This graph below, created by Doug Short, is a very good chart that has been turned to the proof of the ”bear market” case: that we are in for a long, deep, secular bear market reaching out to the mid-2020′s.  If you are inclined to think markets are repetitive, that they “cycle” through phases, like the moon, on a regular basis, then this graph will scare the bejesus out of you.  Your next stop will be the gun store, then out to the Yukon Territory to meet with a realtor about some mountaintop property that can be defended 360 degrees.

I want to deal with this graph over a couple days.  First, today, to describe what chartists and “technical analysts” see in it that leads honest, sane forecasters to think Armageddon is upon us.  Present the bear case.

Then, tomorrow, I want to make the counter-argument, that markets are not pre-destined by their history.  And that this graph is inconsistent in at least one major way, inconsistencies that render it useless as the 21st Cent replacement for tossed chicken bones on the Oracle’s table.

The red line ascending to the right is a regression, the “average” growth of the market’s value over time.  (Yes, you math nerds, I know that’s not quite correct, but it’s close enough.  Pay attention).  Each of the four descending blue lines represent the crest to trough history of one great bear market.  Notice they are parallel.  Well, they are if you fudge the top and bottom points a bit … again, no nitpicking; close enough.  The point the chartist is making is that bear markets “act” the same; descend at the same rate and bear (sorry) a close family resemblance.

Next, notice also that the tops roughly occur at 30-35 year intervals: 1901, 1929, 1965, 2000.  There is a periodicity to the markets, is the argument, a rhythm that as predictable as the phases of the moon.  The bottoms are not that repetitive by date, but notice that each bear market lasted about 20 years.  1901-1920.  1929-1949.  1965-1982.  For our chartist’s analysis, however, it’s the tops the matter, the tops that trigger bear markets and that therefore are the key to the analysis (which is why we skipped over the 1932 bottom for no particular reason).  We had one of those tops in 2000.  Now, if we install a descending line parallel to the other three using the 2000 top as a start we get the bottom of the secular bear in the mid-2020′s.  Ergo, TA-DAAA, we are facing a massive market collapse lasting about another decade or so.

This is a quick and dirty distillation of our chartist’s work, that is true.  There’s a lot more grinding and cranking to this sort of chart interpretation and “technical analysis” involving interim tops and bottoms, cyclical periods of various lengths, and a huge amount of creative pattern recognition to make the arguments sound meaningful and convincing.  But what I’ve offered above is what it all boils done to: the markets are predictable because they are chronologically repetitive and internally cyclic, if only in broad outlines.

Moreover, there’s a usually unspoken but necessary underlying premise at work: markets are and must be cyclical in spite of all the external forces of the political and social world.  They must either be perceived to be in some sense independent of the rest of the world’s drivers, OR they must be perceived to be the primary driver of all of the rest of the world’s forces.  IOW, in this latter dogma, markets are the cause and the rest of the world is the effect.

Lastly, market cycles are determined, in the minds of chartists, not by fundamentals (revenues and earnings and such) in the companies that compose the market’s assets but by the psychology of the buyers and sellers who compose the market’s participants.  The root argument of the technical analysts and chartists is that human society is on a long, predictable 30 year cycle and that markets either reflect or create.  In either case the cycles are predictive.

Play with this graph.  Look for patterns in it, because if you look for them you will find them.  We Homo Sapiens instinctively find patterns in everything whether they exist or not.  Lay on your back and watch the clouds take the shape of things familiar to us.  We are genetically designed to be pattern recognition machines (is that a tiger camouflaged in the grass or just light and shadow playing games with my eyes?).  It’s the lines in the graph above that make the history seem repetitive, is it not?  Take the lines out of the chart and does it still seem so obvious?  This set of lines looks good.  A different analyst would conjure up a different set of line, and different market history and probably a different future.

Technical analysis is in the short term a self-fulfilling prophesy.  There are so many traders who practice this quasi-religion that trading patterns do exist.  An analysis hits the streets and all the traders lock on to it, buying if the analysis is bullish and selling if its bearish.  The buying drives up the price, naturally, or the selling drives it down, purely on supply and demand.  At some point, the analysis declares the bull or bear run over and the market “predictably” reverses.  The chartists call prices moving between the top and bottom of a chart a “channel”.  Most of the automated technical trading software you can buy simply attempts to identify the top and bottom of a channel, which is the self-fulfilling prophesy of the moment for a given stock in the market.  A good logo for one of these software firms, imho, would be a snake eating its own tail.

Tomorrow: the one major problem with all deterministic chart analysis, and why the chart patterns and the markets they are intended to represent are about to change.  Perhaps permanently.