Some of the most popular trading strategies in forex markets involve the use of Japanese Candlestick charts. Given a specific pattern in candlestick formations, traders look to buy and sell currencies in anticipation of reversal or continuations in price. Yet testing the profitability of such concepts is easier said than done. Given that many of these formations are inherently qualitative in nature, it is difficult to develop a reliable quantitative approach with which to test the viability of such strategies. That being said, we will attempt to quantitatively identify specific candlestick patterns and backtest the profitability of trading on such candlestick signals. The first ones that we will analyze are morning star and evening star formations.
[B]Candlestick Formations: Which do we choose?[/B]
Given a great number of different candlestick formations, it would be nearly impossible to gauge the overall profitability of all candlestick strategies in a single study. Instead, we will focus on specific reversal signals that we believe have intuitive value as it relates to market sentiment. In this particular case we will look at both the Morning Star and Evening Star formations as buy-and-sell signals.
A Morning Star formation is a bullish reversal signal for an overall downtrend. Given fairly consistent losses, we see a strongly negative full-bodied first candle. The second candle opens at or below the previous close, trading within a relatively narrow range with the high staying below the midpoint of the first candle. The third candle is strongly positive and closes above the midpoint of the first candle. This tells us that bearish sentiment is unable to push price below previous lows, and risks remain for a reversal in price trends.
The Evening Star is effectively the opposite of a Morning Star, as price starts in an uptrend and the first candle is strongly positive. The second candle opens above or at the previous close, trading within a fairly narrow range and with its low above the previous bar’s midpoint. The third candle is strongly negative and closes below the first candle’s midpoint. This gives warning that bulls are unable to push price to new heights, and a strongly bearish candle hints at further downside potential through subsequent trade.
[B]Backtesting our Candlestick Formations[/B]
Using TradeStation’s EasyLanguage, we translate our Candlestick rules into quantitative code in order to test their performance on a historical basis. (For those interested in seeing the actual Candlestick strategy code in TradeStation’s ELD format, click here. To see the same code in more widely compatible .txt format, click here.) In doing so, we can easily test our concepts across the spectrum of currencies and time frames. We take special care in not adding arbitrary elements to our code that may lead us to over-optimize our results, as our aim is to judge the raw profitability of the simple candlestick formation.
In terms of actual trading rules, we tell our strategy to buy a standard lot of the given currency pair when a Morning Star formation materializes and short sell the currency on an Evening Star. Our initial results do not include stop-loss orders or profit limits. Instead, gains and losses are realized when positioning flips on the subsequent Morning or Evening Star formations. Trading begins on the inception of the euro (12/31/99) and ends at publication date.
[B]Initial Backtesting Results: Could we Improve Upon our Raw Candlestick Strategy?[/B]
Our initial backtest shows that our simple strategy remains modestly profitable across major currency pairs. Yet with sizeable Max Drawdowns, it remains clear that our strategy may do well with slightly improved risk management. Indeed, it may be important to take a closer look at individual exit and entry rules and their efficiency in capturing strong risk-adjusted returns.
Using TradeStation, we are able to quickly assess maximum profit and loss potential across trades with several functions. Of particular interest in this case is to measure each individual trade’s maximum downward move before turning profitable. For simplicity’s sake, we will look at EURUSD trades. The chart below gives us an accurate depiction of how risk/reward on overall trades.
The chart above shows us an interesting pattern in our candlestick-based trades. Namely, our most profitable trades tend to have very little negative incursion. In plainer terms, our best trades are those that are strongly positive from the onset and show very little downward extension. At the same time, some of our most negative trades show very little profit potential from a buy-and-hold perspective. What this tells us is that our strategy would likely benefit from a simple stop-loss measure that protects us from large losses yet does not keep us from making a profit on our more successful trades.
In this particular case we see that six out of ten trades that draw down beyond $2,000 turn out to be losers, while only one of nine that drawdown less than this amount end as net losses. This suggests that adding a stop loss to our strategy can dramatically improve our overall profitability.
The table above is clear evidence of this point, as our overall rate of return improves substantially if we limit or maximum loss to a total of $1,000 per given trade. Looking at our results, we see that our percentage of profitable trades drops dramatically when we set a stiff profit target on our trades. In fact, of 37 total trades, 24 produce a loss when we set our stop to $1,000. Yet the amount in profits we collect on each winner far surpasses the relatively modest loss on each loser. It seems as though our Morning Star – Evening Star candlestick strategy works best with tight stop losses and looser profit targets. The next logical question becomes whether we can improve our strategy with a fixed profit target.
The chart above shows us the exact opposite of our Maximum Adverse Excursion chart, whereby we can judge the efficiency of our profit-taking techniques on our trading strategy. The solid black line represents a 1:1 Maximum Favorable Excursion to trading profit line. In other words, trades closed at the line represent instances in which we have captured the maximum possible profit in a particular trading opportunity. Of course, we all know that this is a near-impossibility to achieve on a regular basis. Instead we look to see if there are any particularly extreme cases in which a set take-profit would have dramatically improved our profitability.
Again we see interesting results when we modify money management rules in our trading strategy. Our profitability improves and our overall rate of return rises when we set a fixed profit target of $12,000 per trade. Though the percentage of trades that are profitable drops to a mere 36%, it is clear that a reward to risk ratio of $12,000 to $1,000 offers very attractive risk-adjusted returns. Below we see the actual equity curve of our trading strategy and final conclusions on the profitability of our Morning Star – Evening Star candlestick strategy.
[B]Final Conclusion and Caveats[/B]
[B]
[/B]
Our attractive equity curve suggests that our candlestick strategy has been profitable in the EURUSD from inception to present. The raw results—with no set money management rules—were slightly less profitable, and tell us several interesting things about the nature of our candlestick strategy. We saw that profitability improved dramatically when we cut losses short and let our winners run significantly longer. This tells us that our candlestick signals are either “wrong” in their reversal signals and not worth holding, or they are “correct” and give us ample opportunity to take profits on a turn in trend. It is undeniable that this strategy does indeed provide many false trading signals. Yet we see that proper management rules allow us to capture solid profits and limit our downside on these declines. Combined with other trading techniques and more discretionary strategies, we would argue that these candlestick formations increase our overall trading profitability.