After assessing the problems experienced last night, we have come to the conclusion that it would be best to move our live trading demo account to a more stable demo server. This will require those of you watching the demo trading live to download the software from an entirely different company. Here are the new instructions.
We chose to trade our demo account with Tradeview because our broker of choice, FXCM, only allows real money accounts and Tradeview has spreads similar to those that we expect to receive from FXCM. Unfortunately, Tradeview is a new Forex company and continues to have problems. On Friday for instance, the servers went down for 2 hours in the afternoon. It didn't have any effect on our trading because we had closed all but one trade by that point. However, with more problems last night drastically effecting our ability to accurately track the effectiveness of the strategies we decided it better to change companies early on rather than continue to fight these problems. We are putting in place error alerts so that if this were to happen on a live account server with real money the trader monitoring the accounts at that time would know instantly of the problem and would be able to resolve the error quickly.
Monday, June 30, 2008
Full System Live Testing Update
I woke up this morning to find that the auto trader had stopped working around 2:00AM for no apparent reason. Looking through the server log we get a connect error around that time and from then on out no new trades were entered and the trades we were in did not close as expected. I believe the bug has to do with the transition from the market close on Friday to the open on Monday. This is the first time we had run the program during the closed weekend period. I will reboot the auto trader and we should begin to get new trades again starting this evening. In the mean time, we will work to find the bug that caused the problem last night so that we won't have it happen again.
Friday, June 27, 2008
Full System Live Testing Update
The full system traded last night with only a few minor glitches. Due to an error on my part the GBP/JPY pair was taken out of trading early on. Also, the AUD/USD was not fully functional. Other than that, the system traded exactly as expected. The results were not real exciting as it turned out to be a rather slow Friday, but it was exciting to see the auto traders executing their orders per our desired settings. As you can see on your demo platform, at this very moment we still have 6 trades open and the account equity is sitting right at $100,100. As soon as the day has officially ended I will post the results. (Here are the instructions for watching the account in real time).
UPDATE (4:02pm EST): Here is the updated trade report for those not watching the demo live. There remains one open trade with a floating profit of $750.50 that will not close until Monday. As of this minute, the account has a closed balance of $99,868.02 and equity of $100,618.52.
UPDATE (4:02pm EST): Here is the updated trade report for those not watching the demo live. There remains one open trade with a floating profit of $750.50 that will not close until Monday. As of this minute, the account has a closed balance of $99,868.02 and equity of $100,618.52.
Thursday, June 26, 2008
Full System Live Testing Update
I activated each strategy and the corresponding auto trader last night. The 10-Pair Intraday ended the day up 0.12%. The EUR/USD and GBP/USD auto traders, however, did not execute their trades. This is disappointing considering both would have had sizable wins. We are assessing the problem and hope to have it fixed in time for tonight's session.
If you have not downloaded the platform to watch the trades in real time, you can do so by following the instructions here.
If you would prefer not to download the platform you can see the updated trade report here.
If you have not downloaded the platform to watch the trades in real time, you can do so by following the instructions here.
If you would prefer not to download the platform you can see the updated trade report here.
Wednesday, June 25, 2008
Watch the system run on a live demo account!
As you know we have been live testing the 10-pair Intraday strategy for nearly two weeks and the EUR/USD strategy for 5 days now and both seem to be running as we had hoped. We have also reached a point in the programming of the GBP/USD auto trader that it is ready to begin it's live testing period. I thought all of you might find it useful to have access to the demo account trading the system as a whole. The account will be running live starting tonight at midnight MST. Keep in mind, there will continue to be changes made daily to the different strategies as we find ways to make them more efficient and profitable.
Update: With the problems we have had with Tradeview's servers, we have opted to change our live demo trading to Alpari-US. As you know, we have been approved with FXCM for our actual real money trading but they do not give the option of trading demo accounts.
INSTRUCTIONS TO OPEN AND VIEW DEMO ACCOUNT (UPDATED JUNE 30th):
1) Download Alpari-US Metatrader 4 Platform here.
2) It will ask you if you want to save MT4setup.exe. Click Save File. Save the file to a directory where you will be able to find it easily.
3) Once the file is downloaded, open the file from the directory you saved it in. Complete all necessary steps for installing the MT4 platform.
4) Open the Alpari-US Metatrader 4 program.
5) A window will pop up allowing you to 'Open an Account'. You do NOT want to open a new account so click 'Cancel'.
6) The Metatrader 4 program may ask you if you want to update. Click yes and wait for the update to complete and restart. If not, proceed to the next step.
7) Click the 'File' menu in the upper left-hand corner, move down and select 'Login'
8) A window will pop open asking for the 'Login', 'Password', and 'Server'. Insert:
Login: 56289
Password: zosk1ul
Server: AlpariUS-Demo
9) Once you are logged into the demo account. You can select 'Terminal' from the 'View' menu to look at the Open Trades as well as the Account History. You can also open the following currency charts to see the trades in action.
GBP/USD; EUR/USD; USD/JPY; USD/CHF; GBP/EUR; GBP/JPY; GBP/CHF; EUR/JPY; EUR/CHF; CHF/JPY; AUD/USD
10) To learn more about the Metatrader software and how to use it, download the user guide here.
Update: With the problems we have had with Tradeview's servers, we have opted to change our live demo trading to Alpari-US. As you know, we have been approved with FXCM for our actual real money trading but they do not give the option of trading demo accounts.
INSTRUCTIONS TO OPEN AND VIEW DEMO ACCOUNT (UPDATED JUNE 30th):
1) Download Alpari-US Metatrader 4 Platform here.
2) It will ask you if you want to save MT4setup.exe. Click Save File. Save the file to a directory where you will be able to find it easily.
3) Once the file is downloaded, open the file from the directory you saved it in. Complete all necessary steps for installing the MT4 platform.
4) Open the Alpari-US Metatrader 4 program.
5) A window will pop up allowing you to 'Open an Account'. You do NOT want to open a new account so click 'Cancel'.
6) The Metatrader 4 program may ask you if you want to update. Click yes and wait for the update to complete and restart. If not, proceed to the next step.
7) Click the 'File' menu in the upper left-hand corner, move down and select 'Login'
8) A window will pop open asking for the 'Login', 'Password', and 'Server'. Insert:
Login: 56289
Password: zosk1ul
Server: AlpariUS-Demo
9) Once you are logged into the demo account. You can select 'Terminal' from the 'View' menu to look at the Open Trades as well as the Account History. You can also open the following currency charts to see the trades in action.
GBP/USD; EUR/USD; USD/JPY; USD/CHF; GBP/EUR; GBP/JPY; GBP/CHF; EUR/JPY; EUR/CHF; CHF/JPY; AUD/USD
10) To learn more about the Metatrader software and how to use it, download the user guide here.
Live Testing Results - Intraday Strategy
Here are the results through today, June 25th. Up 2.4% in 8 trading days. It was a pretty flat day in the market as market players waited to see what the Fed would do with interest rates at their monthly meeting today.
Tuesday, June 24, 2008
Live Testing Results - Intraday Strategy
Here are the results through today. Up 2.1% since we started 7 trading days ago.
Monday, June 23, 2008
Live Testing Results - Intraday Strategy
Here are the results for the 10-Pair Intraday strategy through today. Up a little over 1.5% since June 16th.
GBP/USD Strategy Update
The GBP/USD is nearly debugged. It is to a point where I can backtest it and so I have been doing so. I have tested since 2005 with promising results. Here are the reports for the last 12 months and since the beginning of 2008.
There are some changes that are being programmed that will adjust the results some. As soon as those changes are complete I will post updated reports. We will also beginning live testing this auto trader later this week.
There are some changes that are being programmed that will adjust the results some. As soon as those changes are complete I will post updated reports. We will also beginning live testing this auto trader later this week.
Possible EUR/USD Strategy
While researching different money management plans used by other traders, I came across a EUR/USD strategy that really intrigued me. I have discussed this strategy with the other 3 and we are strongly considering using it as part of our overall system. Again, this strategy is NOT mine. I tend to shy away from using other people's strategies for one main reason. Every strategy works because of certain fundamental factors. And when the market changes such that those fundamentals no longer have the same effect, the strategy will stop working. If I am using someone else's strategy, I may not know what those fundamental factors nor be able to make parameter adjustments to meet the changing market. However, this EUR/USD strategy is extremely simple and I know the reason why it works and why it doesn't work. I believe adding it to the mix would help diversify our system to be profitable across multiple market types.
Here are the preliminary backtest results for 2008 and for the last 12 months.
Here are the preliminary backtest results for 2008 and for the last 12 months.
Friday, June 20, 2008
Live Testing Results - Intraday Strategy
Here are the results of the entire week. Finished up about 1/2%.
Thursday, June 19, 2008
Live Testing Results - Intraday Strategy
Above breakeven for the week again. Here is the report through today.
Wednesday, June 18, 2008
Progress Report (WITH UPDATES)
The Intraday 10-Pair Mini Trending Strategy has had two losing days in a row now due to a flat market. As we had noted in previous posts, the problem with this strategy is that it depends on mini-trends to win and we have seen very few in the last two days. Our overall direction is to line up several strategies that win in varying market conditions. The GBP/USD strategy that relies on entirely different market conditions has been programmed and we are currently working on it's unique and exciting new money management plan. We have also done some preliminary programming on a version of our Intraday strategy that will utilize 6 pairs to win during ranging markets such as yesterday and today. That strategy would most likely only be traded when both the European and U.S. markets are closed since it would not fair well during trending periods.
So here is a summary of our progress thus far.
1) Intraday 10 Pair Mini-Trend Strategy - Programmed; Debugged; Backtested; Being live demo tested currently
2) Intraday 6 Pair Ranging Strategy - 90% Programmed;
3) GBP/USD Strategy - Programmed; In the process of being debugged; Adding new money management plan
UPDATE (July 1, 2008) - We have since added a daily EUR/USD strategy that takes about one trade per month. The Intraday 10 Pair is ready to go. The Intraday 6 pair ranging strategy is still in development and is unlikely to be traded live within the next 60 days. The GBP/USD Strategy is programmed, debugged, backtested, and being tested live alongside the EUR/USD Daily and the Intraday 10 Pair.
So here is a summary of our progress thus far.
1) Intraday 10 Pair Mini-Trend Strategy - Programmed; Debugged; Backtested; Being live demo tested currently
2) Intraday 6 Pair Ranging Strategy - 90% Programmed;
3) GBP/USD Strategy - Programmed; In the process of being debugged; Adding new money management plan
UPDATE (July 1, 2008) - We have since added a daily EUR/USD strategy that takes about one trade per month. The Intraday 10 Pair is ready to go. The Intraday 6 pair ranging strategy is still in development and is unlikely to be traded live within the next 60 days. The GBP/USD Strategy is programmed, debugged, backtested, and being tested live alongside the EUR/USD Daily and the Intraday 10 Pair.
Tuesday, June 17, 2008
Monday, June 16, 2008
Live Testing Results - Intraday Strategy
We finished the lot size/cash management engines for the Intraday Strategy auto trader last week and now the program is running flawlessly. I have opened a new and separate account and will run that account non-stop all this week posting the daily results as they come in. Attached is the link to the results for today. Keep in mind, the updated GBP/USD strategy is not completely debugged yet and so I will not be posting those results until it is running as it should.
June 16, 2008 Statement
Note: This demo account (fake money traded in real market time with real market pricing) is at Rosenthal Collins/Tradeview Forex because FXCM does not have the ability to run demo accounts yet. I would post live results with real money but in order to accurately gauge money management, this system needs to be run on a minimum of $50,000 and I am not comfortable running it on $50,000 live just yet.
June 16, 2008 Statement
Note: This demo account (fake money traded in real market time with real market pricing) is at Rosenthal Collins/Tradeview Forex because FXCM does not have the ability to run demo accounts yet. I would post live results with real money but in order to accurately gauge money management, this system needs to be run on a minimum of $50,000 and I am not comfortable running it on $50,000 live just yet.
Thursday, June 12, 2008
Forward Testing Update
We have been running the auto trader, without stopping, for over a week. We have found numerous additional errors that have been corrected. We will continue this process until we have one week straight without any bugs, problems, or errors. We have had some difficulty calculating lot sizes for the 10 different currency pairs we are testing and so our results are not accurate. In spite of the bugs/errors, inaccurate lot size orders, etc, the system has been profitable over the last week. I am sorry I don't have more to report but at this point it is a matter of allowing the live market to work out the remaining problems.
Saturday, June 7, 2008
Balance vs. Equity - Closing accounts prematurely
I have included, in previous posts, about a dozen links to the standard print off that our auto trader gives when it completes a backtest. Included with that report is a graph showing the movement of the results daily balance. I want to take a moment and make sure everyone understands how to read the graph. While the graph for the most part is very straightforward, there are a couple aspects worth noting including the balance line vs equity line. So here is a basic run down of how to read and understand the graph.
(Click on the image to enlarge it)
Attached is a graph representing the balance changes of the GBP/JPY Daily extension strategy from December 6, 2007 - April 6, 2008. The numbers on the right represent the overall balance, the numbers on the bottom represent the trade transaction number. You can find the actual trade numbers on the report list and use that number to match up the date. There are two lines, the blue balance line, and the green equity line. I really wanted to write this post so that I could point out that green equity line. If we decide to implement this strategy into the overall system, we will have times when our balance is higher than our equity. What that means, is that we have trades open that are in the red that if closed, would shrink our balance to the equity figure. Our current strategy does not have trades open overnight and so this really has not been a factor. If we choose to add this strategy, we all need to understand that there will be times when we have losing trades open for weeks. If someone chooses to close their account, that losing trade would be closed prematurely and the balance would shrink to that equity number.
For comparison sake, here is a graph of our Intraday strategy that does NOT carry positions overnight. Notice no green equity line.
(Click on the image to enlarge it)
Attached is a graph representing the balance changes of the GBP/JPY Daily extension strategy from December 6, 2007 - April 6, 2008. The numbers on the right represent the overall balance, the numbers on the bottom represent the trade transaction number. You can find the actual trade numbers on the report list and use that number to match up the date. There are two lines, the blue balance line, and the green equity line. I really wanted to write this post so that I could point out that green equity line. If we decide to implement this strategy into the overall system, we will have times when our balance is higher than our equity. What that means, is that we have trades open that are in the red that if closed, would shrink our balance to the equity figure. Our current strategy does not have trades open overnight and so this really has not been a factor. If we choose to add this strategy, we all need to understand that there will be times when we have losing trades open for weeks. If someone chooses to close their account, that losing trade would be closed prematurely and the balance would shrink to that equity number.
For comparison sake, here is a graph of our Intraday strategy that does NOT carry positions overnight. Notice no green equity line.
Friday, June 6, 2008
System Extensions Testing
As I had mentioned in an earlier post, we are working on a few system extensions that would hopefully smooth out the profits vs losses. I spent most of the day testing the auto trader created for our potential GBP/JPY daily time frame extension. The exciting thing about this extension is that it is NOT as heavily influenced by changes in markets. Using the same triggers over the last two years it has consistently returned profits. Keep in mind, this is an extension using a daily timeline. What that means is that while the majority of our trades are closed within just a few hours, this extension takes trades based off of daily signals and includes trades that are open for weeks and sometimes months. It has not been decided whether we will trade this extension. While the past returns are impressive, the downsides of having trades rolling over from one month to the next may out way the positives of a extension such as this one.
The extension trades the GBP/JPY. Here is the 6-month past performance, and the 24-month past performance.
The extension trades the GBP/JPY. Here is the 6-month past performance, and the 24-month past performance.
Thursday, June 5, 2008
Detailed Intraday Backtest Results
I have found a way to post the HTML files that contain the detailed data for the results of each of the 10 intraday pairs from December 1, 2007 - May 31, 2008. Click the pair you wish to see below to be redirected to the link containing each individual report.
EUR/USD
GBP/USD (Intraday strategy only)
USD/CHF
USD/JPY
EUR/GBP
EUR/CHF
EUR/JPY
GBP/CHF
GBP/JPY
CHF/JPY
Some things to note about the back test results:
First. Take a look at the Past/Hypothetical Performance Alerts post again. Past results do not guarantee future performance. It will require constant assessment of the triggers to keep the system profitable.
Second. The back testing was done using a different broker(Alpari - UK) other than FXCM. The reason for this is that FXCM does not have precise data going back further than 30 days. This should not have a significant impact on our results but I wanted everyone to be aware of that fact. Alpari's spreads are higher than those being quoted by FXCM and so, in theory, the results would have been better had FXCM been able to provide past data.
Third. While the strategy as a whole is based off of the same specific indicators/signals, the way these indicators are used is distinct for each pair depending on the tendencies of that pair. For instance, the GBP/JPY is more likely to trend (thus be profitable) than the EUR/USD from 4:00 - 8:00am London time because the GBP/JPY is more heavily traded during the Japanese session. Therefore, the GBP/JPY is traded during that time while the EUR/USD is not. Slight variations have been attributed to each pair so as to maximum that pair's potential.
EUR/USD
GBP/USD (Intraday strategy only)
USD/CHF
USD/JPY
EUR/GBP
EUR/CHF
EUR/JPY
GBP/CHF
GBP/JPY
CHF/JPY
Some things to note about the back test results:
First. Take a look at the Past/Hypothetical Performance Alerts post again. Past results do not guarantee future performance. It will require constant assessment of the triggers to keep the system profitable.
Second. The back testing was done using a different broker(Alpari - UK) other than FXCM. The reason for this is that FXCM does not have precise data going back further than 30 days. This should not have a significant impact on our results but I wanted everyone to be aware of that fact. Alpari's spreads are higher than those being quoted by FXCM and so, in theory, the results would have been better had FXCM been able to provide past data.
Third. While the strategy as a whole is based off of the same specific indicators/signals, the way these indicators are used is distinct for each pair depending on the tendencies of that pair. For instance, the GBP/JPY is more likely to trend (thus be profitable) than the EUR/USD from 4:00 - 8:00am London time because the GBP/JPY is more heavily traded during the Japanese session. Therefore, the GBP/JPY is traded during that time while the EUR/USD is not. Slight variations have been attributed to each pair so as to maximum that pair's potential.
GBP/USD Strategy Upgrade Update
For those wondering how our autotrader for the GBP/USD strategy is coming along, here is the latest. The strategy has been programmed but needs debugging. Now that we have begun running the Intraday live testing, we will have more time to finish up the GBP/USD upgrades and get it testing live as well.
Month by Month Intraday Strategy Backtest data
Here is the attached, month-by-month, results of the individual tests done of each pair. We have cut the number of tradeable pairs to 10 due to overly high spreads on the other 3. Keep in mind, this does NOT include the upgraded GBP/USD which still has a little work left before it will be ready to backtest. (UPDATE: Here is the link to the post containing the actual detailed data.)
| Currency | Starting Balance | Ending Balance | Profit | Return % | |
| Dec | EUR/USD | $100,000.00 | $103,912.00 | $3,912.00 | 3.91% |
| Dec | GBP/USD | $100,000.00 | $101,879.00 | $1,879.00 | 1.88% |
| Dec | USD/CHF | $100,000.00 | $102,068.00 | $2,068.00 | 2.07% |
| Dec | USD/JPY | $100,000.00 | $99,932.00 | -$68.00 | -0.07% |
| Dec | EUR/GBP | $100,000.00 | $99,955.00 | -$45.00 | -0.05% |
| Dec | EUR/CHF | $100,000.00 | $102,400.00 | $2,400.00 | 2.40% |
| Dec | EUR/JPY | $100,000.00 | $105,162.00 | $5,162.00 | 5.16% |
| Dec | GBP/CHF | $100,000.00 | $100,674.00 | $674.00 | 0.67% |
| Dec | GBP/JPY | $100,000.00 | $102,626.00 | $2,626.00 | 2.63% |
| Dec | CHF/JPY | $100,000.00 | $100,809.00 | $809.00 | 0.81% |
| Month End Total | $19,417.00 | 19.42% | |||
| Currency | Starting Balance | Ending Balance | Profit | Return % | |
| Jan | EUR/USD | $103,912.00 | $103,740.00 | -$172.00 | -0.17% |
| Jan | GBP/USD | $101,879.00 | $103,354.00 | $1,475.00 | 1.45% |
| Jan | USD/CHF | $102,068.00 | $101,471.00 | -$597.00 | -0.58% |
| Jan | USD/JPY | $99,932.00 | $101,526.00 | $1,594.00 | 1.60% |
| Jan | EUR/GBP | $99,955.00 | $103,247.00 | $3,292.00 | 3.29% |
| Jan | EUR/CHF | $102,400.00 | $100,969.00 | -$1,431.00 | -1.40% |
| Jan | EUR/JPY | $105,162.00 | $108,297.00 | $3,135.00 | 2.98% |
| Jan | GBP/CHF | $100,674.00 | $100,553.00 | -$121.00 | -0.12% |
| Jan | GBP/JPY | $102,626.00 | $109,332.00 | $6,706.00 | 6.53% |
| Jan | CHF/JPY | $100,809.00 | $102,113.00 | $1,304.00 | 1.29% |
| Month End Total | $15,185.00 | 14.61% | |||
| Currency | Starting Balance | Ending Balance | Profit | Return % | |
| Feb | EUR/USD | $103,740.00 | $106,560.00 | $2,820.00 | 2.72% |
| Feb | GBP/USD | $103,354.00 | $101,769.00 | -$1,585.00 | -1.53% |
| Feb | USD/CHF | $101,471.00 | $101,709.00 | $238.00 | 0.23% |
| Feb | USD/JPY | $101,526.00 | $101,656.00 | $130.00 | 0.13% |
| Feb | EUR/GBP | $103,247.00 | $104,457.00 | $1,210.00 | 1.17% |
| Feb | EUR/CHF | $100,969.00 | $101,096.00 | $127.00 | 0.13% |
| Feb | EUR/JPY | $108,297.00 | $109,301.00 | $1,004.00 | 0.93% |
| Feb | GBP/CHF | $100,553.00 | $104,655.00 | $4,102.00 | 4.08% |
| Feb | GBP/JPY | $109,332.00 | $110,965.00 | $1,633.00 | 1.49% |
| Feb | CHF/JPY | $102,113.00 | $101,606.00 | -$507.00 | -0.50% |
| Month End Total | $9,172.00 | 8.84% | |||
| Currency | Starting Balance | Ending Balance | Profit | Return % | |
| Mar | EUR/USD | $106,560.00 | $109,521.00 | $2,961.00 | 2.78% |
| Mar | GBP/USD | $101,769.00 | $104,019.00 | $2,250.00 | 2.21% |
| Mar | USD/CHF | $101,709.00 | $100,449.00 | -$1,260.00 | -1.24% |
| Mar | USD/JPY | $101,656.00 | $103,239.00 | $1,583.00 | 1.56% |
| Mar | EUR/GBP | $104,457.00 | $108,169.00 | $3,712.00 | 3.55% |
| Mar | EUR/CHF | $101,096.00 | $100,869.00 | -$227.00 | -0.22% |
| Mar | EUR/JPY | $109,301.00 | $111,905.00 | $2,604.00 | 2.38% |
| Mar | GBP/CHF | $104,655.00 | $105,897.00 | $1,242.00 | 1.19% |
| Mar | GBP/JPY | $110,965.00 | $112,876.00 | $1,911.00 | 1.72% |
| Mar | CHF/JPY | $101,606.00 | $99,156.00 | -$2,450.00 | -2.41% |
| Month End Total | $12,326.00 | 11.57% | |||
| Currency | Starting Balance | Ending Balance | Profit | Return % | |
| Apr | EUR/USD | $109,521.00 | $114,896.00 | $5,375.00 | 4.91% |
| Apr | GBP/USD | $104,019.00 | $104,259.00 | $240.00 | 0.23% |
| Apr | USD/CHF | $100,449.00 | $106,210.00 | $5,761.00 | 5.74% |
| Apr | USD/JPY | $103,239.00 | $103,155.00 | -$84.00 | -0.08% |
| Apr | EUR/GBP | $108,169.00 | $109,602.00 | $1,433.00 | 1.32% |
| Apr | EUR/CHF | $100,869.00 | $106,877.00 | $6,008.00 | 5.96% |
| Apr | EUR/JPY | $111,905.00 | $113,336.00 | $1,431.00 | 1.28% |
| Apr | GBP/CHF | $105,897.00 | $109,474.00 | $3,577.00 | 3.38% |
| Apr | GBP/JPY | $112,876.00 | $117,885.00 | $5,009.00 | 4.44% |
| Apr | CHF/JPY | $99,156.00 | $99,264.00 | $108.00 | 0.11% |
| Month End Total | $28,858.00 | 26.35% | |||
| Currency | Starting Balance | Ending Balance | Profit | Return % | |
| May | EUR/USD | $114,896.00 | $117,058.00 | $2,162.00 | 1.88% |
| May | GBP/USD | $104,259.00 | $107,544.00 | $3,285.00 | 3.15% |
| May | USD/CHF | $106,210.00 | $109,154.00 | $2,944.00 | 2.77% |
| May | USD/JPY | $103,155.00 | $102,592.00 | -$563.00 | -0.55% |
| May | EUR/GBP | $109,602.00 | $108,055.00 | -$1,547.00 | -1.41% |
| May | EUR/CHF | $106,877.00 | $107,297.00 | $420.00 | 0.39% |
| May | EUR/JPY | $113,336.00 | $114,984.00 | $1,648.00 | 1.45% |
| May | GBP/CHF | $109,474.00 | $109,754.00 | $280.00 | 0.26% |
| May | GBP/JPY | $117,885.00 | $121,003.00 | $3,118.00 | 2.64% |
| May | CHF/JPY | $99,264.00 | $99,712.00 | $448.00 | 0.45% |
| Month End Total | $12,195.00 | 10.61% |
Intraday Strategy Back Testing - Monthly Balance
A challenge I have had in posting the 6 month backtest results is that I can not post non-image files on this site. I am working on finding a way to publish the results (held in html files) on separate webpages and then providing links to those. Also, our back tester does not allow us to test all pairs simultaneously which means that to get a monthly picture of all pairs together, I have to transfer the data from each individual pair to a separate spreadsheet that calculates for me the monthly results of the system as a whole. I have attached below a graph I built that shows the ending balance of the Intraday strategy upgrades collectively. Keep in mind, ending balances would be slightly different(higher) if the 10 pairs were tested together because of the built-in compounding feature of our auto trader. In the following posts, I will include more detailed testing results from our 6 month backtest (December 1, 2007 - May 31, 2008).
| Ending Balance | |
| November | $100,000.00 |
| December | $119,417.00 |
| January | $134,602.00 |
| February | $143,774.00 |
| March | $156,100.00 |
| April | $184,958.00 |
| May | $197,153.00 |
15 day game plan
So now that the upgrades are in place and the autotrader is working, what next? Tradewinds Currency was approved this morning by FXCM to manage customer accounts. I hesitate to have clients begin setting up their accounts with FXCM until we have traded the auto trader live, with our own personal money, for a couple of weeks so that we know what issues we might face dealing with a different brokerage. So, for the next 15 days, sit tight. I will be posting daily result updates from our live, executed trades in the real market. I will also post today the previous six months worth of trade result as well as the not as impressive year prior.
Backtesting the Autotrader
I have completed a thorough backtest of the last 6 months (December 1, 2007 - May 31, 2008). I will post the results today. I find the results on one hand exciting and on the other hand troubling. While the system did extraordinarily well in the previous six months, it did not perform as well in the year previous to that. Of course, the reason we have made changes to the original system is that we have noted a distinct difference in how the market behaves recently compared to last year. These results show that while the system has great potential for growth, it will not perform consistently through all market types (no system I know can). I am working on an extension to the upgraded system that would potentially hedge against the scenario of a changing market.
Debugging Update
Well, it looks like the auto trader has been officially debugged. While we are not 100% sure that it is free of bugs/quirks, it is running successfully and efficiently. It is executing the trades exactly as desired and seems to be amazingly stable considering the complexities of the code and the number of orders occurring simultaneously.
Tuesday, June 3, 2008
Debugging Update
The debugging continues. As with any computer program, every time it seems you have it fixed, another bug arises. We thought we had it 100% ready last night, ran it live on all 13 pairs, and it had problems closing out trades. The entries were executed perfectly but the exit orders were not always accepted by the broker's server. We believe we have identified the problem, made the necessary changes, and now will see how it performs tonight.
The autotrader performs perfectly in backtesting. Because it does not require live server communication it is now completely free of errors when backtesting. So, we have begun the work of backtesting the last 6 months to get a feel for how accurate the system is across multiple months and market types. We already have identified a concern during ranging months that could hurt its ability to be consistent. I have an idea that I believe will allow it to make money even during those periods. We have created an extension to the autotrader to test this theory and if it proves to be worthwhile, I will unveil it to the group.
Because of the massive processing power it is taking to run the autotrader on 13 pairs across multiple timeframes 24/7, we are experiencing some drag on order processing and entries. We are weighing our options and considering placing the autotrader on a high power server located on the same network as the broker we will be trading through. It is either that, or network multiple computers together to utilize the processing power of the computers collectively.
The autotrader performs perfectly in backtesting. Because it does not require live server communication it is now completely free of errors when backtesting. So, we have begun the work of backtesting the last 6 months to get a feel for how accurate the system is across multiple months and market types. We already have identified a concern during ranging months that could hurt its ability to be consistent. I have an idea that I believe will allow it to make money even during those periods. We have created an extension to the autotrader to test this theory and if it proves to be worthwhile, I will unveil it to the group.
Because of the massive processing power it is taking to run the autotrader on 13 pairs across multiple timeframes 24/7, we are experiencing some drag on order processing and entries. We are weighing our options and considering placing the autotrader on a high power server located on the same network as the broker we will be trading through. It is either that, or network multiple computers together to utilize the processing power of the computers collectively.
Subscribe to:
Posts (Atom)