"Statistics are used much like a drunk uses a lamppost: for support, not illumination."
– Vin Scully (among others)
The broad range of data suggests that the U.S. economy slowed in the first quarter. It’s more likely that this is merely “a slow patch” than the start of a more substantial downturn (not gonna use the r-word). The mixed nature of the economic data allows one to make any particular argument one wants and the noise is likely to add to market volatility in the near term.
The Atlanta Fed’s GDP-Now model has received increased attention from the markets. Most economists use a similar methodology in estimating GDP growth. One simply forecasts the components and adds them up (it’s actually a little trickier than it sounds since you can’t add inflation-adjusted figures). Economists will differ in how they forecast the components. The problem with this framework is that your GDP forecast will shift around as each economic release arrives. In late March and into early April, the components have come in generally weaker than expected. The Atlanta Fed’s model is now estimating a 0.1% annual rate. I’m currently at +0.5% (±1.0%), but that will change as more information becomes available.
Note that in the last six years, first quarter GDP growth has generally been a lot lower than in the rest of the year. This raises the issue of whether something may be wrong with the seasonal adjustment. Indeed, the Great Recession may have permanently altered seasonal patterns in economic activity. We can’t say for sure. The sample size here is too small. If the seasonal patterns have changed, the government’s seasonal adjustment will pick that up over time (and eventually, we may see first quarter GDP figures revised higher).
First quarter weakness is much less apparent in the government’s measure of underlying domestic demand. Real Private Domestic Final Purchases can be thought of as the “meat and potatoes” of the U.S. economy. It is made up of consumer spending, business fixed investment, and residential homebuilding (alternatively, it’s GDP less the change in inventories, net exports, and government). PDFP is less volatile than GDP (as it excludes the more choppy GDP components).
Federal Reserve policymakers are well aware of the GDP arithmetic. However, most officials favored a more cautious approach in deciding when to raise rates in mid-March, before the recent string of softer GDP components. Ahead of the March policy meeting, the Fed conducted its regular surveys of primary dealers and bond market participants. Both of those surveys showed a median expectation that the Fed would move in June and follow up with two additional 25-bp moves by the end of the year. Meanwhile, the federal funds futures market is now suggesting about a 64% chance of just one rate hike by the end of this year. The yield on the 2-year Treasury note has fallen by nearly 30 basis points since mid-March, also consistent with a more cautious glide path for the overnight lending rate.
Over the next several weeks, the economic data are likely to remain mixed, but with a more positive tilt overall. Stock market participants may react to the twists and turns in the data reports, but are more likely to use selective figures as support. Earnings reports ought to be important in the near term, but much of the bad news may be already factored in.
Over the last several months, the bond and currency markets have over-reacted to shifts in the Fed policy outlook. All else equal, tighter monetary policy should be positive for the dollar, but by how much? The dollar moves over the last year, especially relative to our largest trading partners, Canada and Mexico, appeared way overdone, and we have seen a reversal. A more gradual tightening path may hasten that correction. A softer dollar would also help the rest of the world to recover.