Economic Surprise Indices have been rolling over for a month or two now. The trend of US macro data has also disappointed in a period when it would be expected (empirically) to accelerate. However, taken anecdotally or cherry-picked managers can find plenty of ammunition to support the to-infinity-and-beyond Birinyi forecast (though often it relies on the most manipulated and adjusted government provided time-series). Overnight's concerns on China show just how quickly confidence can be upset but Goldman's Jan Hatzius sees three main factors for why their GDP-tracking estimate is weakening already (more like 2% than 3-3.5% growth) and that we are seeing slightly softer data already. The end of the inventory cycle, the pulling forward of demand thanks to the warm weather aberration, and the already clear impact on consumption from higher gasoline prices will likely shift from an overstated economic trajectory to more muddle-through or worse for Q2 onwards.
Goldman Sachs: Sticking With Sluggish
The US economic data over the past few months have clearly outperformed expectations. Our current activity indicator (CAI) is running at 3.5% in February given the data in hand so far, and is tracking 2.9% for the first quarter as a whole. However, we expect the numbers over next 2-3 months to slow to a pace that looks more consistent with a 2% overall activity growth pace rather than 3% or even 3.5%.
1. Warm weather has pulled forward activity.
Some of the recent strength in the CAI is likely to reflect the exceptionally mild 2011-2012 winter. To be sure, there are some areas where mild weather has a negative impact on economic activity, such as utilities output and perhaps some retailers that sell seasonal goods. But this is likely to be offset by areas where the impact is positive, as construction and other outdoor activities decline by less in not seasonally adjusted terms than the seasonal factors "expect" and this gets translated into a large seasonally adjusted increase. Overall, we found that the weather has contributed an estimated 0.3 percentage points to the 2.9% annualized growth rate of the CAI in the first quarter so far (see Zach Pandl, "Growth Impact of a Mild Winter," US Daily, March 1, 2012). The impact on a more comprehensive measure of activity that includes both the CAI and GDP would be a bit smaller, perhaps 0.2 points, as GDP is probably less weather-sensitive.
2. The inventory cycle has helped.
The pickup in inventory accumulation from -$2 billion (annualized) in the third quarter to +$54 billion in the fourth quarter contributed 1.9 percentage points to the Q4 growth rate. Moreover, inventory accumulation seems to have picked up a bit further in the early part of the first quarter, judging from the Commerce Department's book-value inventory numbers for January as well as the ISM manufacturing survey for January/February. This suggests that inventories may still be making a positive growth contribution for the time being. But while we are not close to a situation where inventories start to look "heavy," a further positive impact in coming quarters is not likely.
3. Gas prices are starting to cut into real income.
Gasoline prices rose sharply in January and February. Using weekly data from the US Department of Energy, they are now up 9.1% from their end-2011 level on a seasonally adjusted basis, with most of the increase likely to show up in February and March on a month-average basis. According to our models of the link between gasoline prices and growth, such a hit might take 0.3-0.4 percentage points off real GDP growth over the subsequent year. Moreover, using monthly data on the link between gasoline prices and consumption, we find that the impact becomes visible about 1 month after the initial hit, so this would imply that the impact would show up in March and April.
There may be some early signs of deceleration in the data.
The data surprises have indeed turned a bit less positive in recent weeks, although it is too early to say definitively whether this is noise or a more lasting shift. In March so far, our US-MAP scoring system for the economic data relative to consensus expectations has averaged a slightly negative reading, despite the better-than-expected February employment report. Of course, some of this just reflects the fact that consensus expectations have caught up with the better data. However, there are also some faint signs in the more forward-looking indicators that the tone of the data may be shifting down a notch. In particular, the new orders indexes of the February ISM, March NY Empire State, and March Philly Fed survey all fell moderately.
Our bottom line is that there are several reasons to believe that the recent data may have overstated the strength of the US economic data. For the first quarter as a whole, our current best guess is that a broad measure of activity growth that puts most of the weight on the CAI but also some weight on the GDP bean count is currently running at a 2.6% pace; we believe that this might overstate true growth by perhaps 0.2 percentage points because of weather, so that the second quarter is likely to understate growth by 0.2 percentage points (for a "swing" of 0.4 percentage points); the end of the inventory cycle might take a couple of tenths off growth; the impact of the gas price increase might take another few tenths off growth; and we may be seeing some signs of softer growth in the most recent data already. All told, we believe that the numbers are likely to slow to a pace that looks much more consistent with a 2% rather than a 3% or even 3.5% growth pace through the end of the second quarter.