How Investors Can Use Manufacturing Data to Predict Growth
The ISM Manufacturing Survey contains two sub-indices that, when related to each other, give investors a way of ascertaining the most likely trend in future economic growth. Investors need to have an opinion about the economy that is based on sound logic and confirmed empirically. Alerting investors of a tool that can assist investors in this endeavor is the subject of this paper. We will use two sub-indices from the ISM Manufacturing Survey to create an indicator that has proven its reliability in predicting the future trend in real GDP. But before we do that, we must understand what it is that we are trying to predict – real GDP growth. And then we must understand a little bit about the manufacturing sector of the US economy.
Last week, the Census Bureau released the first of three reports on 3rd Quarter real Gross Domestic Product (GDP). Economists use the term "real" to indicate a series of data net of inflation. So, real GDP refers to the total value of all goods and services produced in the US with the effects of price inflation stripped out.
GDP is perhaps the most popular economic statistic. It's the most often used proxy for economic growth and the health of the overall economy. Specifically, GDP is the total value of consumption + investment + government consumption + exports - imports. In our opinion, GDP is only an average measure of economic health. A better measure would be the total value of a country's pool of savings. That is another matter, however. Investors need to be concerned with the performance of GDP simply because we know it's used to formulate monetary and fiscal policy, if for no other reason. Both policy types can have a dramatic impact on capital markets.
Aside from the fact mentioned above that GDP is a poor indicator of a country's economic health, it is also not very timely. It's computed only quarterly and often has significant revisions in the months following the initial release. But because the use of real GDP is so heavily ingrained in monetary and fiscal policy-making, anyone who can forecast GDP with any skill will have a decided advantage when it comes to investing.
It is a loser's game to try to predict the actual real GDP number. However, predicting the actual value is not very useful anyway. Instead, investors should focus on predicting the trend in real GDP growth. The trend is not only more useful to investors, but it is also considerably easier to predict than an actual value. Below, you will see a chart of the year-over-year growth of real GDP since 1970.
There are some economic reports that economists place more importance on than others. One of these economic reports is the Institute for Supply Management's (ISM) Manufacturing Survey. It has stood the test of time as a good indication of conditions in the US manufacturing sector. While manufacturing has been a shrinking sector of the US economy since World War II, the ISM Manufacturing Survey has maintained its correlation with the overall economy very well.
The ISM report is the result of a survey of 400 industrial companies in 20 diverse industries across every geographical region in the US. Survey respondents are asked to grade conditions at their company in nine key areas of manufacturing: production, new orders, inventories, supplier deliveries, employment, imports, new export orders, order backlog and prices paid. For each of these 9 manufacturing variables, respondents are asked if conditions are "Higher/Better," "Same," or "Worse/Lower" than the previous month.
A diffusion index for each of the nine key manufacturing areas is then computed. Simply put, the percentage of "Better" responses is added to one-half of the percentage of "Same" responses. We are then left with nine diffusion indices that range in value from 0 to 100, where any number above 50 is considered expansionary for the economy.
Two of the nine components of the report are New Orders and Inventories. Most people that have worked in manufacturing are aware that these are two very important metrics. New orders are the lifeblood of any business, as orders for the firm's product drive revenues and profits. A constant stream of incoming orders keeps employment levels of the firm relatively unchanged. An increase in new orders increases revenues, profits and employment. Therefore, it's easy to see how tracking new orders across an economy is beneficial to the investor.
As business conditions cycle through boom and bust, inventory levels tend to build and deplete in a similarly cyclical way. A consistently growing level of inventories indicate either overproduction or a declining level of sales. A decreasing level of inventories generally will indicate that sales are growing faster than inventories. If this is the case, production of goods must at some point be ramped up to meet existing levels of demand for the goods. The cyclicality of inventories can be seen in the chart below.
Putting it All Together
One interesting and beneficial way to use ISM manufacturing data is to subtract the ISM Inventories Index from the ISM New Orders Index. Historically, when New Orders minus Inventories (NO-Inv) is negative, a recession follows. When NO-Inv is at 10 or above, it typically leads to robust economic growth. Of course, this should not be used deterministically. When NO-Inv goes negative, the investor cannot be sure a recession will follow. But that's almost irrelevant because slow and decelerating economic growth will impact the direction of the markets the same as a technical recession. Similarly, when NO-Inv is above 10, it does not matter if GDP grows at 3% or 10% - the impact on markets will be the same. The record for NO-Inv predicting recessions and expansions is evident in the following chart.
The relationship between NO-Inv and GDP growth is verified empirically by the chart above. Clearly, there is something there. But in addition to the empirical evidence supporting the relationship, there is also some economic logic behind it. A rising NO-Inv number is indicative of (1) new orders growing faster than inventories or (2) new orders decelerating at a slower pace than inventories. In the first case, new orders ultimately lead to production of additional inventory to satisfy increased demand. Production is economic growth. We would therefore expect excess growth of new orders above growth in inventories to lead GDP growth. In the second case, inventories can only drop to 0. At some point inventory must be rebuilt to meet existing levels of demand – albeit a decelerating level of demand. Again, we expect the growth rate of new orders in excess of the growth rate in inventories to lead to future production, or GDP.
A declining NO-Inv number is indicative of (1) new orders declining faster than inventories or (2) new orders increasing slower than inventories. In both cases, inventories are building without the demand in new orders to warrant that level. When manufacturers see inventories build in excess of new orders, production of goods will slow or temporarily shut down. Since production is GDP, we can see how a declining NO-Inv number would lead to decelerating GDP growth, or even a recession.
Now let's look at this relationship at a more granular level to see what this relationship looks like currently.
A negative NO-Inv does not bode well for GDP growth. The past five recessions were preceded by a negative NO-Inv. In fact, NO-Inv has only been negative five times since 1970 and all resulted in a recession. What does this mean for the investor? Caution is warranted. Equities would sell off again should a double-dip recession occur. With interest rates so low, the upside in bonds in this environment would be limited. The dollar would likely sell off unless other industrialized nations recessed greater than the US. Hard assets, like gold, silver, oil, other commodities and real estate would likely get punished rather severely as they have been the biggest beneficiaries of the "recovery."
A final word about indicators like this. The logic of NO-Inv leading GDP is sound. It has been empirically verified, although the sample size is rather small. Yet, still, every cycle is different than the previous cycle. Quantitative Easing by the Fed is an unprecedented manipulation of the markets. We've never seen this kind of intervention. We need to be sure that while a double-dip recession is possible, we should not bet the farm on it. Just be alert to what is going on in the economy and be prepared to take the appropriate actions when and if economic weakness is confirmed. Keep track of the NO-Inv. The data comes from the ISM Manufacturing Survey and is released shortly after the end of the month.