(Click on image to enlarge)

GDPNow Data from the Atlanta Fed, chart by Mish.

Officially the current GDPNow nowcast is -2.1 percent with Real Final Sales (RFS) at -2.7 percent.

However, GDPNow creator Pat Higgins has confirmed the posted numbers are too low by two full percentage points. He does not like to change the model mid-cycle.

My chart shows his stated, but not shown adjustments.

I covered this on March 7 in Statement from Pat Higgins on Gold Impact to GDPNow Forecast

Although GDPNow does [not] distinguish gold from other imports, the Bureau of Economic Analysis does, in tallying up the total of the net exports, subaggregate within GDP. Removing gold from imports and exports leads to an increase in both GDPNow’s topline growth forecast and the contribution of net exports to that forecast, of about 2 percentage points. The topline growth forecasts also increased today—standard model -2.4 percent to -1.6 percent, “gold adjusted” model -0.4 percent to 0.4 percent—as data from today’s labor market report came in stronger than the model was expecting based on the limited February data the model received prior to that release.

The above snip was from a Linked-In post by Higgins. I added the word [not] which is now corrected in Linked-In.

I commented …

Employment Added to GDPNow

Higgins said “ today’s labor market report came in stronger than the model was expecting based on the limited February data the model received prior to that release.“

As I have commented, it’s not the data that matters (the report was bad), it’s the data vs what the model expected (not what economists expected).

Today, the jobs consensus was 160,000 vs an actual of 151,000 with significant negative revisions too.

But for whatever reason (I strongly suspect wage growth), the model liked today’s report, boosting PCE spending estimates by 0.42 percentage points.

Higgins doesn’t show Real Final Sales reflective of today’s unemployment report but I will take a stab at (+0.8 for Employment – (0.57 – 0.43) for CIPI), a net of +0.64 to (-2.8 + 2.0) = -0.16 percent.

We will find out on Monday [March 17].

CIPI stands for Change in Private Inventories.

The red box now shows my calculation was correct. To a single decimal point, RFS was adjusted to -0.2 percent for March 7.

Real Final Sales

RFS is the bottom-line GDP number. The difference between GDP and RFS is CIPI which nets to zero over time. Since CIPI went up on March 7, I calculated RFS would not increase as much as the base forecast.

In the red box note that GDP went from -0.4 to +0.4 (0.8 percentage points) but RFS went from -0.8 to -0.2 (0.6 percentage points).

Retail Sales

Retail sales clobbered GDPNow and RFS by 0.5 percentage points each.

The current nowcast is -0.1 percent on the base forecast and -0.7 percent for RFS.

For those unaware, I was adjusting GDPNow before Higgins’ Link-In post.

For discussion, please see How Did Gold Imports Exacerbate the Huge Decline in GDPNow Nowcast?

The GDP estimate of GDPNow crashed on February 28. Gold played a big Impact.

My original correction was to subtract out all of the trade report, knowing that was overdoing things.

I have since confirmed with Higgins, by email, that the correct adjustment is two full percentage points for both GDP and RFS.

Message to the BEA

It would behoove the BEA and writers to make a note of top-line vs bottom-line estimates of GDP.

The NBER, the official arbiter of recessions, only considers the bottom-line estimate. This is one reason why two quarters of negative GDP are woefully insufficient to call a recession.

The last time GDP went negative for two consecutive quarters (now revised away), I said “no recession” on the basis of RFS. I tend to be early on recession calls, but I ducked that one.

My last recession call was for May-June of 2024. That hasn’t hit, so I am early again, but not as early as some who called recession as early as October 2023.

Related Posts

January 9, 2025: Biggest Drop in Revolving Consumer Credit Since the Great Recession

Consumer credit unexpectedly declined. But what do we make of it?

March 4: A Global Trade War Has Started – Global Recession Will Follow

The most significant global trade war since Smoot-Hawley and the Great Depression is underway.

March 7, 2025: A Historical Look at Unemployment Rates Heading Into Recessions

In 1994, the BLS made huge changes in the calculation of the unemployment rate. So that’s all the further back I look.

March 10, 2025: Trump and Secretary of Treasury Bessent Discuss the “Detox Recession”

Don’t worry, it’s just a little more pain and inflation disturbance before tariff greatness begins.

March 17, 2025: Retail Sales Barely Rise in February, They’re Negative Factoring in Revisions

Retail sales had another poor month in February. It’s very recession looking.

Also factor in tariffs, trade disruptions, DOGE firings, an abrupt halt to immigration and immigration-related spending, and massive business uncertainty about everything Trump does.

Where are we?

More By This Author:

Retail Sales Barely Rise In February, They’re Negative Factoring In Revisions
China Announces Wide-Ranging “Special Action Plan” To Boost Consumer Spending
Trump Wants A Weak Dollar But Needs A Strong One