From the April 01, 2009 issue of Futures Magazine • Subscribe!

Trading forex extremes

Not only are the forex majors among today’s most liquid and widely traded markets, but historically they also have been excellent trending markets. Frequently, most trading profits booked by many of today’s top trend-following firms are generated trading the forex markets.

While trend is always a function of time, the best trending markets exhibit strong directional movement across a multitude of time frames. The six major forex currency pairs include the dollar in relation to the pound sterling (USD/GBP), the euro (USD/EUR), the Japanese yen (USD/JPY), the Canadian dollar (USD/CAD), the Australian dollar (USD/AUD) and the Swiss franc (USD/CHF).

Here, we’ll outline a simple short-term forex strategy that combines a major Japanese candlestick reversal pattern with the relative strength index (RSI) for the entry signal. This entry criterion is complemented with a basic trailing stop exit mechanism. When properly synchronized into a trading system, these components produce the basis for high probability entry signals coupled with an aggressive low-risk exit strategy.


The RSI was developed by J. Welles Wilder Jr. in his book “New Concepts in Technical Trading Systems,” published in 1978. While RSI has many applications in both charting analysis and system development, the focus for this strategy is merely on the oscillator’s 70/30 interpretation. Traditionally, when RSI is above 70, a market is becoming overbought, and when RSI is below 30, it is becoming oversold.

For this strategy, however, the default parameters of 70/30 were found to be suboptimal. In essence, the 70/30 thresholds generated too few trades to be effective over the time horizon we’re studying. A better pair for this horizon, used in conjunction with this particular Japanese candle formation, is 60/40. In other words, a short entry is signaled when the RSI value is 60 or higher, and a long entry is signaled when the RSI value is 40 or lower.

However, for comparative purposes, system results based on both RSI input pairs will be shown.


Candlestick charting provides a wealth of insight into deciphering the bullish and bearish sentiment within a market. If you know how to interpret them, candlestick price formations can illuminate effective continuation and reversal price signals.

As evidenced by Thomas Bulkowski’s recent book, “Encyclopedia of Candlestick Charts,” candle reversal patterns are not all created equal. Certain patterns are more effective than others at identifying major price reversals. In testing this system, a Japanese candle reversal pattern that statistically yields higher predictive short-term reversals was focused on the morning and evening doji star price pattern.

This distinctive doji star candle pattern is composed of three price bars. In the case of the evening doji star, the first bar begins as a long white bullish candle. The second price bar gaps higher on the open, trades in a small range, and closes roughly at the bar’s opening value. This price action forms the doji star. The third price bar gaps down on the open and ultimately develops into a long black candlestick that closes below the midpoint of the real body of the first bar, completing the bearish reversal. (See the first chart in “Doji patterns”)

In the example of the morning doji star, the same price dynamics are simply reversed. The first bar is a long black candlestick that extends the existing downtrend. The second price bar gaps lower and forms the doji star. The third and final completion bar is the long white bullish candle closing above the midpoint of the real body of the first black candlestick (See the second chart in “Doji patterns,” above).

In his research to unearth the name origin of various candle patterns, candlestick expert Steve Nison discovered that the morning and evening doji star patterns were originally called the three river morning star and three river evening star. According to legend, this term originated from a 16th century military campaign led by feudal leader Nobunaga Oda, whose forces forded three consecutive rivers in a fertile rice valley and thereby secured a decisive victory later resulting in the unification of modern Japan. Given the extreme difficulty of this military conquest, the analogy speaks volumes as to the formidable support and resistance wall that forms at the mid-point of this unique candlestick price pattern.


The last component of this system involves managing the risk during the trade with an initial stop loss that converts into a trailing stop and moves steadily in the direction of the trade.

Following a buy signal, the initial stop is placed three pips below the low of the completion bar. As the price moves higher, defined as successively higher highs and higher lows, the trailing stop is moved three pips below the low of each successive price bar one bar prior to the current bar. By definition, if during a trade an inside bar forms (meaning the entire range of this price bar is contained within the prior price bar’s range), until this condition is resolved, the trailing stop remains three pips below the earlier low.

By way of review, the RSI provides a technical filter that calibrates the strength or weakness of the trend and identifies short-term price extremes. In the evening and morning doji star price formations, the doji star becomes the signaling bar that must fall above or below the required RSI thresholds. When this candle pattern occurs at the RSI extremes, this candle reversal pattern provides a springboard for an optimal price set-up. A simple stop exit strategy both limits risk and moves the stop loss in the desired direction of the trade immediately.


The trades generated from the 30-minute entry signals were far and away the most profitable time frame across the forex portfolio, with five of six (60/40, or “RSI Star”) and four of six (70/30, or “123 Star”) pairs generating a profit (see “Paired profits”).

In the same 30-minute category, across both strategies, the profit factor ranged from a high of 2.07 (for all trades long and short) to a low of 0.22, an outlier statistic for the Swiss franc. However, net profit was significantly lower and, in fact, slightly negative across the portfolio for the more stringent 123 Star strategy.

Not unexpectedly, the total number of trade signals generated from this 70/30 strategy was far fewer as well. While the total number of trades was noticeably higher on the five and 15-minute time frames, the total trades generated were still low compared to the 60/40 RSI Star strategy. This was due not only to the wider RSI thresholds, but also due to the relative rarity of the evening and morning doji star price pattern within the three intra-day time frames.

It was striking to see that across all three time frames in the RSI Star strategy, the percentage of profitable trades only averaged 32.4%. In other words, just over two thirds of all trades were losing trades. Yet despite this fact, the average winning trade was 2.36 to 1.00, or nearly 2.5 times larger than the average losing trade. This positively skewed average dollar win-loss ratio seems to demonstrate that the risk management criteria largely succeeded at cutting losses short and letting profits run.

In terms of the top performing time frame and pair, the five-minute RSI Star USD/GBP pair realized the greatest net profit. The USD/EUR pair was the best overall performer, showing a profit in four of six time frames, respectively. Next was the yen, Australian dollar, Canadian dollar, pound sterling and Swiss franc. The franc was the only market that posted a net loss.

Finally, across both strategies, the importance of taking every long and short signal can not be understated. While more profits were generated on short trades relative to long trades, this reversal strategy was only in the market about 5% of the time. Thus, shorter-term traders with a psychological need for market stimulation and trading activity would likely have a difficult time applying this type of system, due to the combination of lower trade frequency and a higher percentage of net losing trades relative to winning trades.


While the primary trend of the dollar moved through three distinct phases from down to sideways to up over the one-year test period, a longer-term track record would need to be produced to certify the validity of these findings.

On the entry side of the equation, dynamic application of support and resistance zones would help isolate more significant and lasting reversals. Thus, when the extreme point in the pattern coincides with an important support and resistance area, the quality of the signal may be rated higher, resulting in a higher percentage of profitable trades. Also, the inclusion of a volume and price measurement filter within the three-bar reversal pattern might further delineate between strong and weak signals.

On the exit strategy, the first order modification to test would be lowering the initial stop (on long trades) from just below the low of the completion bar to just slightly below the low of the doji star’s extreme low price point. (The opposite would be done to adjust stops on short trades.) While this increases the initial risk in every new trade, the larger question becomes whether this additional risk might be more than fully offset by a larger number of profitable trades as well as higher net profit.

Though vast numbers of newfangled trading algorithms are being thrust upon today’s forex markets, you don’t have to be a rocket scientist to trade them effectively. At their core, all trading systems merely consist of three things: an entry criteria, an exiting strategy to define and manage the risk, and an underlying investment process. Further, whether fully administered by computers or with human intervention, the trading process is still guided by human decision making.

When developing any new system, the human element is not only the most important, but it is also the most fragile. Technically speaking, it is always better to trade an average system with superior risk management, than it is to trade a superior system with average risk management.

System results are available for review at

David Wilder is the chief market analyst for the Delta Society International and publishes advisory reports covering more than 30 commodity markets. Contact him at

comments powered by Disqus