Richmond Fed Manufacturing Shipments Index

The Richmond Fed Manufacturing Shipments Index tracks monthly change in finished shipments versus the prior month. It is 1 of 3 components of the composite index published by the Federal Reserve Bank of Richmond. The Bank is 1 of 12 regional Reserve Banks acting under its Regional Surveys of Business Activity. The factory survey began in June 1986 and took its current monthly form in November 1993.

It covers makers with a NAICS code in the 300s across DC, Maryland, North Carolina, South Carolina, Virginia and most of West Virginia. Senior staff answer voluntarily without pay, with about 100 replies from several hundred surveys sent. It is a diffusion index (rises minus falls across firms), not national output, dollar sales or the size of each change. Readings above 0 mean expansion while readings below 0 mean contraction.

Responses are collected over about 3 weeks, from 2 business days after prior results until the Wednesday before publication. Questions cover change versus the prior month and expectations for the next 6 months, sent by email before 10 AM Eastern with 3 reminders. Results usually appear the fourth Tuesday at 10:00 to 10:10 AM Eastern, shifting around holidays.

Each release posts the composite index plus current and 6-month readings for orders, employment, prices, inventories and other gauges. Charts and history accompany each release, while seasonal factors update each July and errors are fixed in a separate note.

Shipments moves with order books and staffing, since all 3 feed the headline composite index.

• Shipments, 33%: change in finished product shipments versus prior month; captures realized outflow and demand fulfillment.
• Volume of new orders, 40%: change in incoming order books; captures forward demand pressure.
• Employment, 27%: change in headcount; captures labor input.

In April 28, 2020 release the composite fell from 2 in March 2020 to -53, its lowest reading and largest 1-month drop on record. All 3 parts fell, with shipments at -70 and new orders at -61, both record lows, while employment fell to -21. The drop was demand led, since physical outflow and order books collapsed together while employment fell less. A reading of -70 did not mean shipments fell 70%; it meant firms reporting falls far outnumbered those reporting rises.

1. Answers: each firm reports whether finished product shipments rose, fell or held steady versus the prior month, plus expectations 6 months out.
2. Diffusion: subtract the share reporting falls from the share reporting rises. Steady answers count only in the base, so scores run from -100 to 100. For example, 70 firms with 30% rises and 50% falls give -20.
3. Seasonal adjustment: the index is adjusted by an Arima-13 process to remove normal seasonal swings, with factors updated each July; raw readings are also published. The 5% trimming applies only to price data, not to shipments.
4. Composite: combine the 3 adjusted gauges as 0.33 times shipments plus 0.40 times new orders plus 0.27 times employment. As an illustration, gauges of -18, -25 and -16 give -20.26, reported as -20.