I have been developing and optimizing this system for a while, so I've decided it's time to share it with the FF community in order to give back what knowledge I have taken. So I'll start by going through the idea behind this system and the requirements in which you will need(components will be provided). I'm going to try and make this as simple as possible to make it easier for traders to set up the system. Make sure you read the whole post.
So the system was developed using the platform TradeStation II, which is most commonly know throughout the retail FX community as being provided by F_XCM. So in order to trade this system, I highly recommend that you use this platform as it will make getting started much easier as well as algo management. I think there are a few other b_rokers who utilize this platform besides F_XCM, and as far as my knowledge goes I believe Global Futures and Barclay's retail FX do. I know Deutsche Bank used to but unfortunately they stopped. However, if you do choose to use a different platform such as MT4, you will need to gather and optimize the components, which I am sure are not too difficult to find.
Okay now that we have that out of the way let me describe the idea of this system. The one of the main features of this system is that it utilizes a longer time frame chart (8H, for reasons which I will describe later) which eliminates much noise of the market and shows the true tending and reversion capabilities. The main objective here is to spot a trending market and let the algos enter you into position(s). There are many other more advanced ways to use this system given its very high amount of flexibility, however to start I will only talk about the primary use.
Directional Bias
So the first let's look at how we determine our direction for the trade. To do this, we will be using the 8H chart with two indicators- DSS(double smoothed stochastic) which provides us with a slightly predictive edge, and a moving average to determine trend. For the moving average I typically use a Hull MA(150), however with the averages indicator I have attached you can make use of 20 different MA calculation methods. So to satisfy our trading criteria we will follow rules as such:
1. When price>selected MA, then long positions->DSS stream convergence in OS
2. When price< selected MA, then short positions -> DSS stream convergence in OB
*It is very important to note the interpretation of the DSS shapes. I typically look for smoother shapes, divergences, double/triple bounces and 3 drive patterns. As you use the DSS more, you will gain a better understanding of the shapes that the stochastic produces with respect to price action. It it extremely important to understand these relationships as they will keep you out of many undesirable conditions which you are likely to be stopped out. Most important of all, be patient.
Here is an example of the 8H chart setup on GBPUSD.
Risk Management
Okay, so we have the basic framework of the system, now onto the more important part: risk management and execution. With this system, the risk metrics are extremely flexible. So you can alter the parameter inputs of the execution algorithms to your risk tolerance and lot sizing. However, I have optimized the risk:reward for
1:3.07, so you should just focus on position sizing to manage your risk. the algorithms are set use 38 pip stops and 117 pip targets(like I said you can change this, but I would not suggest it). Generally speaking in terms of total Var in your portfolio, I do not suggest exposing yourself to more than 1% at any given time. For this reason, it adds and edge to the system, if you risk 1% per trade or basket(I will get into this method later) you can loose 100 times in a row before you are bust. 100 losers in a row with the system is very, very highly improbable. As I stated before, you should calculate your position numbers and size, and let the algos create your stops and limits, although this is an emergency risk precaution that I always have running.
Algorithmic Execution
Everything in this system is codependent. That is, as I said earlier, using the 8H chart is for a reason and that is this: (You may want to set your terminal to NY EST I believe it's 05:00 GMT) So we want to look closely at the 8H chart and evaluate our trade criteria at every candle close/open. The times on your 8H candles should be as follows 1:00a, 9:00a, 5:00p. So the algorithms can be seen on a 5m chart and should be attached to it for your reference. The execution algorithms are based upon a simple breakout indicator, that creates ranges based upon times you select. However, we will not be using this as a traditional breakout system. I have found, through a few years of market exposure, that most liquidity is created and consumed at the price levels developed during 3 specific periods of intraday ranging: 1:00a-1:02a, 9:30a-9:32a, 5:00p-5:05p. So, you should set the breakout indicator as follows:
1. London Session: Period begin-1:00 Period end 1:02 Box end- 00:59
2. New York Session: Period begin- 9:30 Period end 9:32 Box end- 9:29
3. New York Close: Period begin- 17:00 Period end 17:05 Box end- 16:59
Corresponding Execution Algos
1. London Liquidity
2. New York Liquidity
3. Market Close Liquidity
Here is a screenshot of what your 5m chart should look like. You can color the boxes so you are able to distinguish the different algos that you need to activate.The more you trade and get used to it, you should begin to see the patterns ie. where the liquidity is located that you need to consume.
Okay so now that we have all our charts set up, we are ready to load our algos. If you are not familiar with the TSII this will help. Once you are in the marketscope, on the upper tabs there is an option for "Alerts and Trading Automation". Select this and click "New Strategy or Alert". A box will pop up and select "Import" on the upper right hand side. then just go to wherever you downloaded the algo files and add them.
So let's say we are ready to enter a long position on a pair. We would want to initiate one(two, or three) of our execution algos. This is what it should look like when you initiate it.
So then we can let the algo enter us into the market and manage the position. If you feel it is necessary, you can move stops or limits at your discretion, however if the system is followed to a T, s/l should be kept as same. I would also suggest moving stop to breakeven when you reach around 55-60p profit. I am currently working on the execution algos so that I can integrate a staged trailing stop that will automatically do this.
As you can see, this strategy is very flexible in terms of risk management and goal seeking. There are many other ways in which this system can be traded given specific market conditions. The only precursor for the system to generate alpha is that the market has to be moving, preferably more volatile the better. Ideally we want to be in and out of the market as fast as possible(R/R unaltered). The real benefit of this is that we can loose 3 out of 4 trades and still breakeven(including commission, less rolls). So if the market is very volatile, but you are getting solid signals for execution to be initiated, so you turn on the algos and get stopped out on one trade, you still have the choice to let the same algo run the rest of the day to put you in a profitable trade. Vice versa, if you are getting weak signals, you can choose to only trade once per liquidity level per day.
*IMPORTANT NOTE
I have removed the model components from this post. I had them posted for a few weeks and I see they have been downloaded 280 times. Due to the lack of participation, which I asked from the people who have downloaded the system, I will no longer be providing the system components publicly. It seems that the overwhelming majority of people who downloaded this system were not interested in participating in the group. Not one chart has been posted. I see this as their loss, as I will be substantially improving and continually optimizing the algos throughout 2013. If you would like to join the group, and feel you are able to contribute ideas, want to learn a long term profitable system, or would like to help in algo development, please feel free to PM me or Yelena, and we will decide if you are a good fit for the group.
It is essential that you have experience trading and a solid understanding of how the FX market works. This is NOT a system for beginners or intermediate traders. This system is advanced and you must devote the time to learn how to use it. You MUST demo the system for a while until you get used to it, otherwise you will not understand how it works and you will loose money. I do not have the time to teach the basics of FX trading. Do not ask me in the thread "Where can I download this system?" I will ignore those questions and or ban you from the thread. This is not a get rich quick, "holy grail" scheme. If you think it is, move along nothing to see here. Combined with solid risk management and discipline, this is a consistent money maker over time and the equity model I will soon share utilizes the power of compounding interest to build your portfolio.
I fully expect everyone trying to use the system to contribute to this thread. It will help all who are trying to learn it. It will be very difficult to learn this system on your own, if you choose to do so, and you will probably loose money. Interact with the group, post charts, questions etc. I am constantly optimizing and changing this system as I see fit with the current market conditions, so it would behoove you to stay up to date. You can join a private group if you wish( all free), just contact Yelena for the details. In this private group I will be providing them with individualized video instructing on how to learn the system directly from the creator, me, so it will be much easier for you. I will also be adding the system updates within the group. Everything will be free with no costs.
No I will not provide any backtests, account statements or anything of that sort as the system is partly controlled by the trader, because it is very, very difficult to have a 100% automated system that works in all market conditions, therefore we cannot run a backtest right now on historical data. However, we can backtest short periods where we want to be in the market. So I guess if you went through historical data manually and found periods where you would initiate the algos then ran a backtest and compiled all of the periods of this in excel or something, then you probably could, although forward-looking may be a problem.
The reason for starting this thread is for me to give back to the FF community for the wealth of knowledge I have taken from it. I want to share this trading strategy and try to form a group here so that we can bounce ideas/problems/ algo /system adjustments etc. off of each other to enhance the profitability.
So the system was developed using the platform TradeStation II, which is most commonly know throughout the retail FX community as being provided by F_XCM. So in order to trade this system, I highly recommend that you use this platform as it will make getting started much easier as well as algo management. I think there are a few other b_rokers who utilize this platform besides F_XCM, and as far as my knowledge goes I believe Global Futures and Barclay's retail FX do. I know Deutsche Bank used to but unfortunately they stopped. However, if you do choose to use a different platform such as MT4, you will need to gather and optimize the components, which I am sure are not too difficult to find.
Okay now that we have that out of the way let me describe the idea of this system. The one of the main features of this system is that it utilizes a longer time frame chart (8H, for reasons which I will describe later) which eliminates much noise of the market and shows the true tending and reversion capabilities. The main objective here is to spot a trending market and let the algos enter you into position(s). There are many other more advanced ways to use this system given its very high amount of flexibility, however to start I will only talk about the primary use.
Directional Bias
So the first let's look at how we determine our direction for the trade. To do this, we will be using the 8H chart with two indicators- DSS(double smoothed stochastic) which provides us with a slightly predictive edge, and a moving average to determine trend. For the moving average I typically use a Hull MA(150), however with the averages indicator I have attached you can make use of 20 different MA calculation methods. So to satisfy our trading criteria we will follow rules as such:
1. When price>selected MA, then long positions->DSS stream convergence in OS
2. When price< selected MA, then short positions -> DSS stream convergence in OB
*It is very important to note the interpretation of the DSS shapes. I typically look for smoother shapes, divergences, double/triple bounces and 3 drive patterns. As you use the DSS more, you will gain a better understanding of the shapes that the stochastic produces with respect to price action. It it extremely important to understand these relationships as they will keep you out of many undesirable conditions which you are likely to be stopped out. Most important of all, be patient.
Here is an example of the 8H chart setup on GBPUSD.
Risk Management
Okay, so we have the basic framework of the system, now onto the more important part: risk management and execution. With this system, the risk metrics are extremely flexible. So you can alter the parameter inputs of the execution algorithms to your risk tolerance and lot sizing. However, I have optimized the risk:reward for
1:3.07, so you should just focus on position sizing to manage your risk. the algorithms are set use 38 pip stops and 117 pip targets(like I said you can change this, but I would not suggest it). Generally speaking in terms of total Var in your portfolio, I do not suggest exposing yourself to more than 1% at any given time. For this reason, it adds and edge to the system, if you risk 1% per trade or basket(I will get into this method later) you can loose 100 times in a row before you are bust. 100 losers in a row with the system is very, very highly improbable. As I stated before, you should calculate your position numbers and size, and let the algos create your stops and limits, although this is an emergency risk precaution that I always have running.
Algorithmic Execution
Everything in this system is codependent. That is, as I said earlier, using the 8H chart is for a reason and that is this: (You may want to set your terminal to NY EST I believe it's 05:00 GMT) So we want to look closely at the 8H chart and evaluate our trade criteria at every candle close/open. The times on your 8H candles should be as follows 1:00a, 9:00a, 5:00p. So the algorithms can be seen on a 5m chart and should be attached to it for your reference. The execution algorithms are based upon a simple breakout indicator, that creates ranges based upon times you select. However, we will not be using this as a traditional breakout system. I have found, through a few years of market exposure, that most liquidity is created and consumed at the price levels developed during 3 specific periods of intraday ranging: 1:00a-1:02a, 9:30a-9:32a, 5:00p-5:05p. So, you should set the breakout indicator as follows:
1. London Session: Period begin-1:00 Period end 1:02 Box end- 00:59
2. New York Session: Period begin- 9:30 Period end 9:32 Box end- 9:29
3. New York Close: Period begin- 17:00 Period end 17:05 Box end- 16:59
Corresponding Execution Algos
1. London Liquidity
2. New York Liquidity
3. Market Close Liquidity
Here is a screenshot of what your 5m chart should look like. You can color the boxes so you are able to distinguish the different algos that you need to activate.The more you trade and get used to it, you should begin to see the patterns ie. where the liquidity is located that you need to consume.
Okay so now that we have all our charts set up, we are ready to load our algos. If you are not familiar with the TSII this will help. Once you are in the marketscope, on the upper tabs there is an option for "Alerts and Trading Automation". Select this and click "New Strategy or Alert". A box will pop up and select "Import" on the upper right hand side. then just go to wherever you downloaded the algo files and add them.
So let's say we are ready to enter a long position on a pair. We would want to initiate one(two, or three) of our execution algos. This is what it should look like when you initiate it.
So then we can let the algo enter us into the market and manage the position. If you feel it is necessary, you can move stops or limits at your discretion, however if the system is followed to a T, s/l should be kept as same. I would also suggest moving stop to breakeven when you reach around 55-60p profit. I am currently working on the execution algos so that I can integrate a staged trailing stop that will automatically do this.
As you can see, this strategy is very flexible in terms of risk management and goal seeking. There are many other ways in which this system can be traded given specific market conditions. The only precursor for the system to generate alpha is that the market has to be moving, preferably more volatile the better. Ideally we want to be in and out of the market as fast as possible(R/R unaltered). The real benefit of this is that we can loose 3 out of 4 trades and still breakeven(including commission, less rolls). So if the market is very volatile, but you are getting solid signals for execution to be initiated, so you turn on the algos and get stopped out on one trade, you still have the choice to let the same algo run the rest of the day to put you in a profitable trade. Vice versa, if you are getting weak signals, you can choose to only trade once per liquidity level per day.
*IMPORTANT NOTE
I have removed the model components from this post. I had them posted for a few weeks and I see they have been downloaded 280 times. Due to the lack of participation, which I asked from the people who have downloaded the system, I will no longer be providing the system components publicly. It seems that the overwhelming majority of people who downloaded this system were not interested in participating in the group. Not one chart has been posted. I see this as their loss, as I will be substantially improving and continually optimizing the algos throughout 2013. If you would like to join the group, and feel you are able to contribute ideas, want to learn a long term profitable system, or would like to help in algo development, please feel free to PM me or Yelena, and we will decide if you are a good fit for the group.
It is essential that you have experience trading and a solid understanding of how the FX market works. This is NOT a system for beginners or intermediate traders. This system is advanced and you must devote the time to learn how to use it. You MUST demo the system for a while until you get used to it, otherwise you will not understand how it works and you will loose money. I do not have the time to teach the basics of FX trading. Do not ask me in the thread "Where can I download this system?" I will ignore those questions and or ban you from the thread. This is not a get rich quick, "holy grail" scheme. If you think it is, move along nothing to see here. Combined with solid risk management and discipline, this is a consistent money maker over time and the equity model I will soon share utilizes the power of compounding interest to build your portfolio.
I fully expect everyone trying to use the system to contribute to this thread. It will help all who are trying to learn it. It will be very difficult to learn this system on your own, if you choose to do so, and you will probably loose money. Interact with the group, post charts, questions etc. I am constantly optimizing and changing this system as I see fit with the current market conditions, so it would behoove you to stay up to date. You can join a private group if you wish( all free), just contact Yelena for the details. In this private group I will be providing them with individualized video instructing on how to learn the system directly from the creator, me, so it will be much easier for you. I will also be adding the system updates within the group. Everything will be free with no costs.
No I will not provide any backtests, account statements or anything of that sort as the system is partly controlled by the trader, because it is very, very difficult to have a 100% automated system that works in all market conditions, therefore we cannot run a backtest right now on historical data. However, we can backtest short periods where we want to be in the market. So I guess if you went through historical data manually and found periods where you would initiate the algos then ran a backtest and compiled all of the periods of this in excel or something, then you probably could, although forward-looking may be a problem.
The reason for starting this thread is for me to give back to the FF community for the wealth of knowledge I have taken from it. I want to share this trading strategy and try to form a group here so that we can bounce ideas/problems/ algo /system adjustments etc. off of each other to enhance the profitability.