top of page

US Macro Markets: The ISM Schism

From my read of sentiment (and social media), there is a sense that something doesn't quite stack up... Stock prices are headed higher, yet for the most part the data still sucks.

So what's up with the disconnect? Is the market on borrowed time, or is it on the money?

With the following charts I will help bring some clarity to the issue.

And that's an important point: markets are complex. I find half of the job in generating meaningful insight is simply to bring clarity where others bring complexity.

Sometimes that comes from a single chart, and often times it's the synthesis of the right combination of charts. Today's combination of charts both examine the issue and illuminate the outlook by squaring up the disconnect.

On with the charts...

1. S&P500 vs the ISM Manufacturing PMI - a Slight Disconnect

The first chart shows how a slight disconnect has opened up in the rolling Year on Year percent change in the S&P500 vs the level of the ISM manufacturing PMI. It's certainly not the first time a gap has opened up, and there's probably 2 main points I would highlight. First, the fact that the ISM moves so closely with the S&P500 says a lot about the sentiment component within that indicator (which after all is a survey of humans, and humans are subject to swings in sentiment). Second, though it's hard to make a definitive statement, there is a slight tendency for the S&P500 to lead the ISM. So we could conclude by saying the stockmarket is just leading the ISM, but it's hardly a satisfying conclusion, especially if just based on the chart below.

S&P500 vs ISM manufacturing PMI

What to watch for: the initial rebound in the manufacturing PMI adds weight to the idea that equities are just leading the ISM, so look for follow-through in Nov/Dec (or lack thereof!). Also watch out if the disconnect widens to an extreme.

2. US Cyclicals vs Defensives and the ISM Manufacturing PMI

If the previous disconnect was slight, then the disconnect in this chart is substantial. The chart shows in the red line US cyclical vs defensive sectors relative performance (which I have de-trended because there was a clear trending tendency in the data -- this makes it more comparable with the ISM, which is more of a stationary series). This is overlayed against the ISM manufacturing PMI, and the key point is that there is a wide gap. The problem with this one is that the conclusion is in the eye of the beholder: bears will say it represents a huge risk as cyclicals will need to fall to close the gap - putting downward pressure on the market as a whole; bulls will say the market is right and the ISM will rebound to close the gap...

US cyclicals vs defensives and the ISM

What to watch for: the current situation is virtually unprecedented, but I would note that whenever the ISM has become extremely disconnected in the past it's more often the ISM moves to close the gap e.g. 04/05, 09, and slightly less so in 01/02. The initial move in the ISM then is promising for the bulls, but again, we'll need to see follow-through.

3. Another Disconnect - Credit Managers vs Purchasing Managers

This is a chart I put together a while ago that has just been siting quietly in the corner of one of my excel files (perhaps just waiting for this moment!). It shows the NACM (National Association of Credit Managers) CMI (Credit Managers Index) against the combined ISM (Institute of Supply Management) PMIs (Purchasing Managers' Index). In the past, for the most part the two surveys have walked a fairly similar path, but that changed in the last two years. I would argue that the PMI showed excess optimism in 17/18, but now shows excess pessimism: Purchasing Managers say economy is headed for recession ...yet for better or worse, Credit Managers just aren't seeing it.


What to watch for: probably the most important thing to watch for is any signs that the CMI begins to "catch DOWN" to the PMI (vs my base case of the PMI catching UP to the CMI).

4. Reconciliation Theory: Looking Forward

I could have just left it there, and thrown in an open question and left it to the reader to figure out, but I will be a little more generous than that and let you see one of my power-charts. This one tracks the rolling annual change in mortgage servicing (effective mortgage rate x median house price), which seems to lead the annual change in the ISM by about a year. The economic logic is that lower mortgage rates (and lower financing costs more broadly) help boost the housing market and consumer discretionary incomes; ultimately flowing through to the more cyclically sensitive manufacturing sector. Now it is important to acknowledge that the relationship is not perfect (and doesn't always work), but the punchline is this: if you take this chart literally we'll be looking at an ISM about 60 this time next year. That would close the gap.

What to watch for: really the main thing to watch is that the historical relationship holds i.e. that the ISM does start to inflect as the chart suggests (and tentatively it is doing so). Personally I'm also paying close attention to the key transmission channels of this indicator i.e. refi activity, etc.

Final Thoughts and Bottom Line

Bottom line: my view is we are already at an inflection point in the macro pulse and the disconnects with the market and other surveys are entirely consistent with this view and thus the ISM will probably do much of the heavy lifting to close the gaps.

Finally though, I would also note at least in passing, that the global manufacturing/export sector has looked very weak this year, and it's fairly easy to make the case that we are basically in the middle of a global manufacturing recession. That is, the softness in the global (and US) manufacturing PMIs is not purely a sentiment effect. But the thing is, hard data e.g. export volumes is by definition a backwards looking and lagged indicator, meanwhile at best the PMIs tend to be more of a coincident indicator in practice. Yet when it comes to markets, investors get paid to look forward, not backwards (even if most of the news/noise is focused on the backwards looking data). And while the backwards looking view is not great, the forward looking view is arguably a lot better.

For more detail on this this topic or any other specific questions get in touch...

You can contact us direct or by social media.


Callum Thomas

Head of Research at Topdown Charts Limited

Recent Posts

See All
bottom of page