The Best Forex Trading Systems
It has been over a year now since I first started this website; during that time I have developed, tested and posted several trading systems and conducted a number of tests and experiments for some of the various articles here. The results of these tests and experiments have lead me to some (fairly) settled conclusions as to what works, what doesn't work, and what the best Forex trading systems have in common. In this article I'm going to share with you what I think the characteristics and attributes of the best Forex Trading Systems are, why these trading systems work and how to create your own one and follow it successfully.
Forex Trading Systems: What works and why,
The Edge...
In order to have any chance whatsoever of making a profit, a trading system must have an edge. An edge is simply a set of definable circumstance(s) that when they are present in the market have a statistical significance and are therefore likely to result in an outcome with a better than average probability of success. If a trading system's author can't easily define their system's edge and explain exactly why it works, then the chances are their system doesn't actually have an edge and is worthless. System developers try and gain an edge in a number of different ways. The most common are –
After experimenting with these different methods in various systems and researching their effectiveness for many articles, I have come to the conclusion that the market's underlying long-term trend provides by far the greatest edge. There are a many methods that are used to detect the direction of the trend, but the best ones I have come across by far are momentum and breakouts. Breakouts tend to be significant at around 4-6 months; that is when the price is making higher highs or lowers lows that have not been seen for around 4-6 months then it's a fairly reliable signal of the start of a new trend, any less than 4 months and the signal's accuracy is seriously degraded. Momentum becomes very significant when there are weekly movements so large that they're only typically seen a few times a year – this is also a fairly reliable signal that we're witnessing the start of a new trend. Whilst momentum is a very effective method of detecting the start of a new trend I still, on balance, prefer to use breakouts. It should probably also be noted that using retail Forex sentiment as a contrarian indicator is also very effective. But unlike market price data, retail Forex sentiment data is proprietary and may not be available in the future.
I will say that this then is the first characteristic of the best Forex trading systems. The best technical/mechanical Forex trading systems are trend following systems - they get their edge from the trend. And the best way of using market's price to detect a trend is with Donchian channel type breakouts.
The Time Frame...
When I first tested trading using different time frames the results were stunning, they revealed that a direct correlation exists between the trading time frame and the results one can reasonably expect. The shorter the trading time frame the greater the randomness and therefore the less reliable the trade and the less profit one can expect. When you are looking at a trading chart composed of daily and weekly candles you'll generally notice some random 'noise' attached to the longer underlying trend(s). However, when you're looking at a short time frame such as charts made of 5, 10 or 15 minute candles one will generally observe almost nothing but the market's random noise. The picture below illustrates the difference between a chart made of daily candles and a chart made of 5-minute candles quite nicely -
This then I will say is the second characteristic of the best Forex trading systems. The best technical/mechanical Forex trading systems use long-term time frames. By using a long-term trading system where trades typically last from anything to a few weeks to a few months (or more) the edge is given a chance to work and take your trade into profit. By entering and exiting every five minutes the results will tend to be more random than anything else and the trading costs of such short-term trading is huge – when trades last several weeks or more the cost of the spread is fairly negligible.
Stop-Losses...
Many traders believe that they can consistently pick market tops and bottoms with a great degree of accuracy. But consistently picking a market's top or bottom is very difficult, if not impossible. If picking tops and bottoms is so difficult even for an experienced trader, how much harder must it be for a blunt mechanical trading system to consistently pick a top or bottom?
If a trading system can't pick a market's top or bottom then it stands to reason that almost every trade taken mechanically will be 'in the red' at some point. If the market moves in favour of your trade the moment you enter it, and it continues to go your way so your trade is never under water at any point in it's life time then it is simply down to luck. The market moving against you (particularly at the beginning of a trade) therefore does not mean that your trade was a bad one; in fact it is to be expected. A trading system therefore needs to use a fairly large stop-loss. In fact, adding any stop-loss at all (other than a very, very large one) usually degrades the performance of any mechanical trading system to some degree. Using a small stop-loss with a mechanical trading system means you'll almost always lose; when the system takes a bad trade you'll lose and when the system takes a good trade you'll most likely be stopped out and lose then too.
I will add stop-losses to the list of characteristic the best Forex trading systems have; the best technical/mechanical Forex trading systems use large stop-losses.
Tailoring Systems For Specific Markets...
The 5th system I tested in the Technical Trading Systems Library examined whether it was best to enter as soon as a breakout was detected or whether it was best to wait for a price retracement. The result of this test was that it depended entirely on the market in question. For example, waiting for price retracements seems to work very well on stocks and stock indices but very badly on currencies, especially on the major currency pairs that trend well. The reasons why different markets behave differently are many and I don't want to write about these reasons in this article, for the purpose of this article I think it's enough to simply say that different markets do behave differently, sometimes very differently. One of the things I've often read regarding mechanical trading systems is that a trading system, if it's any good, should work well on all markets and not just a selected few. I believe this is completely false. A good technical trading system should work well on the market(s) it's designed to trade, and it is not a flaw if it fails to work on any of the markets it's not designed for.
This then is another characteristic that the best Forex trading systems have; the best technical/mechanical Forex trading systems are tailor made for specific types of market(s).
Avoiding Curve Fitting...
In mechanical trading, curve fitting is the process of over-optimising your trading system to make it 'fit' with past data. This results in a system with some unrealistic 'back-tested' results that will not be able to be duplicated in the future. When building a trading system, curve fitting is easily done, in fact, often system developers don't even realise that they're doing it. The best way to avoid curve fitting is to test the system's robustness by checking that it works well with other parameters (say 20% in either direction). For example, suppose a trend following system using a 100-day Donchian type breakout with a 50-day time-based exit produced some great results on the EUR/USD currency pair. It could simply be the case that through chance these parameters had worked great in the past. In order to check that the actual method it's self was sound and that we could expect similar results in the future we might test this system by changing the parameters 20% in each direction. How does it perform with an 80 or a 120-day breakout? What about with time-based exits of 40-days and 60-days? What about if we add or subtract 20% to or from the stop-loss? Does it still produce similar results then? It should, because if not the chances are that the system has simply been unintentionally curve fitted to data from the past and is not likely to produce such great results in the future.
In the example above I wrote about a system designed for the EUR/USD currency pair. In this example it would also be wise to ask how the system performed on other similar major currency pairs such as the GBP/USD or the USD/CHF. However, as I mentioned in the section on tailoring systems for specific markets, it wouldn't actually matter how the system performed on the FTSE 100 or the S&P 500, as those are completely separate markets that behave very differently to the major currency pairs. If a system performed well on the USD/CAD for example, then it should be expected to perform reasonably well on other commodity pairs like the AUD/USD. How the system performs on very similar markets is relevant, how the system performs on very different markets is not.
This then is another characteristic of the best Forex trading systems; the best technical/mechanical Forex trading systems work with similar parameters (say +/- 20%) as great care has been taken to avoid curve fitting.
Not Always 'In The Market'...
Tests have shown that systems that don't seek to trade all the time can out perform those that seek to always be 'in the market'. This is because there are always times in any market where the market conditions are not suitable for your trading system – when the odds aren't in your favour you should not seek to trade.
This is another characteristic of the best Forex trading systems the best technical/mechanical Forex trading systems do not seek to be 'always in the market'. The best Forex trading systems only seek to trade when market conditions are favourable.
The Best Forex Trading Systems: My Conclusion,
I will conclude this article by saying the following about the best Forex trading systems. The best Forex trading systems are long-term and trend following in nature, that is they seek to gain an edge from the effects of the longer term underlying market trend and give that edge time to work. They do not seek to pick tops and bottoms and therefore use fairly large stop-losses; their aim is to buy high and sell higher, or sell low and buy back even lower. They are tailored for certain types of markets, but not curve fitted. And they only seek to trade when market conditions are actually favourable and are therefore not always 'in the market'. These are my conclusions, hope they're of use to someone.
|
