US Q1 GDP surprises to the downside

Downward revision to retail sales responsible for GDP miss

While the main drag came from a drop in inventories (which subtracted a whopping 2.7pp from growth), a large downward revision to retail sales also meant consumption was not as strong as expected. Indeed, the Atlanta Fed’s GDP Now tracker had already suggested a big miss the day prior to the release of the GDP report, due to the revision to retail sales. Despite that downward revision, consumption still grew very strongly in the first quarter, by 3.7% annualised, with a 16.9% surge in durable goods consumption responsible for the strength (services consumption growth was much more moderate at 2.3%).

The exceptional strength in goods consumption has been a surprise in the first quarter, given that for much of last year goods consumption had been on a cooling trend. Still, looking at more recent high frequency data does suggest goods consumption has since resumed its cooling trend, with for instance Redbook weekly department store sales slowing sharply of late. At the same time, there has been a tendency for repeated downward revisions to consumption data in the post-pandemic period, likely reflecting difficulty in measuring price effects in the current high inflation environment. As such, it would not be a surprise if the Q1 strength in consumption is further revised away in future GDP estimates.

Mild recession still on the cards, but buffers could delay the timing

As we describe in our April Global Monthly, the significant cash buffers from the pandemic period appear to be delaying the hit to demand from high interest rates. However, the persistence of inflation means that even if the onset of recession is delayed, it cannot be avoided: one way or another, the Fed needs to generate some economic weakness – including a meaningful rise in the unemployment rate – to be sure that inflation falls back to target. We see emergent signs that the recent banking turmoil is having a significant tightening effect on lending conditions, and this could now be the trigger for a more meaningful slowdown in activity, which our base case sees starting in Q2. Should that fail to materialise – for instance, if tighter lending standards prove short-lived – we see a risk that the Fed continues hiking into the summer months following the expected 25bp hike next week (currently, we expect the May hike to be the last of this cycle). (Bill Diviney)