It is said that 95% of all traders lose their money quickly and quit. I have not found the source of that 95% statement on the web, so if someone really did some statistic about the percentage of losing traders please let me know!
Anyway I believe it is true, as the edge that a system gives you is quite small and thus any mistake or irrational behavior will turn a winning system into a losing one. When you backtest systems posted on web forums, which I've done in the last months, you find that almost all are unprofitable. So it seems that the many hapless followers of those systems will join those 95% losers sooner or later. Maybe I'm preaching to deaf ears, but anyway, here I've listed the seven system trading methods that I think are most certain to separate you from your money. Please post your comments if you have made different experiences.
► Buy an unknown system. Many websites sell software products with fanciful names like "Forex Pilot" or "Million Dollar Robot" that claim to make you rich through automated trading. The description of their methods is somewhat cloudy, but they usually have an impressive backtest posted on the website, or a trade list with several 100% profit per year. That Forex Turbo Robot got rave reviews on robot websites and started well, so why has it yesterday night wiped my account? Don't believe everything you read on the web. By paying a little fee, robot vendors get any review that they want, and with easy tricks they can generate any backtest - especially when the motivation is not to develop a profitable system, but to sell many copies.
► Use an untested system. A strategy is worthless as long as it does not come with an equity curve generated in a solid walk forward analysis or an out-of-sample test. Incredibly, trading book authors often describe detailed systems in their books without having properly tested them - and are even proud of being old school and not familiar with programming. They just "know" that their systems work, or have verified them with "eyeballing", or have successfully traded them "since 20 years". I've tested many systems described in popular trading books and found that almost all were losers, and this was obvious even in a simple backtest done in five minutes.
► Use a recommended system. Someone I trust is trading this system since 6 months and made $40,000 with it so far - so that's certainly a proven strategy? When 100 people use similar unprofitable systems, there's a good chance that 20 of them are winning in the first 6 months, while the other 80 are losing or have quit already. Of course, the promoter of such a system will only reference the winners. That's called 'survivorship bias'. Six months trading proves nothing. "Testing" a system by trading it some time on a demo account proves even less. Recommendations are no substitute for a proper backtest over at least 4 years, better 6 years.
► Believe in esotericism. There are many trading methods that have no rational foundation, but on all trader forums you'll find people who swear that they work. There are Elliott waves, Gann lines, Fibonacci numbers, financial astrology, moon cycles, and many other bizarre ways to generate trade signals. Such a strategy might work for its advocates, especially when they publish books and sell trade systems, but it won't work for you. I've never found an irrational system that was profitable in a backtest.
► Use unoptimized parameters. My system is already profitable without any optimization, so I can certainly trade it with confidence, as it can't be overfitted? Think again. Only the optimization process can show you the stability of a system over its parameter ranges. It happens surprisingly often that apparently random parameters are just at a profit peak in the parameter space - especially when you take over this system from someone else. Trade only systems that were either properly optimized, or have no adjustable parameters.
► Increase the trade size after a loss. Such methods are known as Martingale or d'Alembert systems, and are often used by beginners in roulette. You bet on red or black and double your stake after every loss. The theory is that your chance to win increases when you lost often. Unfortunately, it won't. On the contrary, trading strategy results are often autocorrelated, so after you lost a trade there's a better than random chance that you'll also lose the next one. Martingale or similar systems are for the casino only. Of course they lose there also, but you then at least had fun.
► Have a high winner rate. Surprisingly, lots of winning trades can - and in most cases, will - wipe out your account. A high winner rate is usually achieved with a low profit target and a distant stop (or no stop at all). The theory is that prices will eventually move back to the buy level and hit the profit target if it's only small enough. Such a system will indeed generate a win rate high above 90%, even with random trades. The bad news is that the theory is wrong - many stocks and currencies are today far from their prices from one year ago and will most likely never get back there. Thus, such a system will almost certainly sooner or later enter a trade that won't return to the buy level, and give you the margin call. With more capital you can delay that event - and lose accordingly more when it happens. Don't trade systems with >90% winners, and be careful with profit targets anyway. They always increase the winner rate, but usually reduce the overall profit.
Anyway I believe it is true, as the edge that a system gives you is quite small and thus any mistake or irrational behavior will turn a winning system into a losing one. When you backtest systems posted on web forums, which I've done in the last months, you find that almost all are unprofitable. So it seems that the many hapless followers of those systems will join those 95% losers sooner or later. Maybe I'm preaching to deaf ears, but anyway, here I've listed the seven system trading methods that I think are most certain to separate you from your money. Please post your comments if you have made different experiences.
► Buy an unknown system. Many websites sell software products with fanciful names like "Forex Pilot" or "Million Dollar Robot" that claim to make you rich through automated trading. The description of their methods is somewhat cloudy, but they usually have an impressive backtest posted on the website, or a trade list with several 100% profit per year. That Forex Turbo Robot got rave reviews on robot websites and started well, so why has it yesterday night wiped my account? Don't believe everything you read on the web. By paying a little fee, robot vendors get any review that they want, and with easy tricks they can generate any backtest - especially when the motivation is not to develop a profitable system, but to sell many copies.
► Use an untested system. A strategy is worthless as long as it does not come with an equity curve generated in a solid walk forward analysis or an out-of-sample test. Incredibly, trading book authors often describe detailed systems in their books without having properly tested them - and are even proud of being old school and not familiar with programming. They just "know" that their systems work, or have verified them with "eyeballing", or have successfully traded them "since 20 years". I've tested many systems described in popular trading books and found that almost all were losers, and this was obvious even in a simple backtest done in five minutes.
► Use a recommended system. Someone I trust is trading this system since 6 months and made $40,000 with it so far - so that's certainly a proven strategy? When 100 people use similar unprofitable systems, there's a good chance that 20 of them are winning in the first 6 months, while the other 80 are losing or have quit already. Of course, the promoter of such a system will only reference the winners. That's called 'survivorship bias'. Six months trading proves nothing. "Testing" a system by trading it some time on a demo account proves even less. Recommendations are no substitute for a proper backtest over at least 4 years, better 6 years.
► Believe in esotericism. There are many trading methods that have no rational foundation, but on all trader forums you'll find people who swear that they work. There are Elliott waves, Gann lines, Fibonacci numbers, financial astrology, moon cycles, and many other bizarre ways to generate trade signals. Such a strategy might work for its advocates, especially when they publish books and sell trade systems, but it won't work for you. I've never found an irrational system that was profitable in a backtest.
► Use unoptimized parameters. My system is already profitable without any optimization, so I can certainly trade it with confidence, as it can't be overfitted? Think again. Only the optimization process can show you the stability of a system over its parameter ranges. It happens surprisingly often that apparently random parameters are just at a profit peak in the parameter space - especially when you take over this system from someone else. Trade only systems that were either properly optimized, or have no adjustable parameters.
► Increase the trade size after a loss. Such methods are known as Martingale or d'Alembert systems, and are often used by beginners in roulette. You bet on red or black and double your stake after every loss. The theory is that your chance to win increases when you lost often. Unfortunately, it won't. On the contrary, trading strategy results are often autocorrelated, so after you lost a trade there's a better than random chance that you'll also lose the next one. Martingale or similar systems are for the casino only. Of course they lose there also, but you then at least had fun.
► Have a high winner rate. Surprisingly, lots of winning trades can - and in most cases, will - wipe out your account. A high winner rate is usually achieved with a low profit target and a distant stop (or no stop at all). The theory is that prices will eventually move back to the buy level and hit the profit target if it's only small enough. Such a system will indeed generate a win rate high above 90%, even with random trades. The bad news is that the theory is wrong - many stocks and currencies are today far from their prices from one year ago and will most likely never get back there. Thus, such a system will almost certainly sooner or later enter a trade that won't return to the buy level, and give you the margin call. With more capital you can delay that event - and lose accordingly more when it happens. Don't trade systems with >90% winners, and be careful with profit targets anyway. They always increase the winner rate, but usually reduce the overall profit.