Sentiment surveys such as the ISM’s Manufacturing Purchasing Managers Index are not strictly about current levels of production. Even if they were, they still wouldn’t be as straightforward as is presented. Rather, the ISM index or any PMI for that matter is an amalgam of variables ostensibly displaying how economic agents feel these variables are affecting them in any given month.

Harder still is interpreting what any of this might mean. One big piece of any PMI is prices. Rising prices both paid and received are usually taken as a sign of better fortunes ahead, an impression left on us by decades of orthodox Economics. This may be true under certain circumstances such as during a recovery.

If prices fall sharply consistent with a severe downturn, it stands to reason that if they rebound the economy might be doing the same thing. Or, at least commodity price investors could believe as much. Thus, in the case of PMI’s, purchasing managers may think that what commodity investors think is valid.

The interpretative process is several steps removed from what PMI’s are supposed to represent, in part. Over the last decade plus it is one more still, meaning that managers might be seeing rising prices and feeling good about them (more or less) because they believe commodity investors are betting on opportunity without factoring why. If it’s straight economic recovery, OK then. But what if it’s nothing but expectations about the effectiveness of something like QE?

It ends up as something like purchasing managers looking favorably upon commodity investors who are enthralled by central bankers. This explains quite a lot, actually.

False confirmations would then lead to false dawns, several, as Janet Yellen appropriately put it in 2014.

In November 2003, the ISM Manufacturing Index rose above 58 for the first time since the mid-nineties. It was taken in the same way the same result always was. After a mild recession in 2001, the economy failed to recover much in 2002 and most of 2003 (the first of what wasn’t supposed to be normalizing “jobless recoveries”). The sharp rise in manufacturing sentiment seemed to indicate that the puzzling drag on growth (this giant sucking sound) had finally released the industrial side for resumed expansion.

That is, in fact, what had happened. For the next almost three years, while not overly robust US-based manufacturing did return to the boom part of the normal business cycle. The ISM as in 1994 seemed to be validated by history.

There wasn’t any real deflation in the dot-com recession, meaning that what moved the overall index wasn’t much more than production variables. This wasn’t true, however, in the next “cycle” that followed.

The Great “Recession” smashed any number of things, including commodities. While being better than the bottom is an unqualified good thing, it gets murky interpreting what that exactly means. Dead cats don’t just bounce in the stock market.

The ISM reached 58 again as early as March 2010. It would then stumble as renewed monetary difficulties erupted several months later and would remain a prominent feature of the global landscape immediately thereafter (why QE2 in November 2010, after all?) Still, commodity prices would keep going as would recovery sentiment.

In February 2011, the index was just below 60, 59.9, for the first time since 2004. Like seven years before, this indicative “boom” in manufacturing was believed to be one in fact and not just sentiment. The near 60 for the ISM was expected to be the start of a lengthy and profound expansion as it nearly every time before in its long history.

The same month copper prices registered their peak and oil was near enough to its ultimate top, too.

It went like that for a year before the 2012 slowdown and near recession. For the first time in decades, the ISM at 60 was nearer the top of an upswing than the start of a boom. “Something” was obviously different.

To confuse the issue further, or rather to clarify better, at least to those with honest curiosity about the whole economic system, the same process would repeat just a few years later. Though the ISM index didn’t quite reach 58 in 2014 it did come pretty close. The difference between that peak and 2011’s was the absence of inflationary reflation, or commodity prices. They were still on their post-2011 downswing (deflation) in 2014.

It may have seemed like less enthusiasm but in the end it didn’t really matter. But like 2011 anyway, the near 58 in 2014 proved once more to be the top rather than the presumed beginning – even though, quite importantly, manufacturing in the US was still significantly below its prior peak by then seven years in the past.

The peak ISM was met only by a true manufacturing recession here (revealed better only by benchmark revisions taking away positive data bias) and near manufacturing catastrophe around the world. It unleashed big deflationary forces that in many places, especially commodities, ended up in similar proportions to what had happened during the Great “Recession.”

That meant in terms of the rebound or reflation following the 2015-16 downturn, there would again like 2010 and 2011 be an inflationary feel to it as commodity prices came off of a low trough. How they have been interpreted in 2017 and 2018 explains a lot.

In mainstream convention, the ISM Manufacturing having reached 60 several times over the past ten months, and being above 58 in all but one of those ten, it is being viewed in the same way as it was in late 2003 – the start of actual expansion. Why?

Recent past experience and an overwhelming catalog of parallels suggest far closer to early 2011, or the middle of 2014, as the top of a mild upswing. But having learned nothing from those last two swings, the boom is instead grounded in the confirmation bias of circular logic.

The Federal Reserve’s Industrial Production series, the latest estimates released today, includes several diffusion indexes that are similar to PMI’s like the ISM version. They are simpler, however, in that they only deal with production levels rather than include other factors like employment, backlogs, and the sometimes dramatic movements of prices.

The IP diffusion estimates are also meaningfully different than the ISM in recent months, just as they were six years ago in 2011. There was much more consistency in 2014 when inflationary reflation wasn’t too much of an issue.

I think what we are seeing is the mis-interpreting of commodity prices on both sides of them. Investors are anticipating globally synchronized growth this time around bidding up everything from copper to oil and whatever is in between. This pervasive belief is indirectly related to US QE at least in the “rate hikes” US central bankers are attempting to use as confirmation of their views of its success (circular: QE worked therefore we have to raise rates; we have to raise rates therefore QE worked).

That view is seeming bolstered by positive views of other monetary polices elsewhere in the world – primarily Europe and the ECB as well as China and the PBOC. There is this feeling that at long last these policies are working and finally working together. Data is scarce.

It comes out as a bit of serious confirmation bias. Prices rise boosting PMI’s and “sentiment” beyond where they might be without the prices changes (2014 vs. 2017-18) as these presumably inflationary results appear to be consistent with what should happen if recovery was at hand. The ISM is back at 60 because of the influence of oil and copper which is then taken as a sign that oil and copper are correctly interpreting future probabilities.

This is how, like 2011, people miss the underlying deflationary signals and warnings. Just because commodity prices are moving higher it doesn’t necessarily mean inflation. Besides, there should be much more obviousness to the upswing in terms of broad economic processes beyond the overamplified hope of commodity investors betting rather meekly that central banks might finally get something right for once.

In terms of US Industrial Production, there are any number of big time disagreements when there should be none. A real inflationary upswing isn’t perfectly harmonious, but it doesn’t feature diametric opposites in crucial components, either. The energy sector is truly booming again, but that, too, is based on the swings of oil rather than broad economy.

Opposing crude is now the auto sector, unlike before 2015 when they were rising together. Motor Vehicle Assemblies rebounded from a disastrous month in April but only to what is more like 2017 than 2014.

Worse for the ISM, the struggles in the auto sector are more consistent with related factors like income. Lower not gaining incomes lead to lower not gaining sales which forces production cuts not acceleration.

In short, there is way too much 2011 in these 2018 ISM sentiments and very little of 2004 or 1994. The difference in terms of interpretation is profound. The latter would signal, as is written in the mainstream, the start of an actual boom. The former has signaled, twice more recently, the end of a small upswing. With commodities now rolling lower and curves distorting and even inverting, which one might be the more likely interpretation?