Warning:  The following post is offered in the spirit of shameless self-promotion.

Those who prowl the web and consume all manner of economic research probably saw some things late last week that looked eerily familiar.  Maybe you could place them, maybe you couldn’t.  Perhaps you wondered, “Where have I seen that before?”  Well, in at least two cases, you saw it right here at TBP.

The St. Louis Fed published a piece titled A Jump in Consumer Loans? that analyzed the recent “spike” in consumer loans that, upon further review, wasn’t a spike at all:

At first glance, the upward spike in consumer loans during the months of March and April 2010 (see Figure 1) seems to suggest a dramatic expansion in credit. The spike itself could be seen as a strong signal that banks have loosened credit standards or have originated more consumer loans in the wake of an improving economy.  While there have been improvements in consumer loans recently, the dramatic increases over the past few months have been caused by a new reporting requirement issued by the Financial Accounting Standards Board.

They proceeded to show an adjusted chart that took the FASB changes into account:

TBP readers may recall that we showed the following chart on April 26:

(Note:  Difference in appearance due mostly to timeline — I went to 2011, St. Louis Fed only through March 2010 — and a bit to scale — 900 v. 920.)

That consumer loans had miraculously rebounded to a new all time high was, at the time, a badly mistaken talking point that was beginning to gain some traction.

Moving on, we observed after the release of May ISM, and just before May NFP, that there was a relationship between the two that seemed to bode ill for the future of the labor market (see our post here on June 3).  ISM seems to have peaked in April at just over 60 and now appears headed lower.  We looked at the lag between ISM and payrolls and concluded that the payroll situation is bleak because we are not creating jobs rapidly enough in light of the fact that ISM seems to be rolling over.  To illustrate the point, we presented a chart (which I’ve updated through last week’s releases):

Crack economist David Rosenberg presented his readers with the same chart last week, but with considerably more information about the relationship.  (Note that my chart goes back further than Dave’s, but they are identical from 1978 on.)

Said Rosie about this relationship:

ISM leads employment growth with a six-month lead time and with a decent 73% correlation.

The ISM is useful because it is timely, highly cyclical, moves in regular patterns and goes all the way back to 1948!

But here’s the rub. Never before have we hit a peak in ISM with employment growth still negative. That has never happened prior to this post-bubble experience. What is normal, and there are 15 ISM cycles over the past 62 years, is that job growth is positive at the ISM peak, and that peak, more often than not, is around 60 as was the case this time around in April.

On average, at the peak in ISM, year-over-year payroll growth is running at 3% and has another six months to go in terms of acceleration before the trend reverses. The median is 2.5%, as is the mode. And, the range is +1.0-5.3%. This time around, the pace was -1.1%. This means that not only will we never get back to the old pre-recession highs in employment, but that the jobless rate is going to grind ever higher in coming quarters and, in turn, that means so long as the laws of supply and demand are still relevant as far as the labour market is concerned, wages move from disinflation towards outright deflation.

So there you have it –  a twofer of “read it here firsts.”

TBP — worth every dime you pay for it.

Category: Data Analysis, Economy, Employment, Research

Please use the comments to demonstrate your own ignorance, unfamiliarity with empirical data and lack of respect for scientific knowledge. Be sure to create straw men and argue against things I have neither said nor implied. If you could repeat previously discredited memes or steer the conversation into irrelevant, off topic discussions, it would be appreciated. Lastly, kindly forgo all civility in your discourse . . . you are, after all, anonymous.

20 Responses to “TBP “Déjà Vu All Over Again” Post”

  1. Nicely done, Invictus.

  2. yes folks, TBP, where trends are born

  3. hgordon says:

    I wonder to what degree the ISM manufacturing index reflects offshore sourcing of components that go into final assembly of manufactured goods.

    At least in my part of the product development world, there seems to be an accelerating trend toward offshore sourcing of subassemblies and even of design services, so while there would still be a correlation between ISM index and employment, some decoupling is taking place.

  4. Nick Abe says:

    Nice post and all… but the one conclusion that you both jump to is that the declining ISM means we are going to see more job losses in the future. This is not necessarily true, as we are missing some key data. The YoY employment is on a 6 month lag and we aren’t yet sure if it will break above 0% yet. Further, a decline in the ISM to more normal levels does not necessarily imply jobs won’t be created. In fact, if you just wanted to go purely by the chart you could say only if the ISM drops below 45 (or so) can we expect negative job growth.

    What I’m saying is we could still possibly see employment up something like 1% in October (assuming April was the ISM peak) and then slow down (but remain positive). So the logical conclusion at this point is not necessarily that we’re going to lose more jobs, but that this is going to be a very weak recovery in terms of jobs (at best given the data and at worst your scenario will be correct).

  5. vipasyana says:

    ECRI’s Achuthan mentions this here in the video (employment problem): It means chronic high employment and markets can never take off.

  6. kkrathi says:

    Great post. I would like to add that wouldn’t ISM services data be more predictive of shorter term payroll data? By the way isn’t that what is being released this week? The reason I say the ISM services data is more predictive than mfg data is because more than majority of jobs come from the services sector. Agreed the ISM services data is perhaps less reliable than mfg data to provide clues about the economic trend itself. However, because it has such a heavy influence on jobs in US, it may be a better tell for shorter term payroll data trend, no?

  7. DCS says:

    “TBP — worth every dime you pay for it.”

    TBP is so invaluable that I would enthusiastically pay 10x the current price for continued access! :-)

  8. Feel free to hit the Amazon Wish List!

    (I just added a slew of interesting books)

  9. Lamont says:

    I don’t know how employment not rising with the manufacturing ISM has come as a surprise to anyone. This graph should explain it all:


    Manufacturing employment has dropped 35% in over the past 10 or so years, almost all jobs being lost to China and other low cost labor areas. These jobs are not only not coming back, but are continuing to leave the US. Manufacturing employment currently makes up only 11% of total employment. Going into the 1990-1991 recession, manufacturing employment made up over 16.5% of total US employment. During the 1982 recession manufacturing employment made up 20.5% of total US employment. So any increase in manufacturing activity now will result in a substantially lower increase in manufacturing employment. This is why I’ve been paying little attention to the ISMs and production numbers in my analysis of the overall US economy. It’s just too small a piece of the pie currently. Also, the increases in the manufacturing ISM don’t indicate the degree of improvement, just the fact that things have improved from (in this case) extraordinarily low levels.

    IMO, any stimulus will be futile unless the trade deficit is addressed. It can be addressed thru huge incentives for companies to produce in the US (eliminate payroll tax and cut corporate tax dramatically for manufacturers) and to do whatever it takes to reduce our dependence on oil (though cap and trade doesn’t address that; it just punished coal, nat gas, refineries etc, all domestic resources/industries).

  10. hgordon says:

    @Lamont –

    Your post quantifies my basic question about the significance of the ISM index except as a first or second derivative indicator.

    It is my sense as well that the opportunity for long term economic growth will come from investments in the energy sector. In large part, the jobs that will be created are geographically specific and not amenable to export.

  11. shoreb45 says:

    I saw this interesting quote in Rosenberg’s morning piece today as well:

    “The U.S. is now 234 years old and yet over half the nation’s money supply was created since Helicopter Ben took over the flight controls four years ago.”

  12. nades says:

    Nothing wrong with reminding you were ahead of the curve…. Good stuff! :)

  13. Invictus says:


    Two points:

    1) The non-manufacturing ISM has a much shorter history than the manufacturing ISM, so its usefulness is somewhat limited.
    2) It’s the cyclicality of the ISM that makes it more relevant, imo.

  14. bena gyerek says:

    are you trying to suggest that a fed economist has been upstaged by a lowly financial blogger? surely some mistake?

  15. Invictus says:

    @bena gyerek

    Wouldn’t quite say “upstaged.” But in the case of consumer loans, we were clearly first on the scene correcting some misinformation that was leaking into the blogosphere.


    My work, your Wish List? Wassup wit dat?

  16. bear_in_mind says:

    There’s investing and there’s meta trends. I believe what tends to get missed in many of these analysis is subtle yet unmistakable secular downtrend contained in the American macro-economic environment. IMHO, the long-term ISM graphs clearly suggest the deleterious impact of off-shoring in sectors that previously had comprised core middle-class employment opportunities. This race-to-the-bottom of cheaper labor has been coupled with the broad adoption of computers and manufacturing mechanization, amplifying the impact.

    I’ve seen first-hand how outsourcing is rapaciously employed to drive down labor costs and boost profits. The economic damage was conveniently effaced by Easy Al & the Banksters, with the Pied Piper media lackeys selling a McMansion vision of the future. The populace wanted to believe they still retained the social mobility of their parents and grandparents, even if the “debt = prosperity” equation was borne out of complete irrationality. If your wages are stagnant and the CPI is complete bunk (d0n’t get me started on hedonics or ‘silent’ inflation – where you pay the same price and receive 14 percent less in the same package), there’s not a great deal that many can do to change their lot in life. But then banks decide to give away free money so you can ‘buy’ a bigger lifestyle… hmmmm, that’s better than a poke in the eye with a sharp stick , right?! We’ll see about that…

    Bottom line is that we’re in a real pickle. Too many American corporations have elected to put next quarter’s stock price ahead of the stability of communities in 2015… 0r 2025. But it’s going to be pretty tough to find more golden eggs when the flock can no longer eat.

    Turning around the ship? Can do. Get tough, think big, think radically, chart a whole new course. Draft marching orders to reclaim our national independence. Retool to get our country off-the-grid (yes, there’d still be power generation, but it would become secondary only). Place solar and wind generation units on EVERY roof in America by 2020. Increase investments and employment in solar tech, fuel cell / hydro, and commence building major public transit infrastructure in the Top 50 urban markets. And concurrently, increase net efficiency by two percent annually for the next twenty-five years.

    There is no reason why this couldn’t all be done in America — and export the fruits of our labor throughout the globe. We have to demonstrate the value and the will to embrace doing things differently. Unfortunately, it appears too many entrenched interests are keeping the country locked in scapegoating and other meaningless diversions. For now.

    Just curious… anyone seen analysis of what our unemployment would look like if we didn’t have est. 500,000 to 1 million citizens (via armed services, soldiers of fortune, building contractors, defense contractors) deployed in various forms of operations in Iraq and Afghanistan?

  17. d4winds says:

    great blog but not for these examples; the first illustrates the first law of looking at socio-economic data, viz, spikes are anomolies until proved conclusively otherwise (almost never); the second is an interesting correlation curiosity masquerading the prosaic observation that coincident indicators lead lagging ones; having said that in the interest of general skepticism–certainly not of criticism–it’s still a great blog.

  18. Clem Stone says:

    Who says ISM has peaked? I’m not predicting it hasn’t, but look at ’02 on that chart.

  19. Invictus says:

    @Clem Stone

    I have looked at the numbers closely — not just the charts — and can tell you that it’s fairly rare, though admittedly not impossible, for ISM to retake 60 once it starts its descent. Not to say it hasn’t happened, but the odds are against it.

  20. Meteor Blades says:

    kkrathi: Yep on non-manufacturing index, the component employment index of which is in negative territory after an upward blip.

    bear_in_mind: “Can do. Get tough, think big, think radically, chart a whole new course.” Yes, yes, yes. Unfortunately, Congress and the oligarchs: No, no, no.