Am I wrong about velocity?

 I claim that stable transaction rates at the retail level are globally stable. So what happened between 1995 and 2000 where velocity continued to rise?   If the economy takes five years to reach a stable quant, then there is no QM theory. And from 2005 to 2008, why wasn't M1 velocity stable?  Note that the two periods of non-constant velocity ended in recessions.

Looking at the 1995-2000 deviation, a closer look shows the velocity was stable to an uncertainty level up to about 1996, then it rises above 70.  Transaction rates are not perfect spectral lines, but close.  The economy will let spectral production lines vary, up to the point they interfere with adjacent spectral lines.  I don't mind a two year adjustment period to reach stability going up, I would be much more bothered if we failed to crash to a certain historical velocity.  How is it that we crashed exactly to 80-85 after July 2008?  We had stability from 2000 to 2005, look where we ended up in 2010, almost the same velocity, a reversion to a known constant afters a rapid fall.

Another research notice.  In this scale system the stable velocities are 65, 82, 100. Looks like a gain of about 15% at each jump.  Look back at the previous post about city growth, what did the authors claim? A 15% gain from specialization.  The author also notes that any independent channel in a city will show the same growth as any other channel, to within a 15% error.  Then look here, retail sales level dropped by 15%. Looks like a constant of uncertainty which equals the gains from specialization.

Channel theory applied to economies is new, as we use it we will get more accurate definition of transaction rates, sizes, gains from specialization and the rest.   I am not worried, this theory will win the day.