Thursday, July 30, 2009

Circus Tent in Outer Space

The last post introduced the concept of the magician, an aggregate representation of all types of induced feedback in the market. At this point, there are two distinct mechanisms of feedback in the market. The first arises from technical analysis. The second is born from the capability to steer technical patterns with subtle nudges at the right points of instability. Both have their limits, and that's the focus of the post today.

The magician (apologize for lack of an illustrative introduction) uses induced feedback. Induced feedback is feedback that has been triggered by creating the right input signal to cause the system to enter a particular state of self-propagation via simpler feedback mechanisms. Imagine a school bus full of children who are shifting their weight in order to rock the bus. By feeling the movement of the bus and using this to decide when and how much to shift their weight, the children are a simple feedback mechanism. They feed back one signal (the bus is rocking) to create another signal (rock more/less). Induced feedback was what happened when one kid got up, started rocking, and encouraged some of his friends. After the thing was rocking and it was widely perceived as fun, the storm was already in motion.

The magician in the stock market is no different. By utilizing the right stimulation to the right trading pattern, the magician has a butterfly on the end of his wand that creates a tornado. The cost of achieving these tornadoes is very low relative to the storms that are created, which continue to build through their own mechanisms in the absence of further capital input. The magician reaps the benefits. However, there are two limitations to the magicians capabilities:
  • In the absence of any susceptible patterns, the magician's actions are distorted in the haze, and it costs more to create a self-propagating pattern. The profit potential is less and the stability of the resulting signal can be affected.

  • Mechanisms unrelated to feedback cycles cannot be controlled by induced feedback. Unrelated influences are noise to the magician, costing him more to manifest signal stability and causing the patterns he creates to dissipate rapidly.
The second leads us to the showdown between our chaotic cow and the magician. The clouds of chaos are ultimately created by fundamentals. Price elasticity takes over at the boundary regions, and like a giant spring of reality, continues to pull more forcefully on a stock as it meanders into the boundary. Here, the almost random influence of the cow becomes highly self-organized and much more potent, and presents a noise signal to the magician that is too powerful to be profitably opposed. At this point, the magician must bow to the cow, reverse his fortunes, and ride with the herd.

What goes on inside the chaotic boundary regions can be likened to a circus. Complete with a ring leader, lots of people shouting, and no real method to the madness except whenever the chaotic meanderings take clear form and feedback drives them in one direction or the other. Outside the circus, whatever went on inside the circus is no longer important. The feedback mechanisms are dominated by external influences that cannot be affected outside the tent. Faced with the cold reality that the circus games can only last from one side of the tent to the other, the crowds and performers head back into the tent. The magician, having driven the show outside the tent himself, stays inside the ring and charges admission for all returning.

The clouds of chaos float just behind the curtains on the stage. The magician has his tricks. The audience gets to clap, but outside the stage the show is over, and our mystery superhero, the chaotic cow, keeps the show in check. Why do I cast the cow in a positive light? You can make money off the magician. You can make money off the cow. You can predict the movements of the cow. You can't predict the movements of the magician. The timing when stocks will move off the stage is much less certain than when they will get moved back on the stage. Follow the beef.

Illustrations on their way, but I want to keep moving through the material and doing a good illustration can take a good chunk of time.


  1. While I appreciate your posts, I think you are over thinking how the markets work. Stocks that pay dividends trade on the PV of their future cash flows and that is constantly changing with new information but it shouldn’t drastically change too much.

    Stocks that don’t pay dividends become tougher since there is no incentive to hold and one would have to rely on the market place to reasonable value the company. If there is no trust in the market valuing the company appropriately then it tends to trade at a larger and larger discount to compensate for the market conditions. As I can see in my portfolio, stocks that pay dividends trade closer to their “intrinsic” value than stocks that don’t pay dividends.

    Remember time is money and market participants want to be paid for any lost opportunities due to the passage of time.

  2. The model I'm working with is valid for everything from Kellog to Nano-Sci-fi Unicorn LLC. It's just that these complexities won't be expressed with as much force on a company with readily estimable assets and dividends. I went over this more in a post way back called "The Origins of Chaos."


Please give yourself a Turing test to verify that you are human.