19th Ave New York, NY 95822, USA

This is why forex trading is not easy (simple trading systems debunked)

LOREM IPSUM DOLOR SIT AMET

People often want to know whether forex trading is profitable. They see a strategy online promising 1000% returns and millions of dollars in profits and wonder whether they too can become successful forex traders.

On the flip side, though, they will also have heard the stories of those traders who have lost money and in the back of their minds they will wonder if forex trading is no different to gambling.

My opinion on this is pretty straightforward. Yes, it is possible to win in forex but it is not easy!

Why forex trading is not easy

The main reason why forex trading is not easy is because of the cost of trading and the efficiency of global forex markets.

Forex markets are the most liquid markets in the world with an average turnover of over $4 trillion. Because of this, the competition between forex traders is fierce. Big banks, hedge funds, organisations and speculators all battle it out to make money and as I will show, this competition means that simple trading strategies are not particularly successful for retail traders.

Furthermore, while the banks are able to trade directly on the interbank rate, the cost for retail traders is a little higher and this gradually erodes the potential for returns. This is especially true for day traders who ultimately battle against high performance HFT programs and bots.

Simple trading systems perform poorly in forex markets

In order to demonstrate how difficult forex trading can be I decided to test some simple but popular trading systems. You can see the video here:

I bought 20 years of accurate forex data from premiumdata.net and imported it into the back testing platform Amibroker. I then set about testing a number of strategies.

EMA Crossover system

The first test is with a simple EMA crossover. Whenever the 50 day EMA crosses over the 200 day EMA a buy order is entered and whenever the 50 day EMA crosses under the 200 EMA a sell order is entered. (EMA’s are calculated using the close price and orders are entered on the next open).

The system uses no leverage and takes positions in GBPUSD only. The test was run from the 1 January 1990 to the 1 January 2010 and commissions were set at just 0.01%.

As you can see, the results of this simple strategy were disappointing. The system achieved a CAR (compounded annual return) of just 0.36% with a maximum drawdown of -28%.

My conclusion is that the efficiency of the forex markets have made this trend system unprofitable.

Optimising EMA parameters

I decided to optimise the parameters of the EMA crossover from a 5 day EMA up to a 400 day EMA. By doing so I can test whether a moving average crossover is more profitable using different settings.

The optimisation found that the best parameters for the crossover were the 120 day EMA and the 261 day EMA.

But this still only produced a return of 2.36% with a drawdown of -13%.

why forex trading is not easy 2nd table

Of course, just change these settings slightly and the system can easily slip into negative territory. For example, change the parameters to a 110/171 crossover and returns are negative at -1.97% CAR.

forex trading is not easy negative returns

Breakout system

Next, I decided to run a breakout system. This system buys a forex pair when the currency closes at its highest for 120 days and goes short when the opposite occurs. This time I included several of the major forex pairs in the analysis: AUDUSD, AUDJPY, EURCHF, EURGBP, EURUSD, GBPUSD, GBPJPY, USDCAD, USDJPY.

Again, the results were less than stellar. CAR was 3.45% with a maximum drawdown of 23%.

Optimising breakout parameters

Finally, I optimised the breakout parameters to find the best settings for a breakout. This time I included all of the forex pairs at my disposal including exotic pairs such as TRYUSD.

The system performed best with a breakout setting of 165 days, resulting in a CAR of 5.19% and maximum drawdown of 17%.

why forex trading is not easy table

Conclusion

These tests were run on end-of-day data (supposedly the safer option for traders) yet they indicate that simple medium-term trading strategies do not perform very well on forex markets.

Even after optimisation, no test was able to deliver more than 6% in annual returns. (This is somewhat in contrast to strategies on stocks which appear to perform much better.)

This illustrates that forex trading is not easy and traders must be realistic about the trading returns they seek.

Hoping to double your money trading forex, using simple strategies like these, now seems highly unlikely and traders should back test their strategies first in order to understand what their profit potential is.

Furthermore, traders blindly following simple methods as illustrated are likely to be wasting their time.

What can we do to improve returns?

While these results are discouraging, there are many steps that traders can take to improve returns.

First, they can start to trade more than one forex pair at a time, in a portfolio. This might help to find profitable trades more often.

Second, they can add more rules and complexity to the system. Mean reversion techniques as well as trend following. Though they should refrain from adding too many rules or curve-fitting a system to the past data.

Third, they can incorporate other instruments such as equities, bonds and commodities in order to trade a more diversified portfolio. This should be one way to increase overall expectancy.

Fourth, they can experiment with different money management techniques, use stops and try out different timeframes.


Comments (6)

Over all good post with right information and analysis. The last part of the article “What can we do to improve returns?” is the most convincing portion to me. Loved it.

Glad you liked it.

Great article! Would you be willing to take a look at our software/strategies? We have been having some success in these channeling markets with our channeling strategies. I’d be happy to provide you with access to watch the strategies trading in real time and you can evaluate for yourself. Thanks!

Sure, best to email me directly at jbmarwood@gmail.com

Hi JB, thanks for your experiment on Forex.

Just my humble personal opinion here. Forex doesn’t follow a simple trend trading as you know we are dealing with pips here. A person a bet 100, 1000 or even 1000 on just a single pips move.

Typical Forex trader utilize a lot of position sizing, risk-reward ratio and risk management…. Most important is they way how they use leverage. Since you cant expect any long lasting big moves in currency (if that really happens that country is in deep trouble) forex typically takes in profits once a target zone is met. As i saw from your experiment, your this simple strategy will not work on Forex. Your code doesn’t have any consideration on the position size, risk taken for each pip and leverage effect etc.

I hope you will have time to investigate on the effects of money management and position size as i find these are the key to successful trading rather than emphasizing on entry and exit signals itself. Look forward to your analysis.

Thank you~! =)

Hi Kayson
Yes you are right of course. I wrote this article quite a while ago and I was not too familiar with how to set up forex simulations in my back-testing program (Amibroker). It requires a bit of extra leg work such as setting tick values and decimals. Yes I hope to do more work here, both before and after money management. Cheers.

Leave a comment