More or Less
by Brian Pretti CFA, Contrary Investor. September 18, 2009
You already know that a number of headline economists/strategists are starting to “talk up” and anticipate a very meaningful economic recovery bounce ahead based on the fact that the present economic downturn has been so deep in terms of historical context – the old slingshot theory, so to speak. The first table below gives you a little glimpse of historical context. The harder they fall the more violent the upside recovery? Perhaps. Yes indeed, the very deep recessions ending in 1958 and 1975 were followed by on average 7% real GDP growth in the following twelve months post recession end. Will that be the case again in the current cycle? We certainly all wish we knew with certainty.
What is important to investment decision making near term is not necessarily the ultimate fundamental reality of what is to come in terms of strength of economic recovery immediately ahead, but rather investor’s perceptions of what is to come. It’s these collective perceptions that will form consensus and move financial asset prices short to intermediate term. That and a tsunami of liquidity, of course. We know the government is trying desperately to positively kick start two key fundamental drivers of historical economic recovery that are pent up demand for autos and housing. For now, the government has made some positive short-term impact. Sustainability is the bigger and longer-term issue that is a major question mark. But the other two intertwined drivers/characteristics of prior historical cycle economic recovery are currently missing in action. And those are growth in personal income and acceleration in household debt assumption (debt used in good part for consumption purposes). Again, regardless of ultimate fundamental outcomes that will only be told well after investors anticipate that reality, it is gauging investor perceptions that will be important in the months ahead. So what is it that investors may be watching in order to make an educated guess about potential strength of economic recovery to come? Today I’ll take a look at one of those issues, and that is unemployment claims.
The data for the history of initial unemployment claims only travels back to the late 1960’s, so let’s start with a quick look at the big post recession economic recovery snapbacks since that time. Highlighted below are the two largest post recession twelve-month real GDP gains that were seen after the recession conclusions in 1975 and 1982. In very sharp contrast, the post recession initial GDP lift-off experiences in 1991 and 2001 were very shallow.
|History Of NBER Recessions/Initial Expansions|
|Official Recession Period||Real GDP During Recession||Real GDP Growth In 12 Mos. Following Recession|
|12/69 - 11/70||(0.6)||4.5|
|11/73 - 3/75||(3.2)||6.2|
|1/80 - 7/80||(2.2)||4.4|
|7/81 - 11/82||(2.6)||5.6|
|7/90 - 3/91||(1.4)||2.6|
|3/01 - 11/01||(0.3)||2.3|
|12/07 - 2Q09||(3.9)||?|
What is the character difference between these two contrasting post recession outcomes? Certainly post 1975 and post 1982 were the very definitions of “V” recoveries. Well, it turns out that one key differential was the trajectory and longevity of declines in unemployment claims. From a longer-term perspective, you can see exactly this in the chart of initial unemployment claims below. In the post recession 1975 and 1982 periods, once initial unemployment claims peaked for each cycle, they fell dramatically in percentage and nominal body count terms, and fell for elongated periods of time. In contrast, the drop in unemployment claims in the post 1991 and 2001 periods were initially much smaller in percentage terms and the declines in new claims truncated quickly and for an extended period of time before dropping much lower in each cycle as economic recoveries really kicked into gear literally years after each official recession conclusion. As you know full well, in the clarity of hindsight the post 1991 and 2001 recoveries have come to be characterized as jobless recoveries.
Given well-placed concern over current labor market conditions, importantly and necessarily inclusive of the character of wages, I believe this is a meaningful analytical exercise to walk through at the moment. Although there is no guarantee that history will repeat in exact fashion, watching the rhythm and character of claims data ahead can hopefully give us a “heads up” as to whether we may have a strong snapback headline GDP recovery and whether or not the hoped for recovery will be “jobless”, or otherwise. Important why? Again, investor perceptions. Since the reality of US economic recovery character or potential failure (double dip) will not be known with any certainty for quite some time, it’s anecdotal data such as we find with the unemployment claims that the investment community will be watching and responding to. The historical experience linking claims rhythm to macro economic rhythm we’re talking about is no secret or mystery.
So in the spirit of looking ahead and trying to develop some benchmark against which we can anticipate potential macro economic outcomes, let’s put some quantitative numbers to the analytical exercise we will be walking through ahead. The charts will tell the stories for us. Below is initial unemployment claims experience in the post 1975 and post 1982 periods. You can see exactly what I’m talking about in terms of longevity and magnitude. Let’s develop some benchmark numbers. In 1975, initial unemployment claims peaked in the third week of January. The recession officially ended in March of that year. The peak to initial trough drop in claims occurred over a 52 week period and claims declined by 42%. In 1982, claims peaked in the first week of October as the recession at the time officially ended in November. Claims dropped 52% over the following 72-week period.
In pretty dramatic contrast, as is seen below, in 1991 unemployment claims peaked in the third week of March, which happened to be the concurrent month in which the recession officially came to an end. But the decline in claims initially troughed 17 weeks later with a near 20% peak to trough decline - very shallow relative to post 1975 and 1982. 2001 experience is a bit of a different animal in that the 9/11 incident precipitated the pop in claims you see in the bottom clip of the chart below. Twelve weeks later claims fell back to where they were prior to the 9/11 event. In reality if we eliminate the post 9/11 shock to claims, initial unemployment claims were really stagnant in 2001 and 2002, holding at roughly 400,000 per week for a few years. In hindsight, claims peaked in September of 2001 (9/11) and the recession officially ended in November. Claims really did not start to subside in that cycle until 2003, as job growth finally went positive on a monthly basis that year, a near full two years after the official recession conclusion - the very definition of a jobless recovery.
The following table simply summarizes all of the data mentioned. Hopefully we can all use it as a benchmarking mechanism as we move ahead. The message of history is clear. Significant economic rebounds have been accompanied by big and elongated time period drops in initial unemployment claims activity. Jobless, or shallow, economic recoveries witness shallow declines in claims that are truncated within 3-4 months of peaks.
|Initial Unemployment Claims Experience In Post Recession Environments|
|Recession Ends||Month Claims Peak||Time From Peak In Claims To Recession End||Real GDP Growth 12 Mos. After Recession End||% Drop In Claims To Initial Trough In Claims||Weeks From Peak To Initial Trough In Claims|
|3/1975||January||2 Months||6.2%||42.1%||52 weeks|
As of right now, initial unemployment claims in the current cycle peaked in the last week of March of this year. So far the initial trough, if it is to hold, occurred in the second week of July. The decline so far has lasted 15 weeks and claims dropped 22.3% over this period of time. As you know, the reconciliation cycle is still young, so it's too early to try to draw hard and fast conclusions, and that's not really my purpose. THE issue is to watch the character of claims ahead. If we break sustainably to new lows below 524,000 and continue south, we all have to be open to the historical message of this data. Alternatively, it may be that the numbers are already telling us a shallow and jobless characterized recovery is in the cards for the current cycle. Let's hope this data can truly help us gauge fundamental economic character well before the clarity of hindsight illuminates reality.
Before leaving this subject, one last piece of perspective regarding the history of initial unemployment claims. As you know, all the data looked at above was presented in nominal terms. It’s the history of absolute numbers. But it's also important to look at one more relative view of life in terms claims data. The following combo chart looks at monthly data only. The top clip is unemployment claims in absolute terms. But the bottom clip is claims data relative to the total payroll employment base. This gives us a very good sense of magnitude in terms of claims relative to the totality of the working population. Although monthly claims in absolute terms went to new all time highs this cycle, claims relative to the total payroll employment base look much less bad. In fact, you can see that in terms of claims relative to the total payroll employment base, the recent peak in claims looks much closer to what was seen in the 1991 and 2001 experiences – jobless recovery economic environments.
At least as of now, the numbers above tell us to work under the "jobless recovery" scenario when anticipating both economic and financial market themes and outcomes ahead. I’ll leave you with one last tangential comment to ponder that is really fodder for another discussion. I’ve seen many a Street “seer” these days become convinced a “jobless recovery” does indeed lie ahead. But what seems striking is the complacency with which this conclusion is being drawn and used to support investment conclusions. As I see it, the “jobless recovery” post the 2001 recession had one key characteristic – an incredible expansion in household leverage. It was this almost maniacal leverage explosion that both compensated for lack of job growth and drove the economic recovery itself. So if we look ahead and assume a jobless outcome in current post recession experience, will households lever up again to compensate for lack of jobs and wage growth? Not a chance. Not this time. It’s just a good thing the government will do it for them, right? That is a good thing, isn’t it?
© 2009 Brian Pretti