Tracking the Demand of Goods in the US with Equities

I’ve previously shown how the change in the demand of goods in China can affect its largest trading partners economically and financially (i.e. equities, currency, rates). This post is going to focus on how I analyze the demand of goods in the US using economic data and confirm it with the relative performance of equities.

In this post, I’ll break down the data into three categories:

  • New manufacturing orders and utilization
  • The transportation of goods: shipping and rails
  • The purchase of goods by consumers: autos, retail sales, and housing

What this post will show is that the US leading indicators topped months ago and the hard data is currently peaking or starting to decline. Unfortunately, this is already occurring while discretionary spending from the US Government is a tailwind for the economy.


New Manufacturing Orders and Utilization

The ISM manufacturing survey has long been considered a leading indicator of the US economy regarding future demand of goods.  Here’s an example that I’ve shared previously describing why it leads:

Before a consumer can purchase a new car, an order must first be placed to produce all the parts that are needed to make the car before it can be assembled and then transported by rail and truck to the dealership. An increase in demand for new autos will see more parts and autos being assembled leading to companies and workers making more money that can be spent. As these workers make more money, they spent more at restaurants, retailers, concerts, etc. As these businesses begin to see more demand, these workers get paid more and can purchase additional items as well, like a new car or truck, leading to additional new orders at the auto plants to keep up with the increased demand. At some point, the demand for new cars that will slow, which leads to fewer orders being placed, less money being paid to the auto manufacturing workers, who in turn spend less at restaurants, retailers, concerts, etc. As the demand at these businesses slow, the workers get paid less leading them to decrease their spending as well.

Within the ISM survey are various sub-indices. The most forward looking indicator in the ISM survey is the New Orders. Presently, ISM New Orders peaked in January while ISM peaked a month later.

ISM New Orders (white), ISM Manufacturing (red)

Source: Bloomberg

The surveys are considered soft data points because these are estimates, or guesses, as to how the future will look. Even though these are estimates, the results of the survey have been robust for over 100 years.

Confirming the peak in ISM New Orders is the relative performance of the Machinery Industry, which has been very correlated to ISM New Orders and/or global growth for decades. During the last economic cycle (2000-2008), these companies benefited from the growth of China and Emerging Markets. As the growth rates in those areas of the world have slowed since 2011, the Machinery stocks have traded closer to the ISM New Orders like they did in the 1990s.

ISM New Orders (white), Machinery vs S&P 500 (red)

Source: Bloomberg

Since these data points are from surveys (i.e. ISM New Orders, ISM, Machinery vs S&P), there are two other sets of economic data that I track to confirm a peak in manufacturing.

The first set of data points are the new orders associated with Durable Goods, Capital Goods Non-Defense, and US Manufacturing. All three of these report the orders they have actually received compared to the ISM New Orders, which is an estimate of what they believe they will receive.

Historically, these data points peak about six months after ISM New Orders. By the time this peaks, the Machinery and most cyclical stocks have already declined so this is not a leading indicator for these stocks, hence why I use it to confirm the previous data points.

Currently, these three indicators are going through the process of peaking, thus confirming the decline in ISM New Orders and the Machinery stocks underperforming the S&P.

ISM New Orders (green, 3ma Y/Y), Durable Goods New Orders (blue, 6ma Y/Y), Capital Goods New Orders Non-Defense (red, 6ma Y/Y), US Manufactures New Orders (yellow, 6ma Y/Y)

Source: Bloomberg

The second set that I track deals with the efficiency of manufacturing facilities: Industrial Production, Capacity Utilization, and Industrial Production for Manufacturers. These data points are historically less noisy and have a tendency to peak slightly before the New Orders received data series.

Similar to the New Orders received, two of the three indicators look like they are peaking.

Industrial Production (blue, 6ma Y/Y), Capacity Utilization (green, 6ma Y/Y), Industrial Production for Manufacturers(purple, 6ma Y/Y), ISM Manufacturing (yellow)

Source: Bloomberg


The Transportation of Goods in the US

One area that doesn’t get a lot of focus is the shipping data because of it is lack of depth. While the Los Angeles (LA) and Long Beach (LB) data begins in 2000, the third biggest US port (NY-NJ) starts in 2011. However, and this is shown in the charts below, the difference in only looking at the LA, LB and Savannah data compared to all five major ports is minimal (i.e. variance of approximately one month in turning points).

Within the shipping data there are two data points that I follow closely:

  • Empty Containers
  • Inbound Containers

Those that follow the shipping industry usually ignore the Empty containers. However, if the growth rate of Empty containers is declining, this means there is a decline in the expected growth rate of demand for goods. Additionally, when Empties go negative, it means that shipping container companies (i.e. CAI, TGH) would rather pay to have their containers sit ideal at the ports rather than moving it to a different location to be filled with goods to make money.

Inbound container data is noisier than Empties so I use it as confirmation of the trend in Empties. Similar to Empties though, a declining and negative growth rate means a decline in the expected demand of goods.

I don’t focus too much on the Outbound containers because it is half the size of Inbound and tends to be driven by changes in the USD rather than actual demand growth.

Currently, Empties peaked in September 2017 on a 6 month average and should go negative when the June data is released. Inbound peaked a month earlier than Empties but has seen a slower rate of decline.

Source: TEU Data from Port Websites

Source: TEU Data from Port Websites

Finally, in terms of shipping related equities, I tend to focus more on the container companies than the actual shipping companies due to the simpler business models. In 2013, these companies had issues with overcapacity as trade peaked but they recently topped in 4Q17 when Empties and Inbound began to decline.

Textainer Group (TGH) vs S&P (white), CAI International (CAI) vs S&P (yellow)

Source: Bloomberg

The other data point that I watch regarding the transportation of goods is the Intermodal rail volume. Broadly, Intermodal volume deals with the movement of finished goods, which is going to be correlated to the demand of goods by consumers. The other data point that others tend to follow is the Carload rail volume, which deals with the movement of raw materials. The problem with Carloads is that a third of it is coal and the demand of coal is easily affected by the price of natural gas, which the US has ample amounts.

Similar to other hard data, Intermodal volume tends to confirm peaks in the ISM. Currently, Intermodal is still increasing and not showing the same concern as the shipping data.

ISM Manufacturing (red), Intermodal Rail Volume (blue, 26 week average)

Source: Bloomberg

Looking at the Rail Industry, the group continues to outperform the S&P. Rail stocks tend to peak after the Intermodal volume because we usually see Oil declining as manufacturing declines, thus decreasing their operating expenses and increasing their margins.

Rail Industry vs S&P

Source: Bloomberg


The Purchase of Goods by Consumers

Since 70% of the US economy is consumption driven, the purchase of goods by consumers is an important final step in this analysis.

First, the growth rate of auto sales tends to peak early in the economic cycle and then slowly decline until the cycle ends. Therefore, while it’s good to see the current increase in auto sales, I don’t expect it to be sustainable based upon previous economic cycles.

US Auto Sales (3ma, Y/Y)

Source: Bloomberg

Confirming my view on auto sales is the Auto Manufacturers and Components Index, which continues to underperform the S&P 500.

Auto Manufacturers and Components Index vs S&P

Source: Bloomberg

Second, Retail Sales. The most quoted number by investors and the media is the broad based Retail Sales data point, which includes auto and gasoline sales. However, auto sales can skew the data (peaks early and large dollar amounts) and changes in gasoline sales can be driven more by the change in the price of gasoline, not an increase in demand. Therefore, analyzing Retail Sales ex Autos and Gasoline allows me to focus on the actual demand for goods by the consumer.

When looking at Retail Sales ex Autos and Gasoline, we see the growth rate peaked in 4Q17 at 5% and has held steady in the high 4% the past few months. In other words, not accelerating but not declining either.

Retail Sales (red, 3ma Y/Y), Retail Sales ex Autos and Gasoline (blue, 3ma y/y)

Source: Bloomberg

Specialty Retailers tend to outperform the S&P in the early and mid-point of the cycle regardless of the growth rate due to the broad growth seen by the economy and consumer. As we move further into the cycle the group tends to move with the growth rates of Retail Sales ex Autos and Gasoline, as seen with the 2000, 2006, and 2016 peaks in performance.

Specialty Retailers Index vs S&P 500

Source: Bloomberg

Finally, housing. When looking at housing, it’s important to understand that Existing Home Sales are nine times greater than New Home Sales. Currently, Existing Home Sales are seeing a negative growth rate and New Home Sales peaked in February.

Existing Home Sales (blue, 6ma Y/Y), New Home Sales (red, 6ma Y/Y)

Source: Bloomberg

The MBA Purchase Index is also confirming the weakness in the housing data.

MBA Purchase Index Y/Y: 26 wk average (blue), 12 wk average (green), 8 wk average (yellow)

Source: Bloomberg

When looking at equities, and specifically Whirlpool (WHR) and Lennar (LEN), we see that the relative performance topped in 2015. This was slightly before the MBA Purchase Index peaked in 1Q16 and when interest rates (Fed Funds and the National 30yr Fixed Rate) began to increase.

Whirlpool (WHR) vs S&P (white), Fed Funds Rate (yellow, inversed), Bank Rate National 30yr Fixed Rate Average (green, inversed)

Source: Bloomberg

Lennar (LEN) vs S&P (white), Fed Funds Rate (yellow, inversed), Bank Rate National 30yr Fixed Rate Average (green, inversed)

Source: Bloomberg



When looking at the overall demand of goods in the US, we see:

While it is a major concern that everything seems to be lining up to slow all at the same time, there are still areas of the economy doing ok. My uneasiness with the US economy will begin to increase as the discretionary spending by the US Government begins to fade when the new fiscal year begins October 1st.

Something I underappreciated earlier this year was the strength of the tailwind from the fiscal spending provided to the US economy the past few months. However, this turns into a headwind as the spending is reversed with the new fiscal year.

As demand for goods in the US slows, this will put additional pressure on China, Emerging Markets, and the other Developed Economies that manufacture the goods purchased in the US. In this scenario, and since these areas are already slowing, I would continue to expect US equities to outperform on a relative basis and is why I continue to recommend the same allocation I have previously.

On an absolute basis, I am getting more concerned about the ability for US equities to continue to move higher once all the economic data points slow. However, it seems that we are still a few months away from that occurring at this time.


  • Overweight (or long) the USD
  • Underweight (or short) Emerging Market currencies


  • Overweight (or long) long dated maturing US debt
  • Underweight (or short) Emerging Market debt


  • Overweight (or long) S&P 500
  • Underweight (or short) Emerging Markets
  • Overweight (or long) US defensive sectors and industries
  • Underweight (or short) US cyclical sectors and industries