TECHNICAL ANALYSIS OF Stocks and Commodities Magazine
2016 Readers’ Choice Awards:
NeuroShell Tradernamed
#1 Artificial Intelligence Trading Software
14 years in a row
Thanks to all of you who voted for us
FIND THE BEST TIME SERIES FOR TRADING RULES
LET THE OPTIMIZER CHOOSE THE DATA STREAM
If you’ve ever built a Trading Strategy and let the optimizer choose the number of periods to compute a moving average or an RSI indicator, you know how easy it is to set a range of values and let the optimizer choose the most profitable setting for a particular instrument. But did you know that the optimizer can choose which data stream to use in your indicators or trading rules?
You can create a simple trading rule such as a crossover of a fast and slow moving average based on the close. However, with a simple adjustment you can let the optimizer select an entirely different data stream that results in more profitable trades.
In this example we changed the time series in the Moving Average Crossover Above rule from the traditional “close” to a choice between the RSI and the Volume Weighted Average of the RSI.
We suggest that you add any data series of interest to the chart before adding the Trading Strategy to make it easier to select multiple time series.
Note that the Time Series Search Space includes both the RSI and the RSI Volume Weighted version.
Next, insert the condition that will trigger a trade. For our Long Entry rule, we chose the Average Crossover Above indicator with a default value of close. Once the condition has been added to a rule tab, you can click on the “+” sign in front of the rule and display a list of price data and indicators already on the chart that may be included in the optimization search range. In addition to open, high, low, and close, you can choose RSI and the Volume Weighted version of RSI. You also have the option of choosing other instrument data.
NOTE:Our August 2015 newsletter included an article that used a similar technique to select the correct timeframe. You can view a copy of that newsletter in the newsletters section.
The technique created profitable models for Campbell Soup Company, Garman, and Intel.
The optimizer chose to use the RSI indicator in both the Long and Short entry rules.
While the rules shown above both use the RSI indicator, other models mixed a RSI indicator for the Long Entry rule and the Volume Weighted Average version of RSI in the Short Entry rule.
NEUROSHELL TRADER TIPS
Checking for a Sequence of Events
If you wish to determine when a condition, such as close < open, is true over multiple bars, you can use an indicator such as Minimum Value from the Basic category instead of having to combine “ands” and “lags”.
To see if the last 5 bars have had a closing price less then the open price, you could simply verify that the minimum value of Close < Open over the last 5 bars is True (i.e., 1). So for instance:
A=B( Min( A<B(Close,Open), 5 ), 1 )
The real power of this technique is that you can optimize the lookback of the Min indicator inside of a prediction or trading strategy, which allows the optimizer to determine whether 5 bars really is optimal or perhaps 3, 9, or some other number of bars is better.
If you are looking for a sequence of events that is not contained within a certain number of bars, another approach to this problem is to use the Followed By indicator from Advanced Indicator Set 3. You can even nest Followed By indicators if you use another copy of the Followed By indicator as the second condition in the indicator settings. This will enable you to search for a string of conditions.
