The prospects for stronger economic growth in the second-quarter remain in flux, but the latest numbers move the dial a bit deeper into positive territory. The change is far from decisive, but it’s enough to suggest that momentum may be building for a stronger run of economic reports in the weeks ahead.
One reason is this week’s upbeat news that the US international trade gap fell substantailly in April, thanks largely to a hefty decline in imports. “Trade may add to growth in the second quarter, but afterward, for the next two years, we expect it to be a modest drag on growth,” advised Patrick Newport, an economist at IHS Global Insight.
As for the near-term outlook, the narrowing trade gap provides a boost to GDP estimates for the second quarter, including the Atlanta Fed’s revised GDPNow projection for Q2. The current forecast ticked up to a 1.1% growth rate for the second quarter vs. the previous 0.8% estimate. That’s still a sluggish pace, but it’s up from Q1’s 0.7% slide. More importantly, the revisions show a mildly positive trend these days, raising the possibility that the incoming numbers for the rest of the quarter will deliver more upbeat news.
Meanwhile, yesterday’s weekly update on jobless claims suggest that the labor market is still poised for growth. New filings for unemployment benefits fell 8,000 to a seasonally adjusted 276,000 for the week through May 30, which is close to a 15-year low. The decline also marks the 13th straight week of claims below the 300,000 mark, which is considered an encouraging signal for anticipating continued growth in payrolls.
Let’s use the claims data to estimate recession risk via a probit model. Unsurprisingly, the result shows that the probability is low that the US economy is contracting. The basic procedure: transform the weekly claims figures into monthly data via the four-week average, calculate rolling 12-month changes, and run those numbers through a probit regression against the NBER’s historical business cycle dates. By that measure, the current recession risk estimate is virtually nil, at less than 1%.