eminifuturesblog
Junior member
- Messages
- 20
- Likes
- 3
Hi all, thought I'd post another article from my blog, it's about using big enough sample size when you're system testing but it's relevant even when you're not testing. Approaching trading with a commitment to the next 20 trades to ensure that the probabilities you're system gives you plays out fully is vital. Here's the article, hope you you find it relevant:
Consistent profits in trading require the adoption of a mindset which is unique to perhaps only this business and professional poker playing. Fundamental to success is the ability to think in probabilities or rather to approach trading with a crystal clear awareness of the probability associated to a certain outcome on a trade.
It stands to reason that the trader should only be interested in systems which offer a better chance than not of a positive outcome. However, having an awareness of an outcome’s probability is not enough without an awareness of the fact there is a probability associated to the process to that outcome.
What I mean by this is best demonstrated by an example. Suppose a trader has followed the Comex gold futures market for a number of weeks and has noticed it trends well and tends to continue when it breaks the day’s high or low if it occurs after 1pm. Based on his research he’s worked out there is a 60% probability that if the break occurs after that time it will trend for another 20 ticks before stalling.
He’s ready to start trading on a live account so at 1pm he places a stop order 1 tick above or below the high or low of the day and a limit order to close the position 20 ticks away. He knows that according to his research there’s a good chance (60%) that the market will give him 20 ticks if the break occurs. The difficultly for him is that he doesn’t know in what order the trades will occur, if there’s a 60% chance of success, he can quite easily get 4 or even 5 losing trades in a row before his first winner.
Depending on how much he loses on each trade, those four losing trades could be financially and emotionally so painful that he’s convinced his research was wrong and he should move onto another strategy. However, for every losing trade the odds of the next one being a winner increases subject obviously to him having done sound research beforehand. Stopping at this stage is precisely why many traders are unable to convert their research into profits and why they lack the confidence required to follow their strategy in the long term.
During the real money testing stage therefore using a sample size of 20 trades (Mark Douglas in his book “Trading in the Zone” talks about this being an adequate number) is the most effective way of drawing together data on which you can conclude whether your system should become a permanent part of your business plan.
If the results confirm what your research suggested then obviously you’ve got a winning system, if the results aren’t as good as you expected, then rather than drop the system consider why this might be. It may be simply you tried your system on a day where there were lower volumes due to holiday in Europe for example. Factors like this aren’t always apparent in the research stage and won’t necessarily undermine the validity of the system.
The necessity of having total confidence in your system is obvious but this faith can sometime stretch to only the outcome of the next trade. Viewing your system’s performance over a group of trades, requires a fundamental shift in the way you approach your trading. However by doing so, the result of the next trade won’t matter since you’ll know that given enough time and trades overall you are 100% certain you will come out a winner. Cheers Tim, eminifuturesblog.com
Consistent profits in trading require the adoption of a mindset which is unique to perhaps only this business and professional poker playing. Fundamental to success is the ability to think in probabilities or rather to approach trading with a crystal clear awareness of the probability associated to a certain outcome on a trade.
It stands to reason that the trader should only be interested in systems which offer a better chance than not of a positive outcome. However, having an awareness of an outcome’s probability is not enough without an awareness of the fact there is a probability associated to the process to that outcome.
What I mean by this is best demonstrated by an example. Suppose a trader has followed the Comex gold futures market for a number of weeks and has noticed it trends well and tends to continue when it breaks the day’s high or low if it occurs after 1pm. Based on his research he’s worked out there is a 60% probability that if the break occurs after that time it will trend for another 20 ticks before stalling.
He’s ready to start trading on a live account so at 1pm he places a stop order 1 tick above or below the high or low of the day and a limit order to close the position 20 ticks away. He knows that according to his research there’s a good chance (60%) that the market will give him 20 ticks if the break occurs. The difficultly for him is that he doesn’t know in what order the trades will occur, if there’s a 60% chance of success, he can quite easily get 4 or even 5 losing trades in a row before his first winner.
Depending on how much he loses on each trade, those four losing trades could be financially and emotionally so painful that he’s convinced his research was wrong and he should move onto another strategy. However, for every losing trade the odds of the next one being a winner increases subject obviously to him having done sound research beforehand. Stopping at this stage is precisely why many traders are unable to convert their research into profits and why they lack the confidence required to follow their strategy in the long term.
During the real money testing stage therefore using a sample size of 20 trades (Mark Douglas in his book “Trading in the Zone” talks about this being an adequate number) is the most effective way of drawing together data on which you can conclude whether your system should become a permanent part of your business plan.
If the results confirm what your research suggested then obviously you’ve got a winning system, if the results aren’t as good as you expected, then rather than drop the system consider why this might be. It may be simply you tried your system on a day where there were lower volumes due to holiday in Europe for example. Factors like this aren’t always apparent in the research stage and won’t necessarily undermine the validity of the system.
The necessity of having total confidence in your system is obvious but this faith can sometime stretch to only the outcome of the next trade. Viewing your system’s performance over a group of trades, requires a fundamental shift in the way you approach your trading. However by doing so, the result of the next trade won’t matter since you’ll know that given enough time and trades overall you are 100% certain you will come out a winner. Cheers Tim, eminifuturesblog.com
