Gauging the Impact of December Inventories on 4th Quarter GDP Revisions

Two more reports from last week that we want to look at were on December sales and inventories for different sectors of the economy.  We want to know how they might affect 4th quarter GDP revisions.  Here's a simplistic way to think about how sales and inventories in the economy affect GDP: If the economy consumes 10 apples at a dollar a piece in November, and 11 apples at 90 cents each in December, sales have gone down by 1% in December but monthly GDP is up 10%.  If there are 10 apples left on the shelf in both months, reported inventories will go down 10% in December, but the inventory contribution to GDP will stay the same.  With that oversimplified example that in mind, let's look at what was reported..  

The first release on inventories we saw this week was the Wholesale Trade, Sales and Inventories Report for December (pdf) from the Census Bureau, which estimated that seasonally adjusted sales of wholesale merchants fell 0.4 percent (+/-0.9%)* to $454.6 billion from the revised November estimate of $451.7 billion, but were up 1.4 percent (+/-0.9%) from December a year earlier.  The November preliminary sales estimate was revised down by $0.6 billion or 0.1%, and hence was 0.4% lower than October.  Wholesale sales of durable goods were up 1.1 percent (+/-1.4%)* over November and were up 7.3 percent (+/-1.8%) from December a year ago, as wholesale sales of construction materials rose 5.4%, wholesale sales of electrical equipment rose 3.4% while wholesale computer equipment sales fell 2.1% from November.  Seasonally adjusted sales of nondurable goods were down own 1.7 percent (+/-1.1%) from November and down 3.5 percent (+/-1.9%) from last December, as wholesale sales of petroleum and petroleum products fell by 13.7%, largely due to lower prices.  Excluding oil sales, sales of non durable goods rose 2.1% in December, as wholesale drug sales rose 4.8% and miscellaneous wholesales sales rose 4.6%.  Note that the asterisks following the range of likely percent changes indicate that Census does not yet have sufficient statistical evidence to determine whether sales actually rose of fell for the periods indicated.

This release also reported that seasonally adjusted wholesale inventories were valued at $547.6 billion at the end of December, 0.1% (+/-0.4%)* higher than the revised November level and 6.7% (+/-0.9%) above last December's level, while November's preliminary inventory estimate was revised up by $0.1 billion, statistically insignificant.  Wholesale durable goods inventories were up 0.2 percent (+/-0.5%)* from November and up 7.8 percent (+/-1.4%) from a year earlier, with wholesale inventories of computers, peripherals and software up 2.6% while while inventories of electrical equipment and appliances fell 1.7%.  Inventories of nondurable goods were down 0.1 percent (+/-0.4%)* from November while they were up 4.9% (+/-1.2%) from last December, as wholesale inventories of chemicals were up by 3.4% while wholesale inventories of petroleum and petroleum products were down by 6.2%.  Again, excluding inventories of petroleum and petroleum products, wholesale non-durable inventories in December were 0.5% greater than those in November.  In part due to the distortion caused by lower petroleum prices, the closely watched inventory to sales ratio of merchant wholesalers rose to 1.22, up from 1.21 in November and up from the inventory to sales ratio of 1.16 in December of last year, as the inventory to sales ratio for petroleum and petroleum products, which is about 10% of wholesale sales but just 3% of wholesale inventories, rose from 0.31 to 0.34...

Then on Thursday, the Census Bureau released the composite Manufacturing and Trade Inventories and Sales report for December, which is covered in the media as the "business inventories" report, and which estimated the combined value of seasonally adjusted distributive trade sales and manufacturers' shipments was at $1,331.2 billion in December, down 0.9% (±0.9%)* from November, while 0.9% (±0.4%) above the total monthly sales level of December of last year. Manufacturer's sales were estimated at $488,245 million, down 1.1%, while retailer's sales were estimated down 1.1% at $393,208 million and, as previously noted, sales of merchant wholesalers were down 0.4% and accounted for $454,587 million of the overall total.. Once again, most of the drop in business sales was associated with lower oil prices; even manufacturer's sales are more than 10% refinery sales, as we mentioned when we discussed December factory orders..

Meanwhile, total manufacturer's and trade inventories were estimated to have increased 0.1 percent (±0.2%)* from November to a seasonally adjusted $1,764.4 billion at the end of December, which was up 3.9 percent (±0.5%) from December a year earlier.  Seasonally adjusted inventories of manufacturers were estimated to be valued 0.3% lower at $634,786 million, inventories of retailers were estimated to be valued at $562,881 million, a 0.1% increase, and inventories of wholesalers were estimated to be valued at $547,648 million at the end of December, up 0.1% from November.  The month end total business inventories to total sales ratio, the metric which is widely watched to determine if inventories are becoming excessive, was at 1.33, up from 1.31 November and up from 1.29 December a year ago, again likely a distortion caused by record high petroleum inventories... 

Both sales and inventories from this report were included in 4th quarter GDP, but December inventories were only estimated when the GDP report was released two weeks ago, when the BEA assumed that wholesale and retail inventories had increased and that nondurable manufacturing inventories had decreased for the month...in reporting on this report, most analysts assumed that this report indicated that the BEA overestimated inventories and hence marked down their estimates of GDP...for instance, economists polled by Reuters estimated that GDP could be lowered by at least 0.5%; Wells Fargo economists concurred; Macroadvisers estimated a total hit of 0.6% to GDP from a combination of December reports.  However, it's likely that some of these estimates have not adjusted inventories for inflation, a necessary prerequisite to determine their impact on GDP.  In checking this, we have little data to go on; we can't tell from looking at the GDP report (pdf), as it only gives us the quarterly change in inventories, and that at an annual rate to boot, and it doesn't give a deflator for inventories, either.  So we'll try to figure it out by walking through the steps the BEA itself would likely take in arriving at that result..

We'll start with wholesale inventories, which we have already noted were up just 0.1%, largely due to a 6.2% drop in petroleum inventories.  Most of the inventories covered by this report would be deflated by the BEA using the producer price index for December, which showed a 1.2% drop in prices for finished goods, along with a 1.7% drop in prices for intermediate goods and a 5.0% drop in prices for raw goods.  However, most of that was as the result of lower energy prices, which fell 6.6% at the finished energy goods level.  That alone would indicate a real increase in energy goods inventories, which was borne out anecdotally by news articles in December indicating seasonally record high stocks.  Looking at foods, we see wholesale prices were down 0.4% for finished foods, down 0.8% for intermediate foods, and down 6.9% for raw foodstuffs.  The wholesale inventories report indicates that inventories of wholesale groceries rose by 1.1%, while inventories of farm products rose by 0.8%.  Once adjusted for prices, BEA should find that real inventories of wholesale groceries rose by around 1.5%, while real inventories of farm products may have risen by as much as 5.8%, and it's those real inventories that would be applied to GDP.  Meanwhile, wholesale prices less food and energy were up 0.2% for finished goods, down 0.6% for intermediate goods, and down 0.5% for raw goods.  To determine real inventories for each of those goods, the BEA would use the corresponding itemized tables in the producer price report.  Prices for finished wholesale goods, for instance, are listed in table 4 under "Final demand goods less foods and energy".  By way of example, wholesale inventories of computer equipment were up 2.6% in December, while their prices were down 0.7%; that would suggest that an increase of 3.3% in real inventories of wholesale computers for December would be applied to 4th quarter GDP.

Similarly, inventories of goods at retail would be deflated with the various components of the consumer price index for December, which showed a drop of 0.4%.  Again, energy prices were a major factor, but even excluding food and energy, prices for goods less food and energy were down 0.3%.  But we dont have to pick through the CPI report to get the deflator for December, because the BEA has already computed it in their income and outlays report for December, which we looked at last week.  The PCE price index was down 0.2%, already implying an increased inflation adjustment to personal consumption expenditures.. But to adjust retail inventories, we'd have to use the PCE price index for goods, which we find was down 1.0% in December, as was shown in Table 9 of the pdf version of that report.  Applying that to retail inventories, which are not broken down by category, we'd estimate that real retail inventories rose 1.1% in December, in keeping with or even more than the BEA estimate...

We would find the details on factory inventories in the the Census Bureau's Full Report on Manufacturers’ Shipments, Inventories, & Orders for Decmeber (pdf), but again they would be deflated with the appropriate price index for the types of inventories the various manufactures are accumulating.  Many of these price indexes would again be found in the producer prices report.  Inventories of non-durables, which were estimated in the advance GDP report, were down 1.5% in December, but again much of that was due to an 8.8% drop in refinery inventories, which make up 15% of non-durable factory inventories.  Other factory inventories which fell in December included textiles, rubber products, pesticides and industrial chemicals, all of which use petroleum as an input.  Indeed, checking the producer price index, we find prices for industrial chemicals were down 4.4% in December, indicating the 0.6% drop in inventories of chemicals wasn't a real drop at all, but actually a real increase of around 3.8%.  Other components of factory inventories are less clear.  To figure inventories of capital goods, for instance, we might have to estimate using the GDP deflator for equipment, which showed an annualized price increase of 1.1% in the 4th quarter.  That would mean about a 0.1% per month reduction from the 0.3% nominal increase to arrive at the change in real inventories of December capital goods...

All in all, it appears that the BEA estimates for 4th quarter GDP inventories were pretty close to on the mark, and that the estimates of economic forecasters that reduced inventories will cause a major writedown of  4th quarter GDP will prove to be unfounded.  Of course imports, which we discussed last week, is another matter.  Lower prices for those will just make their real subtraction from GDP that much larger...

(crossposted from MarketWatch 666)

Meta: