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.

Strategies & Market Trends : NetCurrents NTCS -- Ignore unavailable to you. Want to Upgrade?

To: fut_trade who wrote (8616)	3/31/2002 3:18:14 PM
From: Dan Duchardt	Read Replies (1) \| Respond to of 8925

fut_trade, What I find is that these systems do make money, but we are looking at maybe a net gain of say $50/trade/emini over say the last 10 years. That is not really a lot of money. That depends on how many trades the system is making in any time period. Using an approximation, $50 compared to the equity required for an e-mini contract (roughly $4,000) is a gain of 1.25% per trade. If you can make 1.25% per trading day, that's 250*1.25% or 312% per year. If you compound it, which is not realistic because you cannot trade partial contracts, the gain would be 2100% per year. What is realistic is to trade additional contracts as you accumulate gains and grow your account. Starting with a single contract, at the end of one year you would be up to 13 contracts with an accumulated gain of over 1200%. I call that a lot of money. Of course systems are not so well behaved as returning 1.25% per day, every day, so there is some uncertainty in any projection, but this last calculation only assumes an average gain of 1.25% over the duration of trading some fixed number of contracts, not that the gain is achieved every day. A general observation about "randomness" or lack of it seems in order after all the recent discussion here. I think that instead of thinking in terms of the market being sometimes random, and sometimes not, all would do well to recognize that there is always an element of randomness in the market. There is a rather extensive field of mathematics dedicated to understanding the behavior of "random variables". I'm certainly not an expert in that area, but the basic idea is that time series that occur in nature consist of a deterministic component (signal) and a random component (noise). Signal processing techniques are used extensively in technological applications to separate these two components and eliminate noise to the greatest extent possible. The ability to recognize any signal depends on how well the nature of the signal to be detected is understood, the nature of the noise, and the strength of the signal compared to the strength of the noise (signal to noise ratio). The market has much in common with other time series. It is never a question of whether or not randomness is present; it is always there. The question is whether there is any real "signal" present (trend, or perhaps relatively long term oscillations containing multiple trends) that can be exploited. The business of trading, whether systematic or discretionary, is the business of trying to recognize signals in the presence of noise and timing entries and exits to take advantage of what appears to be a deterministic component in the time series. Given all of that, you would think the expert signal processors would apply their mathematical wizardry to the markets and develop systems that made tons of money. No doubt many have tried, and I dare say some have achieved some measure of success. And many have extensively studied the nature of the signal and devised techniques to predict trends or turning points (Elliot waves, Gann charts, Fibonacci levels, pivot points, etc.) The problem is the one that you have stated. All kinds of systems can be developed that will identify and separate the signal from the noise in historical data, and they can be optimized to extract the most signal possible over any time frame in that data set. The problem is that no system can ever predict that whatever signal existed before NOW will persist into the future, or even that the noise will continue unaltered. There is no way to know if the next tick is due primarily to a signal, or just the result of random noise. The real problem all of us have is identifying the market conditions that are most likely to favor a signal that can be exploited and locking in on that signal when it is there to make some gains, and then avoid giving it all back while the signal is gone as we over react to every little move in the market. Of course there are still some of us who at times refuse to recognize a signal when we are on the wrong side of it because we just know that it's really only a burst of noise that will soon pass. Dan