Friday, January 9, 2009

Jobs are a lagging indicator (2)

Here's a better version of the chart in my previous post. I've marked the high and low points for both equities and jobs over the past 10 years. The equity market turned down about 7 months before we saw the job losses that marked the 2001 recession. Equities then turned up about 7 months before jobs began to grow in late 2003. This past year things have unfolded much quicker, and the lag between equities and jobs turning down was only about 3 months.

Equities and jobs have fallen faster this time around, and the economy certainly looks weaker than it did in 2001. But if the lag between the bottom in equities (Nov. 20th) and improvement in jobs is again 7 months (and if Nov. was indeed the bottom in equities), then we would expect to see jobs hit bottom in the next few months and begin to turn up by mid-year. Even if jobs fell another 1-2 million in the meantime, that would still leave us short of a depression. Most recessions are policy-induced, with extremely tight monetary policy almost always the driving force behind the downturn. This recession has been essentially fear-induced; the fear of subprime-related losses, the fear of an imploding banking system, and fear of badly-managed government bailouts. It's happened all very fast, so it could easily end just as fast.

This is all highly speculative, of course, there being way too many variables to forecast with confidence. I post it mainly to show that financial markets can and do lead changes in the real economy. With significant improvement evident in key measures of the health of the corporate bond and stock markets over the past several months, with pessimism rampant, with deflation fear high despite extremely easy monetary policy, and with corporate bond and commodity prices rebounding, I don't think it is unreasonable to think that the economy is in the early stages of bottoming, and that a recovery is possible my mid-year.


Cabodog said...

Scott, thanks so much for your excellent commentary and graphing.

Interesting that as a country, even after the job losses of the last twelve months, we're still up a couple million jobs over the peak in 2001 and about four million from the recession trough in 2002.

I suspect one reason that we're coming down in employment so quickly is the result of the rapid growth of employment from 2002 to 2008.

It's tough to see through the panic and bad media these days. Thanks for your excellent work.

Unknown said...

Though I am somewhat heartened by your analysis, I must admit to being spooked by what appears to be an increasingly volatile economy in general.

Such rapid change on so many fronts, and the accompanying loss of a sense of predictability that results, is for me, one of the most worrisome things revealed by your chart.

brodero said...

I noticed that you pick the point of 134 million for the bottom in job losses is that because the 81 and 74 recessions resulted in a 3% drop in total nonfarm payroll employment.

Scott Grannis said...

Believe it or not, that is pure coincidence. I was just trying to estimate what a reasonable level was for job losses over the next 6 months.

If it turns out that job losses in this recession are no more than 3%, then without doubt the market is priced to an extremely pessimistic view of the economy's fortunes.

Scott Grannis said...

sam: I understand your concerns, believe me. I too worry about all the volatility and policy uncertainty. But I think this is already priced into the market. If I'm right and if Obama manages to avoid doing really stupid things (consider all the ill-considered promises that he has already reneged on), then there is a decent chance that the future will be less terrible than the market expects.

brodero said...

This is not related to the graph but could you comment on where you think we are in the deleveraging cycle and do you think a study of total margin outstanding could be a
decent proxy for measuring deleveraging

Chris said...

Speculative or not, I tend to agree with your clarified comments on the lag between job losses and the sell-off in equities. The precipitous onset of this crisis could very well point to a dramatic recovery.

As you've shown, this downturn is supported by fundamentals, yet the drop-off appears to have surpassed levels you might expect to see given these fundamentals. I tend to agree that a collective mind-set, i.e. market psychology has added some "margin" in this decline.

I'm hopeful we'll see a change in thinking in 2009 and a speedy return to health. A substantial and regular dose of Scott's market optimism in place of the routine media gloom might get us there sooner.

Scott Grannis said...

brodero: I can only guess as to where we are in the deleveraging cycle. I think it's quite advanced. Most importantly, I think most of the forced selling/deleveraging has occurred. I base that guess on the higher prices we are seeing for almost all of the things that were the object of leverage's affection (e.g., commodities, stocks, corporate bonds, the ABX and CMBX securities). It would be interesting to see the status of margin debt.

brodero said...

Thanks for your response

atempe said...


Did you get a chance to go back and look at the 1990 recession to see if the same 7 month lag between the stock market bottom and job market bottom was in play then as well? Do you recall if the unemployement rate exceeded 8% during that downturn? It seems to me that it did. Thanks.

Scott Grannis said...

Well, economics is not a science. Turns out that the stock market hit bottom around the same time that jobs hit bottom. As I said in the post, financial markets "can and do" lead the real economy. But the 1990-91 recession is one of the exceptions. What I could say in favor of my assertion is that stocks turned up well before the official end of the recession was announced, and well before the public thought the economy was once again growing. Recall the 1992 election, when everyone thought the economy was still in recession. In fact, the economy had been growing pretty nicely for over a year.