Note: We have included this chapter in our manual in order to give our reader a glimpse into the making of a profitable trading system. Some of the technical language that is included may be new to some of our readers. However the understanding of this chapter is not necessary for the successful implementation of our strategy (which is fully explained in the latter chapters).
In our quest to find the most profitable and at the same time for a “small” trader feasible trading system we have tested and analysed many different trading strategies. The strategies that we have tested were ranging from simple combinations of TA indicators to more complex trading systems that were utilizing support/resistance levels, pivot points, chart patterns etc… However in order to reduce the number of systems that were later scrutinized more closely, we have developed our own system selection criteria. Basically the system that we were after had to have following properties: Simplicity, Efficiency and Consistency.
Simplicity
As we all know forex trading strategies are becoming more and more complex and sophisticated. What does it mean for our average independent trader? It means that our simplicity factor when developing a trading strategy gains in importance. What usage could an average person make from a strategy that requires or presumes a profound knowledge in mathematics at a PhD level and a computing power beyond that of the newest personal home computer? A type of highly complex strategies commonly used by 19
1investment companies are neural networks. A neural network is, in short, a model of interconnected neurons (also known as nodes) that is inspired by the logical neurons in human nerve system. Like the human brain a neural network can acquire, store and utilize experiential knowledge in order to improve its performance day by day. Regrettably, to consistently use a strategy based on neural networks one requires the complex knowledge of how to feed a neural network with history data as well as excessively high computing power not affordable to our average forex trader. Therefore we have set ourselves with a goal of finding a trading strategy that is comparable in it’s profit potential to the most complex professional trading system and at the same time is feasible and understandable to our average trader.
Efficiency
Efficiency of a trading strategy is basically a measure of profit that is realized using the strategy during specified period of time. When comparing different trading strategies, those strategies that show more profit during specified period of time are said to be more efficient.
Consistency
Once we have found a system that is efficient and simple to use our next most important selection criteria becomes consistency. What does it mean for a strategy to be consistent? It means that when the financial market behaviour changes slightly or even drastically, as often happens in times of political and financial crisis, the strategy is still able to make profit.
It means that a strategy with high efficiency and high consistency is a much better and safer strategy than a strategy with high efficiency but lower consistency. It is the consistency of a strategy that permits traders to plan for capital draw downs and potential profit build up.
A consistent strategy shows the following properties:
1 V.V.Kondratenko and Yu. A Kuperin, “Using Recurrent Neural Networks To Forecasting of Forex”, Condensed Matter, (2003)
20
• The strategy is profitable even in turbulent times, as for example directly after September 11.
• The strategy retains positive efficiency if the financial market behaviour changes slightly. This can be simulated changing slightly the parameters of the strategy. For example if a strategy has worked well in the past with a hard stop order of 50 pips it should also perform well if the stop order is changed to for example 55 pips, or 45 pips…
• The probability of losing all the trading capital during specified period of time needs to be extremely small, that is almost non existent. I don’t think that the importance of this property needs to be explained any further. ☺ We came to the result that our new developed strategy based on the ICWR phenomenon was the most powerful trading strategy, as it was the most efficient and most consistent strategy from all the tested strategies and at the same time feasible and understandable for our average trader. We had to put a lot of effort and time into forex market research to come to this conclusion. In the following we want to give you only a short look into our long way to our strategy.
First of all we will show you some of the milestones of the strategy development. That is how to define reliable and consistent market signals based on the ICWR phenomenon (see chapter 2.1.) and how to find the proper long-term filter for enhancing the performance of our strategy (see chapter 2.2.).
Finally we will give you some remarks regarding the high consistency (see chapter 2.3.) and high efficiency (see chapters 2.4. and 2.5.) of our strategy.
2.1. Market signals based on the ICWR phenomenon
Well, as told in the introduction (see chapter 1.3.), our strategy bases on the observation that when putting into relationship the height of a corrective wave to the height of the prior impulsive wave, the corrective wave tends to retrace the prior impulsive wave in Fibonacci ratios. Frequent relationships are 25%, 38%, 50%, 68% and 75%. 21
Our task was to define consistent and efficient rules for generating bearish and bullish signals based on the ICWR effect.
The open points were the following:
1. To find out the proper Fibonacci levels to be used. 2. To find out the proper triggers for identifying an impulsive or a corrective retracement. As you can imagine there exist really a lot of possibilities of defining rules for generating signals based on the ICWR phenomenon. The problem of making the rules too simple is that they don’t cover all of the possibilities that may arise when markets behave unusually.
Such an unusual behaviour is for example a candlestick being greater 3 times or more than its immediate neighbours. That means that there is a huge difference between the highest and the lowest value of that period. Such a candlestick represents a highly volatile time period. For example in the Figure 2.1. below, such a volatile time period at 08:00 on the 02/07/05 is recognized with the candlestick having a height around 60 pips, while around it, the other candlesticks have a height of 10-15 pips. Ok, suppose that the rule for recognizing a bullish signal in the case of an upward movement is the recognition of the price bouncing off any Fibonacci level (0.750, 0.618, 0.500, 0.318 or 0.250). After the upward movement starting at 06:10 and ending at 07:35 we could (following the former simple rule) make around 08:00 following market reading: around 08:00 we see the price clearly above the Fibonacci levels after having bounced off at the 0.382 Fibonacci level (around 07:50) and in consequence this would represent a bullish signal (see Figure 2.1).
But such a bullish signal makes no sense, as such an isolated and highly volatile candlestick has nothing to do with the impact of the ICWR phenomenon into the market. 22
Figure 2.1.
Another factor that we had to take into account is a sideways market, which can very easily generate false signals. For example in the Figure 2.2. below, after 17:00 the market doesn’t show any clear trend.
Ok, suppose the used rule for recognizing a bearish signal in the case of an downward movement is again the recognition of the price bouncing off any Fibonacci level (0.750, 0.618, 0.500, 0.318 or 0.250). After the downward movement starting at 11:00 and ending at 16:00 we could (following the supposed simple rule) make around 18:00 the following market reading: the price is below the 0.250 Fibonacci level after having bounced off it around 17:00 and in consequence this would represent a bearish signal. However it would make no sense, as such a sideways market has again nothing to do with the impact of the ICWR phenomenon into the market.. 23
Figure 2.2.
Because of that, the main effort had to be put into finding efficient and at the same time reliable rules (that is immune to the other effects of the market) for generating market signals based on the ICWR phenomenon. You will find these rules to be defined in detail in the chapter 3. “Intraday ICWR Trading Rules”.
For the sake of completeness let us just remark that in the shown examples (Figures 2.1 and 2.2) we get the right market signals when using our ICWR Trading Rules. 24
Figure 2.3.
In the example with the high volatile candlestick at 08:00 in Figure 2.3 no signal is generated when using our ICWR Trading Rules for two reasons. First, because the whole candlestick is not above the upper confirmation level (0.750). And second, no retracement channel is entered; even if the whole candlestick at 08:00 was above the 0.750 level no bullish signal would have been generated. Not only a false bullish signal is avoided, but also later the ICWR Trading Rules generate a correct bearish signal corresponding to the real market trend.
Figure 2.4.
25
In the example of a sideways market in Figure 2.4 no signal is generated, as no whole candlestick is below the lower confirmation level (0.250). 
2.2. The proper long-term filter
When developing a trading strategy, it makes sense to search for a proper long-term indicator in order to filter out the entry signals from the short-term scale. The reason is that such long-term filters make the strategy considerably more powerful (meaning more efficient and more consistent). In our case the short-term time period for intraday trading is five minutes candlestick and for long-term trading time period is four hours candlestick. Why do long-term filters make a strategy more efficient? The reason is quite simple. Suppose we are doing intraday trading. As we want to let our profits run, we are going to stay in the market typically for a couple of hours and sometimes even for a couple of days. Basically long-term filters are filtering out those entry signals that are not in the concordance with the long-term market behaviour.
2.2.1. Enhancing the Intraday Strategy
In order to enhance the intraday strategy based on the ICWR phenomenon the following TA indicators and rules were tested:
MA(x): The 20-period simple moving average from the x-period candlestick chart is used. The long-term signal is bullish if the actual value of the moving average is above the actual price. The long-term signal is bearish if the actual value of the moving average is below the actual price. X represents 30 minutes, 1 hour, 4 hours and 1 day. For example MA(1h) stays for the 20-period simple moving average from a one hour candlestick chart. RSI(x; 50/50): The x-period Relative Strength Index (RSI) is used (calculated using the last 14 values). The long-term signal is bullish if the actual value of the RSI is above 50. The long-term signal is bearish if the actual value of the RSI is below 50. X represents 30 26
minutes, 1 hour, 4 hours and 1 day. For example RSI(4h; 50/50) represents the 14-period Relative Strength Index from a four hours candlestick chart. RSI(x; 60/40): The x-period Relative Strength Index (RSI) is used (calculated using the last 14 values). The long-term signal is bullish if the actual value of the RSI is above 60. The long-term signal is bearish if the actual value of the RSI is below 40. X represents 30 minutes, 1 hour, 4 hours and 1 day.
CCI(x; 0/0): The x-period Commodity Channel Index (CCI) is used (calculated using the last 20 values). The long-term signal is bullish if the actual value of the CCI is above 0. The long-term signal is bearish if the actual value of the CCI is below 0. X represents 30 minutes, 1 hour, 4 hours and 1 day.
CCI(x; 50/-50): The x-period Commodity Channel Index (CCI) is used (calculated using the last 20 values). The long-term signal is bullish if the actual value of the CCI is above 50. The long-term signal is bearish if the actual value of the CCI is below -50. X represents 30 minutes, 1 hour, 4 hours and 1 day. 2In the Figure 2.5 shown below you can see the result of using the different long-term filters. The red thick line represents the result of the trading without a long-term filter (around 2700 pips of net profit).
2
The net result shown in the pictures is the average net profit, when trading five months in parallel the currencies EUR/USD, GBP/USD and CAD/USD using the ICWR Trading Rules. The average is calculated based on two years historical back-testing.
27
Analysis of different TA indicators for filtering 5 min ICWR signals400035003000]sip2500p[ t2000lus1500e
Rt1000eN5000)))))0)0)0))) m h h d0)0)0)0))))400)0/0/0/0)0) (1 (4 (1/5/5/5/54040400/0/0/; 0/-5/-50
(30AAA 6 h;
h;
d; 0//-50-5AMMM; 50; 50; 50; 50; 60/
6 6d; m(4; 5; 500/
5; 50M m h h dmh;h;0 (1
(4
(1 (30CI (1CI CI (1m hh;
(30SI (1SI (4SI (1SISISICICCC0 4 dSIRRRSI (3RRRC(3I (1I (I (1RRCI CCCCCCCUsed Long-Term FilterFigure 2.5.
As you can observe from the red line shown on the Figure 2.5 above our strategy is already highly profitable even without long-term filter, however when combined with some of the long term filters it becomes even more profitable. For example, when using the RSI(1 day; 50/50) long-term filter a profit of around 3900 pips is achieved which is 56% more profit than without this filter!
The proper long-term filter was found to be the RSI(1 d; 50/50) as it was the one with the highest efficiency.
2.2.2. Enhancing the Long-Term Strategy In order to enhance the long-term strategy based on ICWR phenomenon the long-term filters MA(1d), RSI(1d; 50/50), RSI(1d; 60/40), CCI(1d; 0/0) and CCI(1d; 50/-50) were tested.
In the Figure 2.6 shown below you can see the results of using the different long-term filters.
28
Analysis of different TA indicators for filtering 4h ICWR signals27002500]s2300pip[
tl2100useR
1900teN17001500MA (1 d)RSI (1 d; 50/50)RSI (1 d; 60/40)CCI (1 d; 0/0)CCI (1 d; 50/-50)Used Long-Term FilterFigure 2.6.
The red thick line represents the result of our strategy without a long-term filter (around 2450 pips of net profit). Again, as in the intraday trading the ICWR strategy is already highly profitable however when combined with long term filters it becomes even more profitable. For example when using the RSI(1 day; 50/50) long-term filter a profit of around 2625 pips is achieved (before only around 2450). For our trading rules the RSI(1 day; 50/50) was chosen as the long-term filter, as it was the one with the highest efficiency.
2.3. Consistency checks
Earlier in this chapter we have mentioned how important it is for a strategy to be consistent. In the long run it is the consistency of a strategy more than its efficiency that will make you successful in the trading business. 29
The strategy presented in this book is highly consistent. Showing you all the analysis done in order for us being able to make this statement would go clearly beyond the scope of this chapter, as we would be forced to bore you with pages and pages full of complicated statistical stuff. As this is not relevant for your trading we decided to show you only an extract from our analysis. That is the graphs showing that our strategy is immune to small changes in the given parameters – which simulates a slight change in the forex market behaviour.
The parameters of our strategy are:
• The minimum height of a movement to be considered as an active wave (used is 40 pips for intraday, 150 pips for long-term trading) • The distance between the entry price and the hard stop order (used is 40 pips for intraday, 100 pips for long-term trading)
• The minimum RSI value for the market considered bullish (used is 50 both for intraday and long-term trading)
• The maximum RSI value for the market considered bearish (used is 50 both for intraday and long-term trading)
These parameters are part of our trading rules, which are defined latter in detail in the chapter 3. “Intraday ICWR Trading Rules”.
As you can see from the Figures shown below (Figures 2.7 till 2.12), our strategy performs equally well when parameters are slightly changed. Both for the Intraday ICWR and the Long-Term ICWR Trading Rules.
30
Consistency Analysis of the Intraday ICWR Trading RulesMinimum Height of Active Wave450040003500]s3000ip[p
2500ltus2000e
R1500te1000N50003035404550Minimum Height of Active Wave [pips]Figure 2.7.: Net result [pips] in dependence of the minimum height of the active wave for the intraday ICWR trading rules
Consistency Analysis of the Long-Term ICWR Trading Rules / Minimum Height of Active Wave30002500]s2000ip
[ptl1500use
R1000teN5000100125150175200Minimum Height of Active Wave [pips]Figure 2.8.: Net result [pips] in dependence of the minimum height of the active wave for the long-term ICWR trading rules
31
Consistency Analysis of the Intraday ICWR Trading RulesDistance of Hard Stop Order from Entry Price450040003500]s3000ip[p
2500ltus2000e
R1500teN100050004045505560Distance of Hard Stop Order from Entry Price [pips]Figure 2.9.: Net result [pips] in dependence of the distance of the stop order from the entry price for the intraday ICWR trading rules
Consistency Analysis of the Long-Term ICWR Trading Rules / Distance of Hard Stop Order from Entry Price30002500]s2000ip
[ptl1500use
R1000teN50005075100125150Distance of Hard Stop Order from Entry Price [pips]Figure 2.10.: Net result [pips] in dependence of the distance of the stop order from the entry price for the long-term ICWR trading rules 32
Consistency Analysis of the Intraday ICWR Trading RulesMimimum RSI for Bullish / Maximum RSI for Bearish 450040003500]s3000ipp
[2500tlus2000e
Rt1500eN1000500060/4050/5040/60Mimimum RSI Value for Bullish Signal / Maximum RSI Value for Bearish Signal Figure 2.11.: Net result [pips] in dependence of the minimum/maximum RSI value for the market considered bullish/bearish for the intraday ICWR trading rules
Consistency Analysis of the Long-Term ICWR Trading Rules / Mimimum RSI for Bullish / Maximum RSI for Bearish
30002500]s2000ip
[ptl1500use
Rt1000eN500060/4050/5040/60Mimimum RSI Value for Bullish Signal / Maximum RSI Value for Bearish Signal Figure 2.12.: Net result [pips] in dependence of the minimum/maximum RSI value for the market considered bullish/bearish for the long-term ICWR trading rules 33
2.4. Why is our entry strategy so profitable? 
If you look at the entry signals that our strategy produces in Figure 2.13 you can see that our entry signals are able to predict the main direction that the market will take. This enables us to catch up a long-term intra-day wave after entering the market and therefore pick up a considerable number of pips.
Figure 2.13.
Instead, if you look at the entry signals that would for example have been generated by a commonly used entry strategy, that is the crossing of the 20-period moving average with the 5-period moving average, one recognizes that a lot of the entry signals generated by MA crossings are of really poor quality, as they are not able to predict what will be the main market trend (see Figure 2.14).
34
Figure 2.14.
2.5. Why is our exit strategy so profitable? 
Most traders make a mistake by thinking that entering the trade is more important than exiting the trade, and if they have found an entry strategy with positive expectations, “the job is done”. Nothing could be further away from the truth. Exit strategy is equally if not more important than entry. In majority of strategies that are used by average traders a trailing stop is used. A trailing stop is definitely better than a hard stop, however our strategy goes way beyond regular trailing stops when determining the place of exit. Before we go on, for the ones not knowing the meaning of a trailing stop, we will show you what a trailing stop is and how it works.
A trailing stop order with a moving rate of 30 pips works as follows: suppose you are entering long a position at the closing price of 1.2456 at 10:10 AM (see Table 2.1). Using the above defined exit strategy you will then put your stop order at 1.2426 (30 pips below 1.2456). If the next closing price is at least 30 pips greater than the last stop order, the new stop order will be the new closing price minus 30 pips. For example at 10:15 AM the closing price of 1.2490 is 64 pips greater than the last trailing stop. Because of that the new trailing stop is set to be 1.2460 = 1.2490 – 0.0030. Also at 10:25 AM the closing price of 1.2495 is 35 pips greater than the last trailing stop. The new trailing stop is set to 35
be 1.2465 = 1.2495 – 0.0030. In this example the position is exited at 10:30, because the low of that period of 1.259 is below the stop order of 1.2465. Time Low Close Old stop orderNew stop orderComment Entry position /
10:10:00 1.2446 1.2456 - 1.2426 Stop order at 1.2426Stop order changed
10:15:00 1.2464 1.2490 1.2426 1.2460 to 1.2460
10:20:00 1.2462 1.2483 1.2460 - Stop order changed
10:25:00 1.2478 1.2495 1.2460 1.2465 to 1.2465
Exit position /
10:30:00 1.2459 1.2479 1.2465 - Low below 1.2465
Table 2.1.: Trailing stop with a moving rate of 30 pips So what is the problem with the exit strategy shown above that uses a trailing stop? The problem with an exit strategy using a trailing stop is that it works against the basic fundamental trading rule “cut the losses short and let the profits run”. And if you use a strategy that doesn’t let your profits run you are in real trouble. And why does an exit strategy using a trailing stop work against this rule? Because very often such a strategy fails unnecessarily, it gets you out just at the moment when your trade needed just a little more space…Why? Every market trend, regardless of how strong it is, also shows movements against the long-term market trend. These deviations usually don’t last very long and after them the market moves again in the direction of the long-term market trend.
We will now give you an example that will show you why a trailing stop is not the best exit strategy. This example is based on the same EUR/USD long-term trade example from chapter 7. Let’s have a look at Figures 2.15 to 2.18. Suppose we entered the market long 36
at 16:00 on the 10/15/04 at the price of 1.2461 (see Figure 2.15) and also that we will be using a 150 pips trailing stop, which is an appropriate moving rate for long-term trading.
Figure 2.15.
At 04:00 on the 10/26/04 the closing price reached the price of 1.2802 (see Figure 2.16). That means according to the rules of a trailing stop, the new stop order is placed 150 pips below. That means at 1.2652 = 1.2802 - 0.0150.
Figure 2.16.
37
At 12:00 on the 10/28/04 the candlestick touches the stop order and the position is exited at the price of 1.2652 (see Figure 2.17).
Figure 2.17.
The total profit of the trade using the 150 pips trailing stop was 1.2461 - 1.2652 = 191 pips. Although we exited the position with profit (191 pips), we lost the chance of picking up the amount of pips that were possible in that trade. How much profit was actually possible in that trade? As we will show you explicitly in chapter 7 we made in this trade a profit of 723 pips, when using our exit trading rules. In Figure 2.18 you can see both the profit gained by the 150 pips trailing stop (191 pips) and by our exit trading rules (723 pips), which are almost four times higher. Using a simple trailing stop is not only a pity for the lost pips (532 pips), but it is making trading in the long run unprofitable: two losses of 150 pips followed by a win of 191 pips result in net loss of 109 pips. In contrast two losses of 150 pips followed by a win of 723 pips result in a net win of 423 pips! Do you get the point?
38
Figure 2.18.
The basic error of the strategy with the trailing stop was, that although the market was clearly going in the direction of the trade (on the 10/26/04 a profit spread of almost 400 pips was already reached) the trailing stop didn’t give the trade enough space to run. In particular if a profit spread of 400 pips was already reached it makes sense to risk, let’s say 200 pips to give the trade a bigger chance to double or even triple the profit (at the end 723 pips were reached!).
Basically the greater the profit spread is the more pips we can risk, in order to gain even more. Again, one should not forget, that not every position will make profit and in order not only to come even with the losses but also to make considerable profit, one needs to milk every possible cent out of every profitable trade. This is where our strategy comes into play.
39
We hope you could get some impression of the research done behind our trading strategy. We understand that this chapter was complicated and maybe sometimes a little bit boring; however we had made it as short and concise as possible. If we had included all of the tests and calculations that were needed to produce the ICWR strategy we would had needed at least several hundred pages… In the next chapter you will be shown every aspect of ICWR strategy that you need to know in order to be able to implement it successfully. Thank you for your patience. ☺
