WS and GF

[Luke asks:] For the world systems theorists: have you noticed any peculiarities in WS history when it comes to trade/communications networks, long cycles, or world economies in Afro-Eurasia? ; (peculiar insofar as such structures/processes would seem to transcend the normal analytic tools used to understand economic behavior in history; cyclic patterns via longwave analysis & period specific datasets (appl. to historic periods, say from 1200 to 1500 CE) th/ don’t seem to fit how we know {or think we know} that business cycles should work in economic/financial settings on an ordinary basis). In other words, I’m asking about irregularities in the world system; are there any that you can identify?

[Mike:] I'll discuss the shift of 50-70 years ago as an example.  The Kondratiev longwave from the late Medieval times to the first part of the 20th century has averaged around 50 years in length.  Today it seems to be about 70 years long.  Primary evidence for this comes from secular trends in the stock market. For 200 years US stocks have move in broadly up and down patterns of 10-20 years.  One up and one down trend make up a Stock Cycle.  Two Stock Cycles make up a Kondratiev Cycle.   From 1802 to 1929 there were five sequential stock cycles between 18 and 33 years in length (average 25.4).  Since 1929 there have been two, 34-37 years in length, average 35.5.  These  averages are different at >93.6% probability.  Evidence from interest rates and prices suggest that today we apparently are in the same position in the cycle as we were in 1930.  These observations all suggest a ~70 year cycle length today, up from the centuries historical norm of ~50. 

In my Kondratiev book I present a lot of evidence that Strauss and Howe generations are aligned with the K-cycle.  Their have been two shifts in alignment.  The most recent which developed over 1860-1930 changed from two generations per K-cycle to four.  Generational lengths have fallen from around 25-27 years to about 18 today.  The 18 year length has shown up most recently by the big stock bull market from 1982-2000.  Since I had forecasted the bull market peak to within a year, it was possible to know based on valuation before it happened that it would occur around 1999 and not much later or much earlier. This means 18 is significant.  And then when you throw in the previous secular trends 1966-82, 1949-66, 1929-49, you see lengths of 16-20, centered on the 18 years.

What is even more interesting is how we managed to get an already very overvalued market in 1997 to get going another 3 more years to reach 18.  Congress passed a bubble act in 1997 specifically designed to produce a massive bubble in the stock market.  How could this happen?  Well these cycles are also cycles of politics. Since around 1980 we have been in a conservative era, which have alternated with liberal eras, once against with that 18 year periodicity.

Now why 18?  The old cycle was tied to a 25-27 year generation, which can reasonably be related to a biological generations.  But 18 years is too show for biology.  Besides if anything replication time has increased over the past  decades.  Strauss and Howe use the idea of phase of life, but that doesn't really work with an 18 years generation (originally they assumed generation length was constant as 22 years).  My own idea that what controls these cycles today is psychological in nature, what I term a paradigm.  A paradigm that would let congressmen vote in legislation that would have to produce a stock bubble and eventual crash that could eventually be very bad politically for some if not most of them. 

This idea of a paradigm is close to the Group Fantasy concept.  But how does GF factor into 18 year periodicity?

Mike Alexander,  author of
Stock Cycles: Why stocks won't beat money markets over the next 20 years and
The Kondratiev Cycle: A generational interpretation
http://www.net-link.net/~malexan/STOCK_CYCLES.htm