DESCRIPTION OF THE ALGORITHM

  1. Select a set of asset among all the available assets that GRID Ideabook provides. The selection method is a clusterization algorithm.
    This Technique allows to select the assets that are not recently linearly correlated. The purpose is to diversify the portfolio as much as possible.

  2. Run on recent historical data of these selected assets a bunch of strategies involving financial indicators.

  3. Use a genetic approach to discard the worst models based on different metrics.

  4. Select the candidate assets for which there is a new signal.

STATISTICS ON THIS IDEA

Historical period analysis: 2.0 years
With the candidate model

  • Max draw-down: 5.9%

  • Total trades: 42

  • Total return: 62.0%

  • Average return / trade: 1.0%

  • Average time / trade: 16 days

  • Total success rate: 84%

The confidence on this idea is provided in attachment.