Ranking options

Author: Kornel Mazur

February 26th, 2019

1 Star2 Stars3 Stars4 Stars5 Stars (No Ratings Yet)

When the strategies are generated, every new strategy is backtested on a history data and the results of the backtest are then used to compute the Fitness (rank) of the strategy.
Fitness is number from 0 to 1 and it should reflect the “quality” of the strategy according to the given criteria.

In this screen you can configure:

  • how this Fitness value is computed (Strategy Selection criteria)
  • how many top strategies are stored in the databank
  • which strategies are stored to databank and which are thrown away (Custom conditions)

Databank Options

The best strategies found are continually stored into the Databank.
It is not possible to store every strategy (remember that StrategyQuant can create thousands of new strategies every hour) so we must specify how many strategies should be stored in the databank, how they should be sorted to find out the best ones and which strategies should be thrown away.

Maximum Strategies to Store in Databank
simply the maximum number of strategies remembered in the program

Stop generation when
allows you to define whether the generation process will be terminated when databanks if full or will continue until stopped manually.

Strategy Quality ranking (fitness)

Here you can choose from which data the strategy criteria will be computed and what criteria should be used to compute the total Fitness of the strategy.

Use (main data backtest / cross checks):
You can select if you want to compute fitness (strategy fitness score) from main data or from portfolio (if you use also additional data).
It is generally recommended to use Portfolio, then the strategy fitness will be computed from the sum of all trades across all the symbols in your portfolio.
If you use this setting and you don’t have any additional data, fitness will be automatically computed from main data.

Determine best strategies by
You can choose from the predefined most used criteria, or you can build your own complex fitness function based on multiple criteria, each with a different weight.
Note that if you choose to combine too many criteria, they might “fight” against each other without achieving what you expected. Profit Factor
criterion to maximize Profit Factor of a strategy.

You can find explanation for every ranking criteria here.

Strategy filtering conditions

you can define rules to throw away strategies with bad properties. For example it usually makes sense to throw away strategies that result in loss or these that have too little number of trades.

Automatic filters

automatic filters should be used to dismiss strategies that are obviously wrong such as 0 trades strategies, strategies having no profit or strategies which have no filled orders. All the available options can be seen below.

Custom filters

allow you to specify your own custom rules that will be evaluated for every strategy.
If the strategy matches any of the rules, it will be dismissed.

This is useful to quickly dismissing strategies that result in loss or with too little number of trades.

There are separate values for In Sample and Out of Sample performance, and for Robustness Tests (if used) or Portfolio (if you used additional data).

The abbreviations used in custom conditions are:

  • IS (In Sample) – result of In Sample part of data
  • OOS (Out of Sample) – result of Out of Sample part of data
  • RT (Robustness Tests) – result of robustness tests
  • P (Portfolio) – results for portfolio (if exists)

Results can be evaluated using Money, Percent or Pips values for either long, short or both directions.

Leave a Reply

Please Login to comment