Misunderstood Measure: How to Approach "GDPNow"
The mistake some investors make with the Atlanta Fed's GDPNow model is forgetting it's actually two words: GDP (gross domestic product) and Now, meaning at the moment.
GDPNow seldom grabs the spotlight, but it did recently when it turned from green to red for the first quarter, dropping to as low as –2.8% earlier this month before climbing to –2.4% last week and –2.1% on Monday (with a fresh update due today). The news burdened stocks, reinforcing worries that the U.S. economy might be headed toward recession.
A recession may or may not be in the cards this time around, but GDPNow doesn't provide enough evidence to make the call.
While the model tracks a multitude of data feeding into the government's GDP calculation, it's on a real-time basis. A negative 2.1% or positive 2.1% at any given time isn't a prediction or a final reading. The model continues evolving as data arrive and is updated regularly, around six or seven times per month.
"It should be noted and emphasized that the GDPNow model is a nowcast, not a forecast, which means the –2.4% estimate is by no means a guarantee," according to Schwab experts last Friday in their latest Market Perspective. "It is also subject to change and will continue to do so as data for the quarter come in."
The model isn't an official forecast of the Atlanta Fed or its officials. Instead, it mimics methods used by the U.S. Bureau of Economic Analysis (BEA) to estimate real GDP growth by aggregating statistical model forecasts of 13 subcomponents that comprise GDP.
The last time the final GDPNow measure ended up negative was the second quarter of 2022, at –1.2%, when the first government estimate was –0.93%. The last time it was lower than the current level for the full quarter was during the pandemic, when it measured second-quarter 2020 GDP growth at –32.08%. That was less than 100 basis points above the first government estimate.
The measure being down –2.1% means that, so far, data for the first quarter indicate a 2.1% annualized quarterly decline. Most of that weakness is being driven by the net exports component, as imports recently surged relative to exports. A growing trade deficit can hurt GDP growth, but the spike in imports may have been a temporary element, driven by businesses making import purchases ahead of time to prepare for possible U.S. tariffs.
Data in coming weeks—including February retail sales and existing home sales—will have an impact on GDPNow. The measure includes everything from car sales to gambling expenditures to intellectual property—among many other elements that feed into BEA's GDP report. The BEA typically issues its initial estimate for quarterly GDP a few weeks after the quarter ends, meaning investors will see first-quarter data in late April. But even that's not the end because another estimate follows in May before the final estimate arrives in June.
Many major financial firms also issue GDP estimates, and the average of those is available at the GDPNow site. In addition, other Federal Reserve banks provide estimates, but those often take a back seat to GDPNow, at least when it comes to influence on the market. Analysts' estimates for first-quarter GDP growth remain slightly positive, because their predictions are for the entire period based partly on data that won't be available until after the end of the quarter in some cases.
GDPNow tends to gain in accuracy toward the end of the quarter as more data work their way in, the Atlanta Fed's research shows, and often overpredicts actual GDP slightly. In the third quarter of 2024, GDPNow's final forecast for 2.82% was slightly above the government's first estimate of 2.79%. In the second quarter of 2024, GDPNow of 2.61% compared with the first government estimate of 2.84%, while in the first quarter of last year GDPNow of 2.7% was far above the first government estimate of 1.59%.
In the long run, while the final figure has often been close, a single GDPNow estimate is simply that, and not the last word. Or to be clear, the last two words.