Amibroker is reasonably affordable and will meet your requirements for backtesting index futures with 1 min data. However, I don't believe it will backtest based on constant volume bars. At least I have never found any options for backtesting based on constant volume bars, only for charting.
You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an alternative browser.
You should upgrade or use an alternative browser.
Backtesting Vs Trading Index Futures
- Thread starter lindq
- Start date
Amibroker is reasonably affordable and will meet your requirements for backtesting index futures with 1 min data. However, I don't believe it will backtest based on constant volume bars. At least I have never found any options for backtesting based on constant volume bars, only for charting.
I've spent 1000's of days manually back testing hundreds & hundreds of methods in my three decades of trading, and will always continue to strive to get better, to identify price action quicker. For every entry signal comes the most important aspect of the trade to me, HOW I identify the current price action for the trade that is not going instantly toward my targets. I don't hit the mouse and let my platform stop me out if I can read price and I can get out at breakeven.
I use no greater than 3-5 ticks in ES for stops unless volatility increases, I have watched five minute bars so long that on average depending on time of the day what each bars' range should be. I have learned from my back testing to break up the day in 30 minute segments and keep note of that range of each bar.
Double and triple bottoms/tops will hold the market, but if getting close to when Bond floor session closes, not to fade it.
If the market goes 4-5 ticks against me and not stopped out and comes back to entry, it is a gift to get out. I use time stops as well, if market won't go my way, guess what if I overstay my welcome?
So easy to get in, but my edge is the during. And always limit entries, no slippage and no extending my Ps.
When back testing, I use five minute bars and one tick charts, good sample size is 1000 trades, needless to say, takes me long time to want to change anything.
It might be the degree of ones' back testing whether forward testing/trading with be the same. Only by my microscoping each trade has shown me repeative patterns that allows me to take a breakeven than a sure loss most of the time. Saving just two lossing trades a day can take an "ok" method to an awesome method.
Good trading on Monday.
I've spent 1000's of hours manually backtesting dozens & dozens of methods. Of the few that showed a small edge none ever worked real time. I'm trading a method I developed live right now on the NQ. I manually backtested 9 months from July through March. It made money every month except February when it showed a small loss. I'm on my third week & have lost money almost every day? There was never that much of a drawdown during my testing. Guess I'll never get it?
I use no greater than 3-5 ticks in ES for stops unless volatility increases, I have watched five minute bars so long that on average depending on time of the day what each bars' range should be. I have learned from my back testing to break up the day in 30 minute segments and keep note of that range of each bar.
Double and triple bottoms/tops will hold the market, but if getting close to when Bond floor session closes, not to fade it.
If the market goes 4-5 ticks against me and not stopped out and comes back to entry, it is a gift to get out. I use time stops as well, if market won't go my way, guess what if I overstay my welcome?
So easy to get in, but my edge is the during. And always limit entries, no slippage and no extending my Ps.
When back testing, I use five minute bars and one tick charts, good sample size is 1000 trades, needless to say, takes me long time to want to change anything.
It might be the degree of ones' back testing whether forward testing/trading with be the same. Only by my microscoping each trade has shown me repeative patterns that allows me to take a breakeven than a sure loss most of the time. Saving just two lossing trades a day can take an "ok" method to an awesome method.
Good trading on Monday.
I've spent 1000's of hours manually backtesting dozens & dozens of methods. Of the few that showed a small edge none ever worked real time. I'm trading a method I developed live right now on the NQ. I manually backtested 9 months from July through March. It made money every month except February when it showed a small loss. I'm on my third week & have lost money almost every day? There was never that much of a drawdown during my testing. Guess I'll never get it?
Quote from Handle123:
If the market goes 4-5 ticks against me and not stopped out and comes back to entry, it is a gift to get out. I use time stops as well, if market won't go my way, guess what if I overstay my welcome?
So easy to get in, but my edge is the during. And always limit entries, no slippage and no extending my Ps.
When back testing, I use five minute bars and one tick charts, good sample size is 1000 trades, needless to say, takes me long time to want to change anything.
I have never used time stops, unless it was for EOD trading, and then usually the time stop is that an earnings announcement is getting too close. What duration do you use for time stops for intraday trading?
IronFist,
Yes I'm still using my method. April ended up a slight profit but with commissions was a slight loss. May has been back on track. April was a period the method had a hard time, much worse than any of the backtesting months showed.
I still manually backtest. I use Ninjatrader which has backtesting capabilities but I've never been able to figure it out. I plan to take their next paid backtesting course for nonprogrammers when/if they offer it. Wanted to take their last offering in April but was gone on vacation.
I've found that selecting what days & time of day not to trade greatly improves my results. I've gotten a handle on that from looking at my backtest results.
Yes I'm still using my method. April ended up a slight profit but with commissions was a slight loss. May has been back on track. April was a period the method had a hard time, much worse than any of the backtesting months showed.
I still manually backtest. I use Ninjatrader which has backtesting capabilities but I've never been able to figure it out. I plan to take their next paid backtesting course for nonprogrammers when/if they offer it. Wanted to take their last offering in April but was gone on vacation.
I've found that selecting what days & time of day not to trade greatly improves my results. I've gotten a handle on that from looking at my backtest results.
I have found for my style of trading whether monthly, weekly, daily or five minute bar, three bars works for me. In long term trading, I need to be at "breakeven" at end of three days cause when I sell futures, also buying calls as hedge and don't want to lose much time decay. But I don't buy puts when long futures, but still have three day rule. But even with intraday, my first target of 1.00, if market can't even go one point in three bars and going sideways, no need to linger.
I have never used time stops, unless it was for EOD trading, and then usually the time stop is that an earnings announcement is getting too close. What duration do you use for time stops for intraday trading?
I have never used time stops, unless it was for EOD trading, and then usually the time stop is that an earnings announcement is getting too close. What duration do you use for time stops for intraday trading?
Quote from IronFist:
In for more info. Are you still using this system? Did you figure out what was going on?
I'm manually backtesting stuff right now and it takes forever. I'm also looking for some good software (that is reasonably affordable) to backtest index futures with 1 min and volume candlestick data. Any suggestions?
Thanks.
The attached image shows an excel spreadsheet I created for backtesting ES.
Charts can be set to any time frame since they are created from tick data.
Instructions, tick data and the file are downloadable here.
(If "volume candlestick data" refers to constant volume bars, I have never used them but it would take me about 5 minutes to edit backtest.xls to create constant volume candlestick bars)
Attachments
bluedemon77
Guest
Quote from Retief:
Amibroker is reasonably affordable and will meet your requirements for backtesting index futures with 1 min data. However, I don't believe it will backtest based on constant volume bars. At least I have never found any options for backtesting based on constant volume bars, only for charting.
Yes it will, but you have to use a tick database, which is a problem because the database only holds 500,000 bars. I found a workaround solution which is to collect data in a 5 second database and then save it as a tick database. Then you can do volume based testing. The only problem is the amount of data is still limited (like 3 months), but it can at least give you an idea if your model is totally whacked or not.
Sorry I'm new to this backtesting thing.
Do I understand this correctly that after I buy backtesting software I then have to go buy data from someone else?
Are there any compatibility issues between certain backtesting software and certain brokers/data vendors?
In what format is data that one buys?
What is the average price?
Are there any reliability issues with any data sources?
How good at programming does one have to be to code a backtesting strategy?
Can I just go buy Inverstor RT or something and then start coding backtest strategies? Do backtesting applications require brokerage accounts?
I assume backtesting software shows all trades and exits on a chart, along with account balance after the trade, etc.
The only backtesting I've done has been at stockfetcher.com and they only let you do it with EOD data and there are still some goofy issues with it.
Do I understand this correctly that after I buy backtesting software I then have to go buy data from someone else?
Are there any compatibility issues between certain backtesting software and certain brokers/data vendors?
In what format is data that one buys?
What is the average price?
Are there any reliability issues with any data sources?
How good at programming does one have to be to code a backtesting strategy?
Can I just go buy Inverstor RT or something and then start coding backtest strategies? Do backtesting applications require brokerage accounts?
I assume backtesting software shows all trades and exits on a chart, along with account balance after the trade, etc.
The only backtesting I've done has been at stockfetcher.com and they only let you do it with EOD data and there are still some goofy issues with it.
Quote from IronFist:
Sorry I'm new to this backtesting thing.
Do I understand this correctly that after I buy backtesting software I then have to go buy data from someone else?
Are there any compatibility issues between certain backtesting software and certain brokers/data vendors?
In what format is data that one buys?
What is the average price?
Are there any reliability issues with any data sources?
How good at programming does one have to be to code a backtesting strategy?
Can I just go buy Inverstor RT or something and then start coding backtest strategies? Do backtesting applications require brokerage accounts?
I assume backtesting software shows all trades and exits on a chart, along with account balance after the trade, etc.
The only backtesting I've done has been at stockfetcher.com and they only let you do it with EOD data and there are still some goofy issues with it.
You need to get data from somewhere to backtest on. Amibroker has an IB plugin so if you have an account with IB, you can feed it with IB data at no additonal cost. The difficulty with IB data is that backfills are not very long, so you cannot backtest over a very long period unless you buy data elsewhere, such as from Esignal or IQfeed.
Some platforms come with data, such as Trade Station. Traders Studio has a deal where you get their platform plus one year of data for $1400.
If you're dead set on backtesting, my suggestion is to use very pessimistic assumptions, such as assuming that you buy at the high of the bar and sell at the low.