Uncategorized
GDP arithmetic
admin | May 10, 2019
This document is intended for institutional investors and is not subject to all of the independence and disclosure standards applicable to debt research reports prepared for retail investors.
The arithmetic for translating monthly readings of various components of GDP to quarterly averages is rather simple, but it is not necessarily transparent. Choppy monthly results for consumer spending since last fall have led to substantial volatility in the quarterly averages. A quick tutorial on how the quarterly real consumer spending figures relate to the monthly data helps explain the soft first quarter results, and why the market should expect a robust acceleration in the spring.
December ruins Q1
Even though anecdotal reports on Christmas retails sales were mostly robust, both the Census Bureau retail sales and the BEA consumer spending results for December were disastrous. In real terms, consumer expenditures according to these numbers fell by 0.6% in December, the largest monthly drop since 2009. While the December hiccup helped to drag down the Q4 average, outlier results for the last month of a calendar quarter usually do the most damage to the next quarter.
The dip in December outlays brought the level at the end of the year well below the Q4 average (Exhibit 1). As a result, consumer spending spent much of the first quarter digging out of the hole created by the gap between the Q4 average and the low December level—a gap of almost $40 billion. A solid bounceback in consumer expenditures in January (+0.4% in real terms) barely brought the level of spending back up to the Q4 average. As the chart shows, if not for the explosive gain in March, real consumer spending would have barely grown at all in Q1, even though the January (+0.4%) and February (flat) moves averaged out to a trend-like +0.2% per month, which is a 2.4% annualized pace. The tremendous gain in March managed to boost Q1 to a still-anemic 1.2% annualized rise, slightly less than half of the pace registered in 2017 and 2018. Thus, in some sense, Q1’s fate for consumer spending was all but sealed in December, before the quarter even started.
Exhibit 1: Real Consumer Spending
Source: BEA.
March boosts Q2
The same arithmetic that doomed Q1 is going to boost the Q2 average considerably. As Exhibit 1 shows, the explosive rise in real consumer outlays in March pushed the level of spending at the end of the first quarter sharply above the quarterly average (by almost $60 billion). If real consumer expenditures were absolutely flat from the March level for April, May, and June, the Q2 average would still work out to an annualized gain of 1.8%, not great but only modestly shy of the underlying trend (and better than the Q1 result!).
More realistically, if spending in April, May, and June each grow by, say, 0.2% per month, the Q2 average would work out to a 3.4% annualized increase. My very preliminary forecast for Q2 is slightly more optimistic, assuming monthly increases of 0.25% per month, a 3% annualized clip, not much different than the 2017 and 2018 performances (2.7% and 2.6% respectively), and produces a quarterly annualized advance of 3.8%.
Consensus estimates too conservative
At this point, many readers are probably thinking that this arithmetic is ridiculously simple and should be obvious to everyone. Perhaps, but it clearly is not. The latest Bloomberg economist survey released just a few days ago calls for real consumer spending to post a gain of only 2.8% in Q2, which implies that the consensus forecast for average monthly gains over the April to June period is barely above 0.1%. This is probably the primary reason why my Q2 GDP forecast (2.8% as of this writing) is far above the consensus forecast of 2.0% (from the same Bloomberg survey).
You would think that economists would be wise to the routine by now. The March/Q2 pattern should be familiar. Last year, real consumer spending surged by 0.6% in real terms in March, which set the stage for a robust bounceback quarter (3.8% annualized in Q2, up from a rise of only 0.5% annualized in Q1). However, in the same Bloomberg survey from a year ago (i.e. as of early May), the consensus for Q2 consumer spending was 2.9%, nearly a full percentage point shy of the eventual result. Similarly, in March 2017, consumer spending in real terms surged by 0.7%, yet again setting the stage for a strong Q2 (in that case, by May, economists’ forecasts roughly matched the eventual Q2 result for consumption). Likewise, in 2014 and 2015, above-trend gains in March real consumer outlays laid the groundwork for explosive Q2 advances that exceeded the consensus forecast as of early May.
Thus, while it may seem foolhardy to have a far above-consensus call for real consumer spending and, in turn, GDP growth for Q2 before we have any hard data for the three months of the actual calendar quarter, that is a limb out onto I will gladly shimmy.