(Bloomberg View) -- Many market pundits point to the Institute for Supply Management manufacturing purchasing managers' index as a reason for caution. Growth is just too good, so it can't last, or so the thinking goes. The ISM stands at a historically high level, about a full standard deviation above its long-run historical average. That means economic conditions must deteriorate soon, right? Maybe not.
The ISM manufacturing PMI is easy to follow because of its simple construction. The survey asks about 300 purchasing managers whether activity is up, down, or the same relative to the prior month. A level above 50 is consistent with factory expansion; a level below 50 signals a contraction. The index is timely, released on the first business day of every month, and upon casual inspection, lines up reasonably well with the broad ups and downs of the economy.
That said, if the ISM is about to drop because it is already so elevated, it’s worth highlighting that the index is quite error-prone, sending more false signals than correct ones. Let’s assume the ISM indicates a turning point when it registers a value below 50 on average over a three-month period. Lining up this series against U.S. recession periods shows that ISM is often wrong. The index is as likely to be early in signaling a peak as it is late in signaling a trough. If anything, false signals have become more frequent in recent business cycles, perhaps owing to the factory sector's declining share of U.S. gross domestic product.
Despite its flaws, there is no shortage of investment strategies based on the ISM manufacturing index that circulate across Wall Street. The “ISM investment clock” is a standard research piece, devising simple rules for investors based on different levels of the ISM. Over the years, we’ve heard many such rules of thumb: Three months below 50, either a recession is near or a Fed rate cut is on the way, and so investors should pile into bond surrogates in the equity market, such as utilities. Another popular one is that once the ISM levels off, investors would do well to rotate away from industrial stocks and industrial commodities.
Of course, no two cycles are the same, and these rules tend to assume that relationships are stable from one cycle to the next. For example, the ISM peaked in May 2004 and fell steadily into 2007. Yet during this time, industrial stocks generally outperformed the benchmark index. By contrast, in recent years, the ISM manufacturing PMI has tracked the relative performance of industrials reasonably well. Still, we’d hardly describe the relationship as seamless. With markets, as with the economy, the ISM is as likely to tell you what will happen as it is what has happened.
We can analyze this in a more systematic way. First, we isolate the local peaks and troughs for the ISM manufacturing index. Second, we devise a trading strategy where we short industrial stocks relative to the market at the peak and hold the short until we reach a trough. Likewise, we go long industrials relative to the market at a local ISM trough and hold until we reach a local peak. This trading strategy has underperformed a traditional buy-and-hold strategy back to the early 1990s. Sometimes, our ISM-based trading strategy beats the market and other times, it disappoints -- hardly an indicator to hang your hat on.
In short, an elevated ISM is not reliable evidence that growth will slow. That the index remains comfortably in expansion territory should be taken as a positive, not an indication of impending weakness. The factory sector in the U.S. is expanding. That’s good. Moreover, playing for a peak ISM makes inconsistent assumptions about investors’ forward-looking behavior. After all, if the ISM is known to be a mean-reverting series, won’t investors already assume an increasing likelihood of declines the higher the index moves?
What matters for markets is not the direction, but the direction of the index relative to expectations. Finally, the ISM’s usefulness as a proxy for the economy is oversold. The monthly employment report is far more useful and released almost as early as the ISM. The ISM is useful, but no substitute for hard measures of economic activity.
This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.
Neil Dutta is the head of economics at Renaissance Macro Research, responsible for analyzing global trends and cross-market investment themes.
For more columns from Bloomberg View, visit http://www.bloomberg.com/view.
©2017 Bloomberg L.P.