Backtesting strategies for Euro/USD FX

[jimmy1]

Hi,
I am looking at the EURO/USD FX future on globex as it seems to move a bit like the dax a year or 2 ago, and want to test some of my old stratgies on it....
do you know which time frame i should use for the trading day? as it is a 24hr mkt, but it does not move the same in the 24hrs due to volume diff's should i use 7.00 a.m. to 18.00 etc? all advice is appreciated as i feel if i back test it over 24hr it will not optisimse correctly

 

[JTrader]

Hi jimmy1

I can't answer your questions, but I do know that EUR/USD futures isn't quite a 24 hour market. Here is a link to the Euro FX contract specification, including trading hours.

http://www.cme.com/prd/overview_EC2465.html

jtrader.

 

[a_gnome]

Hi Jimmy1
It's difficult to give an answer to your question because you haven't said what sort of market characteristic your system is looking to exploit. If for example you require volatility then you could limit the trading to the most volatile trading times. Why not add "StartTime" and "EndTime" parameters to your strategy and then experiment with using different values to see which works the best.

Hope this helps.

 

[Newtron Bomb]

HI Jimmy

I beleive everything works some of the time on all TF's

15 mins may be a good a place to start as any. I would suggest that you first look at the tie frames and setups that you used when trading other instruments and see how the characteristic's differ.

The patterns will form on all time frames imo so it is really down to what time frames you want to trade with as they are all good.

Happy Trading
NB

 

[JTrader]

Hi jimmy1

Generally, the timescale of a chart helps determine several things.

One-minute charts will produce more frequent entry and exit signals than 5, 15, 30 or 60-minute charts. Although the number of signals will be to the same scale - the higher the timescale, the longer the time between trades - thus fewer trading opportunities within the course of a trading day. The longer the timescale, the longer the trades can be expected to last.

The timescale also determines the size of the profits achievable within individual trades. Trading signals produced on a 1 minute chart will produce opportunities to make smaller and quicker profits in terms of points profit, in comparison to 60 minute charts that will produce signals for longer lasting trades with bigger price movements.

Therefore the timescale of a chart will also determine the size of the stop-loss that should be used. The bigger the timescale of a chart, the bigger the price range within the individual time based price bars on the chart. When trading with a 1-minute chart, we should expect to trade more frequently within a trading day, and to collect smaller profits than when trading using 60-minute charts. Therefore a trader also needs to use bigger stop-losses when trading in bigger timescales, in order to retain the same proportion of trading opportunities (influenced by being stopped out too often) and the same proportion of winning to losing trades.

jtrader.

 

[jimmy1]

Hi jimmy1

Generally, the timescale of a chart helps determine several things.

One-minute charts will produce more frequent entry and exit signals than 5, 15, 30 or 60-minute charts. Although the number of signals will be to the same scale - the higher the timescale, the longer the time between trades - thus fewer trading opportunities within the course of a trading day. The longer the timescale, the longer the trades can be expected to last.

The timescale also determines the size of the profits achievable within individual trades. Trading signals produced on a 1 minute chart will produce opportunities to make smaller and quicker profits in terms of points profit, in comparison to 60 minute charts that will produce signals for longer lasting trades with bigger price movements.

Therefore the timescale of a chart will also determine the size of the stop-loss that should be used. The bigger the timescale of a chart, the bigger the price range within the individual time based price bars on the chart. When trading with a 1-minute chart, we should expect to trade more frequently within a trading day, and to collect smaller profits than when trading using 60-minute charts. Therefore a trader also needs to use bigger stop-losses when trading in bigger timescales, in order to retain the same proportion of trading opportunities (influenced by being stopped out too often) and the same proportion of winning to losing trades.

jtrader.

cheers j trader and all

 

[stevespray]

A number of traders appear to favour intaday breakout systems which use the 7am - 10am (uk time) as the initial range.

Hope this helps,
Steve.

 

[c6ackp]

Hi Jimmy1

Unless you are intending to trade completely automatically then don't test across 24hrs. Back-test for the time periods you intend to actively trade.

For example, you need to be careful with macd based signals in the Asian session as this session is often noisy, relatively non-directional and may give too many false signals.

IMHO 07-18 is correct for EUR and cable.

Regards
c6