Is Atlanta Fed GDPNow broken again?

Do US GDP estimates need to adjust for exports "other precious metals" given the ongoing silver, platinum and palladium rally?

Since early January, Atlanta Fed GDPNow data have shifted up significantly and point towards Q4 US GDP growth (due out in late February) clocking in above 5%.

This upward revision in Atlanta Fed GDPNow estimates reflects a sharp improvement in the model’s estimate of net exports which swung from -0.31 pp contribution to Q4 US GDP to +1.993 pp contribution of Q4 US GDP on Jan 8.

Stated simply, GDPNow suggests a notable increase in US net exports. Under the GDP identity, higher net exports imply higher US GDP.

What happened? The model’s shift followed the Bureau of Economic Analysis’ (BEA’s) release on January 8th of Oct. US trade data.

The Oct US trade data show sharp gains in US exports of both gold and non-gold “other precious metals” (think silver, platinum, palladium).

Some M-FAB readers may remember that the AtlantaFed’s model had a widely covered issue in Q1 2025 — when its GDPNow estimates collapsed and showed the US economy headed for an imminent recession.

Changes to the model were made in Q1, but it seems like the AtlantaFed’s release of a “gold adjusted forecast” through April 29, 2025 may have been too limited in its scope.

The chart below shows BEA data on non-monetary gold imports (yellow spike created issues in Q1 2025 estimate); green spike are non-monetary gold exports (now creating opposite issue in Atlanta Fed GDPNow Q4 2025 estimate).

The BEA in its US GDP estimate exclude non-monetary gold trade flows as GDP should not capture investment flows. However, BEA does not make comparable adjustments to GDP for other precious metals, including silver, platinum and palladium which are a derivative trade of gold’s rally.

A number of precious metals gained more than gold in 2025. Are none of the US trade flows in these precious metals investment related? It seems unlikely.

It also is understandable that other precious metals have generally flowed out of the US in Q4 because prices are higher overseas than domestically. The chart below compares silver prices in China with those in Europe/US and shows metal prices in China to be much higher.

What should M-FAB readers take away from this post?

My key points are:

• It appears likely that Atlanta Fed GDPNow once again is misfiring due to large precious metals trade flows — albeit now in the direction of over-stating its estimate of Q4 US GDP. Be cautious if you use this indicator in investment or business decisions.

• In light of the precious metals rally extending beyond gold and the export of other US precious metals abroad, the US trade balance shows improvement. This not “tariffs working” — the US trade data overstate the improvement as these precious metal flows are more investment in nature.

• The lack of BEA adjustment in its measurement of US GDP for exports of other precious metals like silver, platinum and palladium suggests upside risks to Q4 GDP coming in hotter than expected, but which also are not indicative of a stronger US economy. This risk in a major US macroeconomic indicator has implications for US fixed income trading and other business decisions.

• US exports of precious metals do not represent an improvement in US trade competitiveness and economic growth. Rather, these exports of precious metals are moving out of the US at very elevated prices for investment purposes.

  • A key question is whether these assets are moving on to the balance sheets of the physical metal ETFs (like SLV, PPLT) who generally maintain their physical metals positions outside of the US but may still be partially owned by US citizens or are US precious metal exports actually moving assets into the hands of foreign nationals.

  • If these precious metal exports are mostly shifting ownership to foreigners, such exports, indeed, are indicative of a lack of a rising confidence in the US dollar.

• The proceeds of US exports of precious metals abroad represents significant investment income and is likely boosting US aggregate consumption and potentially also investments in other asset markets. But, US households and businesses have finite quantities of precious metals to export abroad. Thus, an eventual slowdown in US precious metal exports may have broader US macroeconomic implications, potentially including less aggregate US consumption and a slower US economy in the future.

  Wishing M-FAB subscribers a good start to the week!