Friday, February 21, 2014

The Insured Unemployment Rate Volatility Pump: Unemployment Rate About to Plummet

The following shows the headline unemployment rate against the insured unemployment rate derived from continuing state unemployment insurance claims for the last three business cycle: 1990 to 1994, 2000 to 2004, and 2006 to 2014. The picture provides relevant and key information, not based upon significant statistical analysis but common sense. Immediately one can see that this time things were very very different.  This cycle was a once in a century solvency crisis, and that is qualified as not from its intensity, which was massive, but rather the makeup and reasons for the event were rare.  In a solvency event the collateral value, net worth, suddenly crashes and the Minskian "hedged" or even worse the "Ponzi" economy crashes.  The nature of business that produced the assets, the industry is not changed per se, but does have to respond to the calamity in culling employment and rapidly downsizing.  So the answer for a solvency event is Kindleberger/Bagehot massive injection of liquidity to banking system as lender of last resort and is not really a matter of rates.  but of course rates drop.
So the following different picture is presented for the 2008 crisis versus all previous business cycles. All other business cycles but for perhaps the Great Depression, but even then it was not a solvency crisis, show a discrete or move per move that is  highly correlated between both measure of unemployment - the experienced (by definition) insured labor pool and the general labor force, many who are not able to claim unemployment insurance. 
But the 2008 solvency event is very different.  It shows a very large shock, non-discreet leap in both measures and then the insured unemployment rate gaping down with little change in the national rate. 

This differentiation is also picked up by the Beveridge Curve and the plummeting Labor Force Participation Rate and the Employment to Population ratios.  They have been posted so often I will not post them here.  But what the drop in Insured UER shows that normal cyclical repair does occur and as the rate drops, eventually the national UER follows, but now with a large differentiation that does not occur in all other previous business cycles.   Again that is because 2008 is a not a business cycle but a solvency event.

This also means one has to be very careful in using any array of data that includes the current event with all past cycles. LFPR and EP and so on are basically without any usefulness.  Things have changed.  The massive drop in net worth has permanently altered, well at least for our lifetime, the Cobb Douglas function depiction of the ratio of labor to capital that makes the base economy potential.  Along with that the measurement of output gap that is most important to the Fed is changed.  The clock is restarted with a solvency event.  Back to base 100 in terms of growth potential to figure out when inflation will occur.  The large differentiation between insured and uninsured is permanent.  Output gap post solvency event requires a Hodrick Prescott filter definition rather than the ex post trend growth.  Will let Williamson outline that for you:

But there is very good news now being shown by the above picture.  Note that the relationship between insured and national UER rates has assumed its usual relationship in previous cycles.  Large swings up and down come about in the insured UER as a more steady trend occurs in the UER national rate.  Why? This is because of the experienced folks entering and leaving the work force, with the larger the volatility in the change in the insured rate shows ever greater dynamic activity in the economy as a whole range of effects and remedies provide repair to the economy - Schumpeter's "creative destruction".

There is a solid signal from the volatility of the insured UER during a recovery phase that has prescience.  First heightened volatility in the insured UER rate confirms that the event has ended and recovery has started,  Then the greater the volatility the more powerful the flows are to bring recovery. Every large drop in the national unemployment rate requires and is signaled by this increase volatility in the insured rate.

I call this the insured unemployment rate "pump" as when it mounts it pumps the national unemployment rate down. My rough sketch shows that a move in excess of .30 vol in 6 month insured rate must occur before the final move downwards in the unemployment rate occurs.  We are there now.

The following is the smoothed volatility of the insured UER, showing it is increasing and now a strong enough engine to drop UER to level well under 6%.