Tweet
Re: Astrology and markets
Bart,
Every person must decide what to trade on, on their own.
If the Bradley siderograph works for you, then so be it.
From my viewpoint, the whole problem with any numbers based, star/planet based, or other mumbo jumbo is that there is some relationship between almost anything and anything else.
Its called min/max. The stock market never goes to zero, nor does it attain infinity. As it is theoretically based on some semblence of reality, the range is actually smaller than 0 to infinity - rather the range is more like 25% to 400%. Thus any system based on anything which keeps its output in this said range is going to bear some resemblence (with some set of tuned variables in a given time range) to actual behavior.
The PHI or Fibonacci or whatever number Lukester chooses works because of this - much as the so called 'fractal' behaviors are replicated in nature. Because in nature, very few systems exhibit exponential behavior: bacteria colonies and nuclear reactions being a couple of notable examples. Most natural systems have a physical upper and lower bound, hence 'fractal' modeling and behavior.
But the key here is the statement about modeling and tuning. Just as a climate model, or a circuit simulation model, or any other model can be tuned to predict behavior - the prediction is only good if the model covers all possible ranges of behavior.
As someone with intimate experiences with circuit modeling if not necessarily a statistician, I have seen firsthand numerous instances where new circumstances (i.e. smaller geometries or highly sensitive analog designs) bring about behaviors completely not predicted nor predictable by previous models. The models are then updated until the next unanticipated situation arises.
My problem with astrological, numerological, or whatever crap is that they are nothing more than smokescreens for precisely this type of modeling and tuning. Sure, with enough variables and backtesting for past behavior, they theoretically could have predictive value if past circumstances repeat.
But the markets encompass far more variables than the 500 or so in a circuit simulator model, offer far fewer Monte Carlo range data points to validate against, and furthermore are completely non-natural systems.
Why then should these numbers have true predictive power beyond any other type of modeling?
And should these models be relied on when we are clearly in a situation with very few precedents?
EDIT: I call the stock market a non-natural system because a stock market is like a Ponzi scheme. In nature there are no Ponzi schemes, because there are few substitutes for food. Animals eat, or they die. Animals can invest other animals' food, but they can't loan food or charge food interest. Animal IPOs happen all the time, but there are no animal venture capitalists. And there definitely aren't animal bailouts - only humans do that.
Bart,
Every person must decide what to trade on, on their own.
If the Bradley siderograph works for you, then so be it.
From my viewpoint, the whole problem with any numbers based, star/planet based, or other mumbo jumbo is that there is some relationship between almost anything and anything else.
Its called min/max. The stock market never goes to zero, nor does it attain infinity. As it is theoretically based on some semblence of reality, the range is actually smaller than 0 to infinity - rather the range is more like 25% to 400%. Thus any system based on anything which keeps its output in this said range is going to bear some resemblence (with some set of tuned variables in a given time range) to actual behavior.
The PHI or Fibonacci or whatever number Lukester chooses works because of this - much as the so called 'fractal' behaviors are replicated in nature. Because in nature, very few systems exhibit exponential behavior: bacteria colonies and nuclear reactions being a couple of notable examples. Most natural systems have a physical upper and lower bound, hence 'fractal' modeling and behavior.
But the key here is the statement about modeling and tuning. Just as a climate model, or a circuit simulation model, or any other model can be tuned to predict behavior - the prediction is only good if the model covers all possible ranges of behavior.
As someone with intimate experiences with circuit modeling if not necessarily a statistician, I have seen firsthand numerous instances where new circumstances (i.e. smaller geometries or highly sensitive analog designs) bring about behaviors completely not predicted nor predictable by previous models. The models are then updated until the next unanticipated situation arises.
My problem with astrological, numerological, or whatever crap is that they are nothing more than smokescreens for precisely this type of modeling and tuning. Sure, with enough variables and backtesting for past behavior, they theoretically could have predictive value if past circumstances repeat.
But the markets encompass far more variables than the 500 or so in a circuit simulator model, offer far fewer Monte Carlo range data points to validate against, and furthermore are completely non-natural systems.
Why then should these numbers have true predictive power beyond any other type of modeling?
And should these models be relied on when we are clearly in a situation with very few precedents?
EDIT: I call the stock market a non-natural system because a stock market is like a Ponzi scheme. In nature there are no Ponzi schemes, because there are few substitutes for food. Animals eat, or they die. Animals can invest other animals' food, but they can't loan food or charge food interest. Animal IPOs happen all the time, but there are no animal venture capitalists. And there definitely aren't animal bailouts - only humans do that.
Comment