US Census says manufacturing new orders improved month-over-month. Our analysis shows the rolling averages declined, and the rolling average growth is in contraction year-over-year.

Analyst Opinion of Census Manufacturing Sales

According to the seasonally adjusted data, it was ships & boats and civilian & defense aircraft that caused the improvement in the headline data. The data in this series is noisy so I would rely on the unadjusted 3 month rolling averages which declined and remains in contraction.

What should be concerning is the continual reduction of backlog/unfilled orders this year - although it improved this month.

Note the headline numbers are not inflation-adjusted.

(Click on image to enlarge)

US Census Headline: 

• The seasonally adjusted manufacturing new orders were up 0.3 % month-over-month.
• Market expected (from Econoday) month-over-month growth of -0.3 % to +0.7 % (consensus +0.3 %).
• Manufacturing unfilled orders were up 0.1 % month-over-month.

Econintersect Analysis:

• Unadjusted manufacturing new orders growth accelerated 1.1 % month-over-month, and down 1.3 % year-over-year.
• Unadjusted manufacturing new orders (but inflation-adjusted) up 0.5 % year-over-year.
• Three-month rolling new order rolling averages decelerated 1.0 % month-over-month and are down 2.3 % year-over-year.
• Unadjusted manufacturing unfilled orders growth accelerated 0.1 % month-over-month, and down 1.6 % year-over-year
• As a comparison to the inflation-adjusted new orders data, the manufacturing subindex of the Federal Reserves Industrial Production growth was down 0.6 % month-over-month, and down 1.4 % year-over-year.

Seasonally Adjusted Manufacturing Value of New Orders - All (red line, left axis), All except Defense (green line, left axis), All with Unfilled Orders (orange line, left axis), and all except transport (blue line, right axis)

(Click on image to enlarge)

The graph below shows sector growth year-over-year.

Year-over-Year Change Manufacturing New Orders - Unadjusted (blue line) and Inflation Adjusted (red line)

(Click on image to enlarge)

Now, look at the manufacturing component of industrial production. While it is true that these are slightly different pulse points (inventory not accounted in shipments) - they should not have different trends for long periods of time.

Comparing Unadjusted Year-over-Year Change - Manufacturing Industrial Production (blue line) to Inflation-Adjusted Manufacturers Shipments (red line)

(Click on image to enlarge)

Using employment to confirm manufacturing growth says this industry's growth is growing.

Employment Growth - Manufacturing (Seasonally Adjusted) - Total Employment (blue line) and Year-over-Year Change (red line)

(Click on image to enlarge)

The health of manufacturing is gauged by the growth of unfilled orders. The 3-month rolling average rate of growth is currently slowing.

Unadjusted Unfilled Orders - Total Current Value (blue line, left axis) and Year-over-Year Change (red line, right axis)

(Click on image to enlarge)

A declining unfilled orders backlog could be a recessionary indication as unfilled orders generally decline in poor economic times. Keep the score on surveys, the following is a comparison of surveys to hard data - this Census data is the orange bars.

(Click on image to enlarge)

Caveats on the Use of Manufacturing Sales

The data in this index continues to be revised up to 3 months following initial reporting. The revision usually is not significant enough to change the interpretation of each month's data in real-time. Generally, there are also annual revisions to this data series. The methodology used by the US Census Bureau to seasonally adjust the data is not providing a realistic understanding of the month-to-month movements of the data. One reason is that the US Census uses data over multiple years which includes the largest modern recession which likely distorts the analysis. Further, Econintersect believes there has been a fundamental shift in seasonality in the aftermath of the Great Recession of 2007 - the New Normal. Econintersect determines the month-over-month change by subtracting the current month's year-over-year change from the previous month's year-over-year change. This is the best of the bad options available to determine month-over-month trends - as the preferred methodology would be to use multi-year data (but the New Normal effects and the Great Depression distort historical data). This series is NOT inflation-adjusted -Econintersect uses the PPI - subindex All Manufactured Goods. However, this is a rearview look at the economy. Manufacturing new orders or unfilled orders generally correlates to the economy - but it is not obvious in real-time whether a recession is imminent. So in context to economy watchers - manufacturing by itself cannot be used as an economic gauge.

Adjusted Value - New Orders (blue line) and Unfilled Orders (red line)

(Click on image to enlarge)

The same issues are also evident if manufacturing backlog is used as a recession gauge.