Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Gold/Mining/Energy : Gold Price Monitor -- Ignore unavailable to you. Want to Upgrade?

To: Cage Rattler who wrote (86013)	5/28/2002 11:39:28 AM
From: E. Charters	Read Replies (2) \| Respond to of 117319

The dormant-stock-revival cycle is the simple tendency for some small companies that are in periodic money raising and development cycles to become dormant and then revive. In mining stocks it goes in 2.5 and 5 year cycles variously. It can be read by looking at history of various indicators like price, price/volume, direction/volume (on balance volume), rate of change, price over moving average price of 6, day. 30 day, 200 days and price compared to 200 day high. Also, moving average of volume, volume/price, and rate of change of these can be looked at. This is done at canadianmarketwatch.com (shameless plug) Rejecting things out of hand is perhaps necessary. But at least careful thought is necessary to know why we reject interplanetary beings from being the cause of stock market manipulation. Too much parsimony can be bad however. Obviously some effort is required to divine what other people think is too complex, as in discovering exactly how to get a gumball machine to reject phony coins. Once discovered, a real killing can be made in gumball machines. Shannon used Fourier theory to validate his sampling and information theory. This theory is known. Its exact implementation with reference to the market is not known as he did not explain it in that context. www-ccrma.stanford.edu You can see however how simple it would be with a modern computer to work your way through this stuff and predicate a prediction method from it. Shannon's pass on the market would be worth $8,385,000.00 today. While that is obviously not much worth to put on a mathematician, it was only meant as a demo. Shannon is thought to be truthful. He had little reason to lie. He was not selling the system per se. And he was certainly of good reputation. BTW another person, who I knew well used a variant of info theory on the market for 6 months and it worked. He put himself through university with options on the theory that the market is like an animal or group of animals in a psyche experiment on reward-punishment schedules. You don't have to worry about secrecy. A good market system will not be believed, and it will be sufficiently hard to implement that if it is, not many will try to beat you. If they do, they too will be obessessed with secrecy and not try to give it away to you by trading against you. The system depends partly on an efficient market's hypothesis. Time delay is enough that some efficiency, certain to be there, is a factor over time. If people are good, they depend on some kind of system. A surprising number depend on the inside info of people they can make money for. The trading of these people can be seen with some delay in the volume price cycles mentioned above. There may be no long lead time on these. Perhaps 2 weeks. Correlation need not find exact causation. It needs to find the cyclical association that can be predicted by its regularity. There is no "right or wrong" in association, just the rightness of its predictability component. Removing it and finding no effect on the cycle does not argue it had no predictability. Its effect may be extant but is immaterial extant or not. Anyway, what you do is "principle component" or "factor analysis" to find the factors that vary in synch with the flux of interest. Public confusion is ok as long as it's predictable. The market is predictable of course. It will fluctuate up and down with some monotonic increase or decrease. It is not easy to exactly predict on a fine grained basis. (That is why buy and some hold is safer relatively than quick trading.) But EVERYONE who trades presumes to predict it. If the market were dead easy to predict by the average person then it would swiftly become hard to predict within a short time. It could not remain easy to predict. Or would it? Lets imagine everyone could see a stock going high. They would all rush to buy. The stock would skyrocket. Everyone could see this and they would be able to predict its high. Then they would rush to sell it. It would tank in the blink of an eye. 1/3 would make money with better traders. 1/3 would break even and 1/3 would lose. The thing, is not everyone would see "it" given that the stock market is large and there are many people. The info is diffuse. Some stock rises come out of nowhere. It is kept secret. Stocks are in fact manipulated by insiders. The guys who make money are the guys how consider paper worthless and SELL it to suckers who think they will make money. Selling is where you make money at the low end. At the other end you have Warren Buffet who in taking a majority of a company increases its value by being a holder of huge scads of stock that do not hit the market. In time the companies fortunes increase. If they need money, Buffet in fact can lend it to them, by taking more of the treasury stock. But if Buffet dumped his stock tomorrow, it would be worthless. Even with seemingly perfect predictability or at least very good rules, the market is too diffuse to worry about "everyone" having the info. Now the insiders and marketers "have the info" as they are the ones supporting and selling the stock and running the company. And they cannot predict the price of the stock. They can barely control it. But yes, a perfectly predictable market could not exist. The actions of people make the market, so they cannot predict the results of their actions if everyone modifies their actions based on the prediction. It would be like having a time machine. EC<:-}