Behind the Misperception Veil: What's that Data Behind the Curtain ?
It occurs to me that if all you're able to do is read the headlines the world has to be a tad confusing right now. It takes some time and effort to dig into the various economic, market and business indicators and make some sense of them. Our plan for this weekend as the markets reacted wildly and positively to mildly "good" economic news was to focus on what the actual data said instead of what the headlines made it say. And then Friday arrived....whee. Was that fun or what ? Now as it happens it was personally gratifying in the sense of being consistent with our many months held stance. It was actually puzzling and non-gratifying in other ways because, when you dig into things, Friday's payroll numbers were part of a consistent set of trends and patterns that have been building for months now. And are consistent with many...many data points. So we're going to circle round back to the plan and review some of the weeks economic indicators.
But this headline problem deserves some thought. Barry Ritholz inadvertently highlight's the dilemma with his Sun. morning post pointing to a Floyd Norris (chief financial correspondent NYT) post on the YoY change in unadjusted private sector jobs, which was -125K, as a major recession indicator. Actually that's true....but YoY% change works as well, is consistent with seasonally adjusted data on either a quarterly or monthly basis; and all those data setii are entirely consistent with the trends and patterns of the others we're going to look at. At the bottom of Barry's post is another pointer to James Pethokoukas who has a similar job at US News and also blogs away. Hallucinogenically according to Barry's claims since he's in the "Dude, Where's My Recession ?" camp. Actually he's in the real optimist wing of commentators arguing that the economy is accelerating albeit slowly.
Having established, respected, well-positioned observers make such contradictory observations creates enormous cognitive dissonance for the casual observer. On the other hand it does make a horse race. Your problem is to sort it out. And that's where we try to make our modest contributions: by making comments that are grounded in the data, in the real structure and patterns, and doing it in such a way that the reasoning is clear, easily observable and verifiable. In other words we try to provide you with an easily decoded "dashboard" of indicators who's accuracy you can check for yourselves, either by inspection since we report the data just as it is in context. Or by replicating our approach.
When you do that for the data series we looked at this week here's what you'll find (charts and more detailed discussion after the break). We've run across a few other bloggers btw who take similar approaches. Beyond BigPicture and CalculatedRisk - the big gotos - we highly recommend BeYourOwnEconomist.
1. Durables goods orders and PMI/ISM were reported as growing MtM, which they did. And both have been showing an anomalous uptrend in the very short-run which turns out to be export demand. Both have now turned back down and never violated a longer-term downtrend in any case.
2. Hours, Pay and Unit Labor Costs are all showing downturns consistent with the early stages of the beginnings of a recession. And there do not appear to be any anomalies either. In fact on a quarterly YoY% basis Hours are decelerating much more rapidly than Employment ! Can you spell canary.
3.Initial vs Continuing Claims are both showing continuing downward declines, entirely consistent with downward trends. Though that trend is accelerating. Again the headline was about how Initial Claims dropped but even restricting our view to it there's no trend violation. Continuing Claims though is particularly interesting since on an absolute numbers basis it's patterns were almost completely synchronous with initial claims until this weak and poor recovery. NOW continuing claims have remained at an anomalous high. Really stop and think about the implications of that for a minute, please. Do the words "boiling frog" mean anything ? And it's been going on since '02. That really doesn't bode well for the long-term outlook at all, does it ?
4. Employment, Hours and Unemployment: it was Unemployment that scared the markets so badly. As in a way it should since YoY it jumped from an 11% increase last month to approx. 25% this month ! But when you look at the data and the charts again the patterns are consistent, continuing and no surprises. That is, if you believe the downtrend acceleration thesis we're arguing.
We haven't put in another of our favorite and more often used HF indicators, Real Retail Sales, since we updated the set of HF indicators last week. Though the headline reports of same-store sales were so "good" this week and also supported earlier in the week market surges. But we did use it to kick off the discussion of the Retail Sector (Retail Industry: Plus Ca Change...or Bend Over and Kiss...) and pointed out that it too is accelerating downward, has been for some time and the rate is picking up. Again we'd ask you to link back, at least in your minds, to that detailed dive on Retail, and earlier on Autos, for investment, employment and outlook implications. This all keeps coming full circle indeed.
Durable Goods and NAPM
Two relatively important series that have come out recently are the durable goods orders and the NAPM indexes for purchasing activity in the old ("manufacturing centered") and new ("services") economies. Now one other meme we need to disabuse you of is the notion that there's a wide divergence been old-line and services sectors. Anyway the top chart shows new durable goods orders, with and without aircraft, compared to industrial production. Both were reported as showing MtM increases but when you look at the patterns and trends you do see two things. First the anomaly we've been puzzling over with the recent sharp rise. Now we've figured out that exports, as various detailed numbers and analysts have pointed out, are due to exports. And both are now in YoY decline - consistent with our "the world is slowing too" thesis and re-coupled world economy argument. As you can see looking at PMI (the new lable should be NAPM after they changed their name but...) for oldline and ISM for the services sector. Again you see a nearly identical pattern but no violation of trend. At least so far.
Hours, Pay and Labor Costs
Another interesting set of data that in some ways might/ought to get more attention occasionally is hours worked. Along with pay and unit labor costs. As you can all three on our favor YoY% basis are synchronous and following similar patterns, e.g. they all show the last recession with slight differences in timing. Just to put it back in context with our regular quarterly payroll and GDP analysis the bottom chart compares Hours, GDP and Employment. Notice that while GDP is slowing but definitely NOT in a recession as yet that Employment has been slowing at an increasing rate and Hours even more so. Circling back to the top ask yourself would Pay and ULC also be showing downturns if a) the employment market was tight ? And if b) inflation was about to be a structural problem ? That's one bet, based on deeply informed analysis, that the Fed is winning.
Initial vs Continuing Jobless Claims
The data series that really got all the koolaid drinkers excited this last week, and led to the big run in the markets supposedly on Th., was initial unemployment claims. They did go down on a weekly basis. But when you look at the timing and patterns, again, you see them rising rather rapidly on both an absolute numbers and YoY% basis. BtW, a small whine :). This is weekly data and it took a fair amount of effort to convert it to monthly and quarterly data so you won't find this indicator presented in many other places; none that we are aware of ! The really interesting thing, to us, was that on this much longer time frame initial and continuing claims are nearly identical on a YoY basis BUT....on a numbers basis there's a clear and fascinating anomaly where Continuing Claims have remained MUCH higher than initial claims during the last several years. This is consistent with our long-held views and arguments that this has been a very...very weak recovery where growth never achieved organic takeoff. It also explains the sense of malaise you find among voters that's built up so rapidly. Yet another Boiling Frog syndrome.
Employment, Hours and Unemployment
Now to the baxx-buster...the one that cause the markets to panic as the veils of cognitive dissonance got swept away a little, though not completely. Nor, as we believe, consistently in the sense of there aren't any real surprises here if you've had the right filters and dashboards in place. The top chart is quarterly YoY% changes in Employment, the Hours Worked Index and Unemployment (on a reverse scale note). The bottom is monthly for Employment and Unemployment. Unemployment jumped big time from 11% to ~24% YoY ! OUCH. But again what would appear to us to be right along the trend. Which again, is a downtrend that's accelerating.