Dynamic strategy

Portfolio strategy

In estimation settings you can include one or a handful of investment strategies. For the calculation function of the optimisations for all strategies, refer back to the frontiers module. Predefined strategies include:

  • Equally-weighted: by selecting this strategy, it creates an equal-weighted portfolio for the full sample (not limited by in-sample / out-of-sample setting)

  • Fix-weighted: same as the "equally weighted" strategy, but this can add fixed weights for individual assets.

  • Minimum variance

  • Target mean (variance)

  • Equal risk contribution (variance)

  • Minimum expected shortfall

  • Target mean (ES)

  • Equal risk contribution (ES)

In-sample / Out-of-sample configuration

The "In-sample date range" sets the date range to calculate the portfolio optimisation using only the in-sample period data. The sampling method can choose :

  • The fix method only calculates the in-sample portfolio optimisation once and applies the optimal weight to the entire out-of-sample period.

  • The rolling (window) method calculates the in-sample optimisation first and applies the optimal weight to the next period only. It then moves the window forward one frequency and recalculates the optimisation process.

  • Moments: please refer to the moment module for the same functions and settings.

  • Conditioning information: please refer to the frontiers module for the same functions and settings.

  • Portfolio weights: plots the time-variation of portfolio weights through the entire out-of-sample period. By clicking each date or selecting from the date field, the user can check the asset allocation at each time point.

Last updated