Trading Methods vs. Mechanical Trading Systems

[Joe Ross]

Method vs. Mechanical

Recently, I received a private email from a member of T2W asking me why I am offering mechanical trading systems. I want to assure you I am not. I do not believe in mechanical trading systems, and have never taught anyone to use or develop them. No one who has ever been employed at Trading Educators was taught by me to trade in that manner.

There are reasons why I do not believe in mechanical trading systems, especially those of the kind called “black box.” There are many reasons, but I will set forth just four of them here:

1. You will never learn to trade using a mechanical trading system. You will always be dependent upon the accuracy and validity of computer generated signals. This is not trading in any sense of the word. All that is involved is the discipline needed to follow the dictates of a robot. In my opinion, the degree of discipline needed to blindly follow a mechanical trading system is far greater than the discipline needed to follow a discretionary method of trading.

2. Mechanical trading systems condition you to always need a crutch. You can never depend on your own discretion and good judgment to save you when things go wrong.

3. While a mechanical trading system may deliver the short-term euphoria of some successful wins, it will eventually fail and you will need to adjust either the system or your trading. Since you will not have truly learned how to trade, you will not know how to adjust your trading, so you will be compelled to adjust the system. If it happens to be a black-box system, you will have no way to make that adjustment.

4. Trading mechanical trading systems without ever really learning how to trade impairs your ability to develop as an individual. Learning to trade involves major beneficial personality adjustments. You must learn discipline and control of yourself. I ask: Is it not in the long run more desirable for you as an individual to adjust your character traits, than it is for you to adjust a mechanical trading system?

I ask all those who market black-box mechanical trading systems that sometimes work, but eventually fail, how they can call themselves traders? As everyone should know, as the markets change mechanical systems will stop working.

I am a trader who has learned how to trade. Although I do not trade every day, preferring to be extremely selective and discretionary in my trading, I actually do trade! I trade real markets with real money. As a real trader, I also teach others how I really trade. I do it live in front of others, which is much better than showing anyone a cooked up trading statement. If I win, you see me win. If I lose, you see that as well. I am not proud of my wins and I am not ashamed of my losses. Trading is a business and I do my best to teach you what I know. I do not pretend that my approach and way of trading is suitable for everyone. I do not pretend that I am winning when I’m not. I tell it like it is - the good with the bad.

When I reveal a trading method, I do not hide anything. The method is fully explained. I believe in full disclosure. I tell you what to expect and that the method will not always work. I show you the method at its worst, as well as at its best. I also tell you when I have stopped trading it, so that you can stop trading it. I want you to have real discretion in following any method I show you. Methods offer a way to earn while you earn. As one person took the trouble to write me --“Yours is the only place an aspiring trader can go to get the truth.”

Joe Ross

 

[NLT]

Is your business going that bad that you need to make publicity for your trainings?

I traded as a discretionary trader for different company's and I was able to take money from the markets. I almost always used some rules. 1 day I was saying to myself "why not use those rules in a trading system?". I tested it and it made money. Now I'm using this mechanical system at home for 2 years and it still makes money! you think that is luck? If I was trading discretionary I would do almost the same. 1 difference. Risk management is easy when there is a risk manager breathing in your neck, but is hard if you are trading alone at home.'

there is 1 thing that really makes the trader imo and that is DISCIPLINE. My system trades automatic so the only discipline I need is to start is in the morning and that's it.

You (maby) make money your way, I make my money my way. But it's a bold statement to say the other method don't work if you don't have enough experience in it.

 

[forexoft]

Just wanted to say, I also had not believed in mechanical trading systems until I developed one for myself. Though I still work to make this system more reliable and less risky, I know that it will make money on real account, because now it makes money on demo account.

Also, I know that there are many other automated systems that makes profit.
And most likely you will not find them in public forums.

 

[Joe Ross]

I agree! There are mechanical systems that work. It is most important that you keep them confidential. The longer you keep them privately and to yourself, the better the chance they will work and continue to work.

However here is the caution: Do not tade with too much size. If you get too big, you will acccomplish the same thing as happens to mechanical systems that end up in the hands of too many traders. Sooner or later you will see the market adjust to your system and your stops will be run. I have proven this to be true countless times.

The mechanical systems I am against are those that are sold to the public as black-box systems or those in which there is no possible chance for discretion of any kind.

I once developed a mechanical system for trading currencies. I kept it to myself and it made a steady 12 ticks every time i received my entry signal. This went on for 9 months. I was trading 30 contracts and all I had to do at the time was to stick in a MIT order 12 ticks from my entry and take the money to the bank.

After 9 months, I began to notice I was getting only 10 ticks, so I adjusted to 10 ticks. But the market kept changing and after a couple of months I was getting only 8 ticks and then only 6. At 6 ticks I had to use 60 contracts, a size which at that time I was not comfortable using.

So I stopped trading the system but kept an eye on it for about 1 year. All of a sudden it was mkaing 12 ticks again. The only answer I could come up with is that someone had created a system based on the way my own system worked. When I stopped trading it, his/her system also failed. When I came back, the 12 tick objective worked again for over 18 months.

 

[forexoft]

You are right Joe, black box systems is not very good choice.
Trader must completely understand algorithm of MTS in order to trade using it effeciently.
Also I think that system must be evolving, I mean new builds must be released, bugs are fixed, features added. If you use system that you bought a year ago - there is probability that this year will have a lot of loss trades, because some systems are optimized for specific time period and new market conditions may discover new weak places in it.

Also I think that trader should always keep an eye on MTS while it is trading, MTS is just computer program, it can make mistake in unexpected conditions, so I prefer to be informed about what is MTS currently doing to be able to fix any undesired issues. It can be implemented as SMS or email alerts, etc.

 

[TWI]

I find it hard to believe some of this rubbish about Mechanical systems.
Try telling Jim Simmons that black box systems do not work. Consider his performance over the past decade. Since its inception in March 1988, Simons' $ 5 billion Medallion fund, has amassed average annual returns of 35.6 percent. Medallion's cumulative returns are 2,478.6 percent. Among all offshore funds over that same period the next-best performer was Soros' Quantum Fund, with a 1,710.1 percent return.
All mechanical systems and only the top of a big stack of automated funds that have great track records.

You can even apply for a job with Jim
http://www.aas.org/jobreg/backissues/2001/january/no11818.html

 

[forexoft]

twalker, it seems like Jim Simmons and Soros do know very well their own systems, so they are not black box for them, right? they know how to trade with it.

And what about ordinary people who bought some black box systems, do you know a couple of millionaires among them?

 

[7thSignalTrader]

Method vs. Mechanical


1. You will never learn to trade using a mechanical trading system.

You will if you were the one who did the R&D, Engineering, Testing and Certification of the system. By definition, in order to create a "system", one had to fully integrate trading experience, theory, methodology, tactics and strategic trading concepts. Without those things, a reliable system cannot be built. Without an understanding of what moves prices and how to measure those movements and project normative behavior of the market, one has no technical basis for building a system.

One does not use a trading system to learn. One uses a trading system to apply what one has already learned. So, you actually have to be a good, effective and consistent trader already in order to build and then use a "black box" system. Using someone else's system that one did not build, is quite another story.

You will always be dependent upon the accuracy and validity of computer generated signals. This is not trading in any sense of the word. All that is involved is the discipline needed to follow the dictates of a robot. In my opinion, the degree of discipline needed to blindly follow a mechanical trading system is far greater than the discipline needed to follow a discretionary method of trading.

CPU based trading systems are far superior because they eliminate all but the design error from trading. In a “Method” based system (which is a Manual Trading Method), human error runs ramped and has the opportunity to inject itself at ever turn, no matter how disciplined one might be. Trading is not trading merely because a “manual” decision is made.

Is NASA not conducting space operations because it is relying upon computers? Does NASA not have a myriad of variables that they need to prepare for and have answers for should they (when they) arrive? Is NASA not in the business of solving highly complex problems within a universe filled with unknown variables? Is the Forex Market not filled with unknown variables? Should NASA get rid of it’s fully integrated Flight Operations Center Computer based Network which has been said to be the most sophisticated Computer Network in the world? Manual does not mean “better” just because it is “manual”. There are times when precise calculations are necessary to navigate waters that are filled with high levels of variable structures. If those variable structures have patterns, then the human brain is out-paced by the CPU in real-time computational regimes, when it comes to locating and projecting those patterns.

Clearly, the discipline required to follow a trading system that opens the door to a vast array of potential human error, has got be far more difficult to manage than one that eliminates every form of human error outside of human design error which can be cataloged, tested and removed at will with a zero chance of ever impacting trading results adversely in the future. Can’t do that with the always persistent human error of manually based systems.

2. Mechanical trading systems condition you to always need a crutch. You can never depend on your own discretion and good judgment to save you when things go wrong.

You can design the system to make the same strategic decisions that one would make outside of that system. You can design a level of discretion and judgment into the system dependent upon your level of risk tolerance and revenue expectations. You can design the system to discern, detect and then act upon anomalistic events that occur in the market in real-time. In fact, because it is a CPU based system, it will discern, detect and act with a level of efficiency and precision that the human brain (in real-time) simply cannot accomplish to anywhere near the same degree of consistency and proficiency. Mechanical systems don’t have to be dumb – they can be smart. They can even be taught to learn in real-time. But, most of all, the trader him/her self has ultimate authority for every decision that is approved by the system with a simple click of the mouse. So, the trader is never taken completely out of the decision support loop AND the trader is able to benefit from a level of computational control that the trader him/her self otherwise would not be able to replicate consistently.

3. While a mechanical trading system may deliver the short-term euphoria of some successful wins, it will eventually fail and you will need to adjust either the system or your trading. Since you will not have truly learned how to trade, you will not know how to adjust your trading, so you will be compelled to adjust the system. If it happens to be a black-box system, you will have no way to make that adjustment.

There are trading system that are mechanical that trade to very high levels of accuracy and have done so for prolonged periods of time. 90% of most traders in the business fail. Most of them do not use genuine trading systems of any scope and/or technical depth. That leaves only 10% of successful traders. What are they using? In order for me to make the determination that all mechanical trading systems fail, I would have to have tested and proved all mechanical trading systems in existence. Since I have not done that, then I really don’t know whether or not all mechanical systems fail, or whether or not there are some that succeed a high percentage of the time.

If I’ve designed the system and it fails, I can trace the failure, locates its source and point of origin, and then make the determination as to whether or not the failure was within nominal specs as defined by my goals, or whether or not a code change is in order. So, if I’ve actually built the system, I have ultimate control over its output and thus, its results. I am never out of control of a system that I design. If I did not design the system, then I have no control over source changes.

4. Trading mechanical trading systems without ever really learning how to trade impairs your ability to develop as an individual.

Absolutely, true and correct.

Learning to trade involves major beneficial personality adjustments. You must learn discipline and control of yourself. I ask: Is it not in the long run more desirable for you as an individual to adjust your character traits, than it is for you to adjust a mechanical trading system?

If one has already done the research (years of study) then one should already know what is possible in the market and what is not within their reach. Depending on how much work one does in determining what the market data will reveal about itself, one will alter their trading character to be in-line with their expectations. And, those expectations will be based on empirical evidence.

I ask all those who market black-box mechanical trading systems that sometimes work, but eventually fail, how they can call themselves traders? As everyone should know, as the markets change mechanical systems will stop working.

I don’t market black-box systems, sell them, intend on selling them in the future, or sold them in the past. But, I find this question rather interesting – indeed this entire post, as I am a so-called “black box” engineer/designer/developer/researcher. The assumption that all will fail is only a valid assumption is all have been tested by the one making the assumption. Otherwise, the premise cannot be proven. Thus, the ruling law of the universe known as “cause and effect” gets violated and that cannot happen anywhere in the known universe. So, before I can make that determination (that all “black-box” systems fail), I have to have the empirical data that demonstrates that fact. The only way to get that data would be for me to test all systems in existence and that is not very likely to happen. Many excellent black-box traders won’t go public for their own reasons which I fully understand. So, the premise can’t be proven until all are tested – until then, it is personal opinion, that “all” must fail.

This also does not deal with the matter of not all failure being bad long-term. Just because the system fails does not mean that the system is bad long-term. 9 failures out of 280 trades equals 96.78% accuracy. If that is accuracy to a specified target, then the results are augmented far beyond what the percentage implies. If only 40% of those trades were 96.78% accurate to a specified target and the other 60% failed to hit their targets but broke-even, then my system’s overall accuracy is far less than 96.78% to target, but my growth of capital is well beyond the bottom 90% of all traders in the business. If all I did was focus on the 9 failures and ignore the other 271 successes, then I’ve probably not focused on the things that are going to make me a better trader.

Well, in closing – I don’t have anything to sell and never will. I’ve spent 6 years researching and developing a new way to measure market behavior that does not include any conventional or traditional technical analysis. It is a highly technical system and one that would be classified by many as a “black-box”. However, I find that many who use the term “black-box” are not engineers themselves or mathematicians. Rather, those who use the term in a negative sense in order to explain why they have had no long-term success with mechanical systems. Even the term “mechanical” is used by many in this business to denote something wrong with the approach. I disagree, wholly – for a myriad of different reasons.

The factual truth about trading is that ALL Technical Traders are using Black Boxes and many of them without even knowing it. That’s the reality of the market place. But, in order to understand that, one will need to stop and think about the proposition for a moment.

What is a “Black Box” anyway? That’s a very good question and there is a very reasonable answer.

A “Black Box” to a mathematician, physicist, engineer, chemist, biologist, etc., might all have slightly different meaning, uses and purposes. A black box to a Trader could indeed have yet another meaning. In general, a black box simply means that both the Input and Output into a system is processed outside the primary control of the observer. That’s it. In the trading world, people tend to use the term “black box” to give meaning to an underlying premise that one is attempting to make.

But, the bottom line, regardless of industry type or professional discipline is that there is a clear purpose for using the so-called “black box”. You use a black box for a number of different reasons, but one of the more common, is to ensure consistency and regularity in computational methodology. That is a very real, meaningful and useful purpose – especially in trading when one is trading on patterns derived from the market data itself. So, black boxes can be very helpful for ensuring consistency in analysis, scoring, testing, measuring and anomalistic detection.

A black box is nothing more than a set of rules, relative to trading. Rules set forth the boundaries and establish the parameters within which discipline will be measured. A rule simply starts out with a simple question such as: “If, When, And, Or, Not, Where, etc.” So, every trader (whether a Fundamental Trader or a Technical Trader) has a set or domain of rules within which he/she will make trading decisions. There is no escaping the existence of rules for either trader. One trader (the Fundamentalist) like to say that he/she has no rules because they trade on news and fundamentals. Yet, that very statement itself is a Rule! So, if he/she wanted to, they could codify that rule and use it inside a CPU based mechanical system, if they have the coding skill to do so.

So, all traders have their own rules for making trade decisions. The only question is whether or not they have coded those rules into a mechanical paradigm to eliminate human Input and processing error. Rules themselves can be as flexible as one can be creative in their mind. One rule might have 735 inputs and 1 output (buy or sell). But, there is nothing that says those inputs cannot be news, fundamentals, technical data, or a combination of all three! Yet, it would still be a rule and rules are what make black boxes possible.

So, the question is not whether or not one trades using a “black box”, because we all trade with rules with varying degrees of input, output and variability. It is either a coded black box, or an un-coded black box. Or, put another way, we all have either coded rules, or un-coded rules.

The real question becomes: How Black is your box?

 

[fxmarkets]

The real question becomes: How Black is your box?

I Beg your pardon? moderators.... I, I.... talk about being direct

hmm, intensive post....... agree that if you can really crack trading, then the next challenge could well be, for some, can I code that...........