One month ago, the common media narrative held that the US economy was about to overheat. Now, folks fret a slowdown! What gives? After the Atlanta Fed’s GDPNow forecast on February 1 spooked some worried about inflation by predicting 5.4% Q1 GDP growth, softer data helped drive the GDP projection down to just 2.6% as of February 27. Now it sits at 3.5%—a roller coaster! In our view, this speaks to the futility of obsessing over short-term data—particularly since stocks generally look further beyond the next couple months.
We don’t highlight GDPNow to punch it, but because its swings show how variable data can be. With little else to go on, GDPNow initially gave January’s higher-than-expected manufacturing PMI (59.1) a huge weighting—even though PMIs don’t actually measure growth’s magnitude and manufacturing is only 11.5% of the economy.[i] January employment data—next on the Atlanta Fed’s docket—reduced manufacturing PMI’s weighting and sliced the GDPNow projection to 4.0%. In our view, this is also odd considering employment data are backward-looking—January’s shed light on the economy’s state in late 2017.
Then a few more weak January readings came out, including retail sales’ -0.3% m/m decline, industrial production’s -0.1% m/m dip and durable goods orders’ -3.7% m/m drop. January real personal consumption expenditures, published March 1, also showed a slight dip (-0.1% m/m). All this has folks thinking Q1 growth will be much slower than previously thought—and “overheating” worries morphed to “lost momentum” fears. But these figures are a partial look at one month—not exactly a trend, nor worth fretting. Moreover, plenty of other data (like the aforementioned ISM survey) suggest continued growth.
We also suspect people may be overlooking seasonality. To an extent, economic growth fluctuates throughout the year in predictable patterns—shopping, for example, typically pops around the holiday season, while poor winter weather might slow spending or disrupt output. Statisticians attempt to smooth out these fluctuations by “seasonally adjusting” the numbers. However, measured Q1 GDP growth has lagged the other three quarters’ for decades—and the gap has grown during this expansion. In recent years, “hard data” from the Commerce Department and Federal Reserve have also trailed “soft data” like the ISM surveys throughout Q1.
The Commerce Department’s Bureau of Economic Analysis tried to improve its seasonal adjustment methods in 2015 after a large Q1 GDP lag elicited skeptical headlines. In 2016, a low Q1 GDP read spurred another BEA investigation, which showed seasonality remained.[ii] They plan to tinker again this summer. Maybe this vanquishes the seasonality gnomes, but for now, it is quite possible weaker monthly data merely signal lingering math problems.
All this is Exhibit 1,001,299.4 that economics isn’t a science.[iii] Measurements aren’t precise, and statisticians adjust and revise them repeatedly over time in a never-ending quest for accuracy. They are still important, but always approach them skeptically—and remember very near-term data generally don’t dictate longer-term market moves. Stocks typically look at how political and economic drivers likely play out over the next 3 – 30 months. Any further out is getting speculative, any sooner is probably old news and already priced in. As we enter Q1’s final month, we very much doubt stocks are hanging on narrow data points showing how certain segments of the economy were faring in its first month—or predictions for the remainder of the quarter. Those confirm what stocks already reflect, but they won’t likely sway markets for long.
They grab headlines, though, making it easy for investors to fixate on them. Our advice: Don’t let short-term data dictate your thinking or decisions. They change too fast and mean too little for forward-looking stocks.
[i] Source: Bureau of Economic Analysis, as of 3/2/2018. Value added as a percentage of GDP in Q3 2017.
[ii] As a result of its investigation, the BEA also revised up Q1 2015 GDP growth—which an earlier estimate had as low as -0.7% annualized—to 2.0% growth. Today, it stands at 3.2%—an example of just how much data can change after the fact.
[iii] That exhibit number is also not a precise measure.
If you would like to contact the editors responsible for this article, please click here.
*The content contained in this article represents only the opinions and viewpoints of the Fisher Investments editorial staff.