Figure 1. An example of a simple 3-bar price pattern - long entry
Each bar in the formation in Figure 1 has a high, a low and a close price. It may be seen that the close of the last bar, or today's close, is higher than the
high of bar 2 (the high of 2 days ago). This relationship can be expressed mathematically as follows:
Close of today > High of 2 days ago
By the same reasoning as above, it is clear from Figure 1 that:
Close of 2 days ago > Close of yesterday
The complete description of the price pattern shown in Figure 1 can be obtained by following the same reasoning:
High of today > Close of today AND
High of 2 days ago > High of yesterday AND
Close of today > High of 2 days ago AND
High of yesterday > Low of today AND
Low of today > Close of 2 days ago AND
Close of yesterday > Low of yesterday AND
Close of 2 days ago > Close of yesterday AND
Low of yesterday > Low of 2 days ago
The above 8 inequalities uniquely describe the price pattern formation shown in Figure 1. These inequalities (referred to as the "pattern logic") can be
combined with appropriate money management, trade entry point and trading time frame into a complete system. For instance, if the trade entry is on the
open of the day following the price pattern formation, the profit-target is T and the stop-loss is S, both expressed as a percentage of the entry price, a
trading system model for long positions can take the following form:
{Time frame: daily}
If {long pattern logic} then
Buy tomorrow on the open with
Profit target price at Entry Price x (1+T/100)
Stop-loss price at Entry Price x (1 – S/100)
By following a similar methodology, every simple price pattern formation can be incorporated into a complete trading system model and used in
back-testing or in generating trading signals. For example, Figure 2 shows a 3-bar pattern for short positions.
Figure 2. An example of a simple 3-bar price pattern - short entry
The complete description of the price pattern shown in Figure 2 can be obtained by following the same reasoning:
High of today > Close of yesterday AND
High of 2 days ago > High of yesterday AND
Close of today > Low of today AND
High of yesterday > close of two days ago AND
Low of two days ago > Close of today AND
Close of yesterday > Low of yesterday AND
Close of 2 days ago > high of today AND
Low of yesterday > Low of 2 days ago
The above 8 inequalities uniquely describe the price pattern formation shown in Figure 2. These inequalities (referred to as the "pattern logic") can be
combined with appropriate money management, trade entry point and trading time frame into a complete system. For instance, if the trade entry is on the
open of the day following the price pattern formation, the profit-target is T and the stop-loss is S, both expressed as a percentage of the entry price, a
trading system model for short positions can take the following form:
{Time frame: daily}
If {short pattern logic} then
Sell tomorrow on the open with
Profit target price at Entry Price x (1-T/100)
Stop-loss price at Entry Price x (1 + S/100)
It may be seen from the examples in Figures 1 and 2, as well as, from the pattern logic of the two patterns, that there are no parameters to adjust in order to
affect the timing of the entry signals generated by such trading system models. This is true unless additional logic is introduced, which will alter the basic
structure of the trading system models. What can be changed though is the timing of the exit signals, by changing the profit target and stop-loss levels, T
and S. Such change may eventually affect trading system performance by turning some entry signals into winners or losers depending on the exit price
levels and may result in some kind of optimization of the exit part. However, regardless of any optimization made to the exits, the entry part of such trading
system model cannot be affected. Thus, one can claim that while most indicator based systems are subject to full optimization (i.e. optimization of their
entry and exit part), price pattern systems can only be half-optimized (i.e. in the exit part). A stronger version of this statement is that price patterns that
exhibit a robust performance for a wide range of exit parameters do not belong to the class of optimized or fitted systems.
