1% to +0.3%) May forecast
+0.2% (0.0% to +0.4%) quarter-to-date forecast
In April, the actual came in on the low side relative to the above regression model. Will be interesting to see if that repeats or not for May.
The consensus forecast for May PCE is +0.1% nominal, 0.0% real if adjusted for CPI like above. This would give 0.1% qtd, or 0.6% annualized.
The above points to weak consumer spending so far for Q2 2025, about 1.2% annualized (only +0.6% from consensus), which would have a corresponding weak contribution to Q2 GDP.
GDPNow shows Q2 PCE of 1.9%, which seems a tad high relative to the above forecast and consensus.
The shocker is going to be in the construction industry. The firms can not plan ahead. They are not adding new projects. That means a lot of different poor economic results including laid off construction workers.
For May, PCE was -0.1% and real PCE was -0.3%, so lower than the simple regression model predicted (+0.2%), just like for April, and lower than economists’ consensus forecast (+0.1%).
Through May this year, using latest numbers from FRED, monthly consumer spending (real PCE) this year has been
Jan -0.56%
Feb -0.10%
Mar +0.68%
Apr +0.09%
May -0.28%
So consumer spending’s contribution to GDP growth has been near zero to slightly negative this year.
The consumer economy is not growing.
Crude ballpark numbers for other components of GDP:
inventories ~0.0% qtd
new housing sales ~0.0% qtd
durable goods +7% qtd
trade deficit -33% qtd
For Q2 2025 through May, it looks like wild swings in the trade deficit will be a large data overlay on top of flat consumer spending and weak investment.
With today’s data, GDPnow ticked down to 2.9% GDP growth, dominated by a decrease in the trade deficit (which adds to GDP growth) but has PCE at +1.7%, which seems too high given consumer spending numbers we have seen for Apr and May.
June Advanced Retail Sales (ARS) were +0.6% versus May, much stronger than expected (+0.1%).
Using the regression model from upthread, forecast PCE (consumer spending used in GDP) projects to a forecast of about +0.6% PCE (nominal).
June CPI inflation was +0.3%.
This would give an estimate of June real PCE of +0.3% versus May (0.6% - 0.3%).
Then through June this year, the above forecast for June together with actual numbers for earlier months, monthly consumer spending (real PCE) this year has been
Jan -0.56%
Feb -0.10%
Mar +0.68%
Apr +0.09%
May -0.28%
Jun +0.3% (est)
The above continues to suggest near zero growth, maybe weakly positive, in Q2 and YTD for the consumer spending economy, about +0.4% and +0.2% annualized, respectively.
GDPnow estimates Q2 growth of 2.4% with real PCE of 1.5% (annualized), which seems high compared to the above.
If we think real PCE is really about +0.4% in Q2 (vs +1.5%), then GDPnow’s contribution to GDP (+1.0%) is too high and maybe should be closer to about +0.1 or +0.2% which would put GDP near 1.6% when using GDPnow’s estimates for all other components of GDP.
Actual June real PCE and Q2 GDP numbers are released at end of July, so we can compare above to actuals then.
California’s unemployment rate rises to highest in the country as layoffs mount California’s unemployment rate rose slightly, by 0.1 percentage point, to 5.4% in June, tying Nevada for the highest rate in the U.S., according to new federal data released Friday. The state lost a net 6,100 jobs, including 9,900 layoffs in business and professional services. Health care and government saw job gains, but other sectors all shrank.
San Francisco’s unemployment rate rose 0.7 percentage points to 4.2% in June…“The change is rooted in the ongoing slowdown in the tech sector, the high costs of living, and local and state employment policies that make hiring more costly than other parts of the state,” said Bernick, former director of the state Employment Development Department.
Widespread tech layoffs have continued, with companies like Scale AI, Microsoft and Intel cutting staff. “Although the nation’s tariff chaos is certainly creating broad uncertainty, these losses aren’t about tariffs,” said Christopher Thornberg, founding partner of Beacon Economics, in a statement. “The declines were in Manufacturing, Information, and Professional Services and centered in the Bay Area, meaning they are a reflection of the ongoing contraction in the tech space, which in turn, is being driven by AI.”
Which is only an unemployment rate of 4.2% which you noted. Which is much less than the whole state at large. So I think your hypothesis that it’s the tech space problem is incorrect, some might say it is Hyperbolic.
It’s not my hypothesis and does not apply to the state as a whole. Please just read the San Francisco Chronicle article that was linked by Leap. It is of interest, however, to those who are concerned about the effects of AI.
Initial claims for state unemployment benefits dropped 4,000 to a seasonally adjusted 217,000 for the week ended July 19, the Labor Department said on Thursday. Economists polled by Reuters had forecast 226,000 claims for the latest week…
Though job growth has slowed from last year, the labor market remains stable and has provided scope for the Federal Reserve to delay resuming cutting interest rates while watching for potential inflation from the import duties…
The number of people receiving benefits after an initial week of aid, a proxy for hiring, increased 4,000 to a seasonally adjusted 1.955 million during the week ending July 12…
A decline in labor supply amid reduced immigrant flows means the economy now only needs to create roughly 100,000 or less per month to keep up with growth in the working age population.
As suggested, from time to time, with many thousands being deported, and many thousands more gone to ground to avoid being grabbed, is a seller’s market for straight, white, Christian, male, labor being created?
TIG is completely naive when it comes to economics and foreign policy.
The bluster and fire and brimstone gives way to failure. He was very surprised by the reception Putin gave him. He will be surprised by where the tariffs lead.
Rationale for badgering Powell, he thinks lower interest rates will allow first time buyers into the housing market. He can not think far enough ahead to see prices rising even faster shutting them out. Naive.
