US Retail Inventories

There is a growing concern regarding the recent increase of inventories in the US economy, including in the retail sector. A recent article in the WSJ illustrates that: Retailers’ Full Shelves May Force Holiday Discounts.

Main takeaways

  • Inventory/Sales ratio in the retail sector is increasing since 2012 and jumped substantially since late 2014.
  • Inventories are reported monthly for kind of business representing 80% of the total inventories in the sector.
  • Among these business, there is no common trend in the I/S ratio.
  • Consolidating all those business categories shows a stable I/S ratio in the aftermath of the recession and a minor jump by early 2015.
  • So the uptrend (and jump in 2015) in I/S ratio is due to “others" – business for which the US Census Bureau does not report monthly breakdown on inventories.
  • Looking at the breakdown of sales in the “other" business, one can infer that lower gasoline sales (lower prices) are behind the drop in overall “other" sales.
  • Meanwhile, the US Census Bureau estimates show “other" inventories growing at the same pace as overall inventories – this seems to be an error that is causing the I/S ratio for “other" business to increase materially and affect the shape of the overall I/S ratio that everyone is concerned about.

The bottom line is that the increase in I/S ratio in the retail sector is likely caused by an error in estimating gasoline inventories.

Facts

There’s not much information available on retail inventories. The annual survey breaks down total retail inventories in 12 groups and the monthly survey brings information on just 6 groups that account for around 80% of the total retail inventories.
Total retail sales (excluding food services) was USD 395bn in September/15 (seasonally adjusted). This is equivalent to a pace of annual retail sales of USD 4.7tn (26% of GDP). Retail inventories were USD 584bn in Sep/15 (seasonally adjusted), representing around 1.5 months of sales.

The chart below plots the monthly seasonally adjusted inventories/sales ratio since Jan/1992.

The chart shows that I/S ratio first reached a low of 1.34 months on Oct/2011 and has since increased to 1.48 months (Sep/2015).
It is difficult to square the uptrend in I/S ratio observed since Oct/2011 with the comments in the WSJ report, which only mentions the most recent few months as a reason for concern. It could be the case that the initial build up on I/S ratio was a healthy adjustment in the aftermath of the crisis, but the recent pick up in inventories was an error… however this also does not match the fact that retail sales growth (excluding gasoline) continues in a healthy trend – see US Retail Sales – trend remains unchanged (Oct/2015).

Breakdown of I/S ratio

Total retail inventories are split into the following kind of business:

  • motor vehicle and parts dealers (33%)
  • general merchandise stores (15%)
  • building materials (9%)
  • clothing (9%)
  • food and beverages (8%)
  • furniture,electronics and appliances (5%)
  • others (21%)

Motor vehicle and parts dealers alone account for about one-third of the retail inventories, but the chart below shows that the increase in inventories/sales is not related to the auto sector (indeed, I/S for the motor vehicle sector has been sideways).

Motor vehicle and parts dealers

General merchandise stores

Building materials, garden equip. and supplies dealers

Clothing and clothing access. stores

Food and beverage stores

Furniture, home furn, electronics, and appliance stores

Others

The remaining category “others" is calculated by residual since its breakdown is not available in the monthly data. The “other" group includes health and personal care; gasoline stations; sporting goods, hobby, book and musical instrument; miscellaneous retailers; nonstore retailers (mail order, online).

Additional comments

One can see that increasing inventories in clothing stores and “others" seem to be behind the overall increase in I/S ratio. Unfortunately, there’s no individual kind of business that dominates the “others" inventories. Based on annual data, “others" inventories are split in:

  • nonstore retailers (30%)
  • health and personal care stores (30%)
  • sporting goods, hobby, and music instrument stores (17%)
  • miscellaneous store (14%)
  • gasoline stations (10%)
The chart below plots the I/S ratio for total retail sales excluding motor vehicles and “others". One can see that there was an increase in I/S ratio from late 2014 early 2015 but there is stability in recent months. Moreover, the increase in I/S ratio is materially less pronounced than what is observed in the total I/S ratio (see first chart at the top).

The next chart plots the sales of the “other" categories relative to the total retail sales excluding “other" and vehicles.

The next chart plots the inventories of the “other" categories relative to the total inventories excluding “other" and vehicles.

The charts above illustrate several interesting economic shifts. For instance, the downtrend in I/S ratio of the “other" categories from 2000 to 2008 was led by a faster increase in sales of the “other" categories relative to total sales. This is very likely the result of increasing e-commerce. Indeed, the share of nonstore retailers' sales in the “other" category increased steadily from 25% in 2000 to 38% in 2014.

But for the purpose of understanding why I/S ratio in the “other" categories jumped in late 2014, we need to look further.

The chart below excludes gasoline sales (gas stations and fuel dealers) from the “other" category and plots its sales relative to total sales excluding vehicles and “other". The result? “Other" sales are booming!

The charts above show that “other" inventories are increasing at the same pace of overall inventories but “other" sales collapsed. We learned that the collapse in “other" sales is due to gasoline sales – indeed, “other" sales excluding gasoline are increasing faster than the remaining sales. Unfortunately, the US Census Bureau does not measure monthly gasoline inventories (it is only measured annually) and it seems that they are failing to account for the drop in gasoline inventories as measured in current dollars – since there is no reason to believe that the volume of gasoline inventories has increased substantially in the retail sector to offset the fall in gasoline prices.

Conclusion

The overall tentative conclusion from the charts above is that gasoline inventories are mismeasured and, therefore, are distorting the overall inventory/sales ratio in the retail sector.

Dr. Paulo Gustavo Grahl, CFA (2015-11-23)