Testing A Forex/FX Strategy
How can back-testing help my trading?
(Forex-FX-4x.com) It is imperative that a trader has faith in the system they intend on trading.
Why is this?
If trust is not built through the testing process the trader could be tempted to deviate from the system when experiencing a prolonged drawdown. Back-testing can help the trader become familiar with the inevitable equity swings they will encounter and allows fine tuning (not curve fitting) at an early stage of the system development life cycle.
How do I back-test a discretionary strategy?
Back-testing can offer a great deal of useful statistical feedback related to a trading system and can be executed on a mechanical or discretionary basis. There is validity in both approaches and the system you are intending to test will determine which method you utilise. One benefit of manually testing a discretionary strategy is that, as mentioned above, you will gain a familiarity with the approach you intend on trading live.
The concept and initial principles of the system should be documented in full.
Let’s look at a simplified example (simple can sometimes be good…):
- Entry criteria. Break of the inside day high. Enter the trade when price moves 5 pips above the inside day.
- Exit criteria. Exit when a 30 pip trailing stop loss has been hit.
- Special notes. Only enter between 08:00 – 15:00 GMT Monday – Thursday.
The above simple rules could form the basis of testing. Make sure you test the system on different market conditions including ranging and trending periods. I also try to test on numerous currency pairs and have found that different strategies seem to work better on some currency pairs rather than others.
Document the testing results.
Here are some of the factors I document when conducting a manual back-testing session, in addition to the rules I specify as my primary system triggers:
- What was the optimal exit from each trade. In other words how far did the trade ultimately run.
- Would this system work better, over a series of trades, with a larger or smaller stop loss?
- How does the system perform in ranging markets?
- How does the system perform in trending markets?
- Is this system valid on all time frames?
- How is the system affected if I remove periods of low liquidity i.e. no trading during holiday periods.
- What is the maximum drawdown experienced?
- What was the biggest series of losses the system experienced?
- Does this trading strategy work well on some currency pairs and not on others?
- Any other notes?
Analyse the results.
Some of this information can be provided by the trading simulation software (if used) and some you can log in a spread sheet. The resulting data can be used to work out whether the system is right for you. After you have finished testing the following questions can be answered:
(You should ask as many questions as possible once you have gathered your back testing statistical information).
- Are you comfortable with the maximum drawdown?
- Would you keep your cool and stick with the strategy if experiencing the largest series of consecutive losses as per your back test data?
- If the system performs better on a lower time frame chart – could you find the time to trade it?
- How would you cope with seeing large profits turn into break-even trades? Think about this when considering trailing stop losses.
- How did event risk influence the results?
After testing, the average gain versus loss; combined with the win to loss ratio, can be used to help determine optimal position sizing if you subsequently take the system to forward testing.
One important point to note is that discretionary (non-mechanical) back testing can lead to a familiarity with the data you are testing. The back testing will never completely remove the need to forward test but I find it to be a great filter before I invest too much time (and potentially money).
What software do I use to test Forex strategies?
Some discretionary systems are so simple that they can be tested using software like metatrader. The user simply opens the platform and scrolls back to a given time. The f12 key is then used to progress through the charts. Other, more advanced, expert advisor indicators can also be used to test your Forex strategy.
My favoured software is “Forex Tester” as it very affordable compared to some of the high end software packages but still allows me to test a strategy reasonably quickly and gather some great stats. Forex tester allows traders to do their Forex training when the markets are closed. The current version is Forex tester 2.
See this Forex Tester review for further information.
If you do use a dedicated trading simulation software package it should provide the flexibility to customize the most relevant parameters. The software should allow modification of everything from the spread cost through to the daily rollover charges and much more as can be seen on the following images.
Forex Tester “Data center” gives access to important symbol properties.
The Forex tester platform allows modification of default buy and sell parameters.
Right clicking on the chart gives the following options. In this instance we want to add a simple sell order and stop loss.
The sell order and stop loss have been now been added.
Now the trading simulation position has been opened. A multitude of exit strategies can be utilised to exit the test trade.
The information above is not meant to be a detailed, all inclusive, guide to testing a Forex strategy. Hopefully some traders will benefit from the insight into how I approach testing and possibly spend some time testing their strategy in order to decide themselves if it is fit for purpose.