Solanum trading system parameter optimizations for the period 2006/01/01 to 2014/08/31.
Optimization is the process whereby back tests are run for a large number of different parameter values and the resulting metrics compared to determine which combination of all possible parameters are best matched with your trading goals.
Solanum was designed to use the standard MACD parameters of 9, 12 and 26 MACD(12,26,9), but what about the more exotic MACD(21,81,2) or MACD(5,10,4). Do those parameters perform better or worse?
Optimization allows us to test a range of possible values and Amibroker will let you test up to 10 parameters during an optimization run. The size of the value ranges are only limited by your available processing power and time.
For the Solanum optimizations, I used the following parameter ranges:
The actual Amibroker optimization script snippet:
pMACDFast = Optimize( "pMACDFast", 9, 4, 16, 1 );
pMACDMid = Optimize( "pMACDMid", 12, 6, 20, 1 );
pMACDSlow = Optimize( "pMACDSlow", 26, 16, 40, 1 );
pStoL = Optimize( "pStoL", 80, 10, 90, 10 );
pStoH = Optimize( "pStoH", 40, 10, 90, 10 );
This results in 394,875 possible scenarios (unique parameter sets) or individual back tests. The optimization was run for the 15 minute, 30 minute and 1 hour periodicities. On the computer that ran the optimization, an Intel Core 2 Duo E4600, it took 55 hours.
The date range was updated to 2006/01/01 to 2014/08/31, the initial equity set as R 50,000 and the margin as R 30,000. Commissions and slip are not accounted for.
By default the spreadsheet results are sorted for best Net Profit, but you should sort on other columns to check your preferred metrics, net profit is not the best indicator of a robust trading system.
The first 5 columns show the parameter values used to create the result row. Since the results spreadsheets are very large I have removed parameter rows with no trades.
Only trade an optimized or the original version of Synergy Solanum after doing your own research and make sure you are aware of the negative aspects of over optimization, such as curve-fitting. The original version of Solanum has proven itself to be fairly robust in live forward testing and the parameters should only be altered with caution.
Information, opinion, results and statistical analysis published on this website must not be construed as trading advice.
All your trading and investment decisions must be based on your personal interpretation of the information and your financial situation. Derivatives trading may result in large financial losses. You are solely responsible for your trading decisions.
Solanum Swordfish Script and Results
