This module calculates linear replication following this paper's method. In short, this method applies the regression model without the interception term and uses the estimated coefficient as the weights of replicated assets. There are a few options for users to adjust:

  • Weights sum to 1: generally, the sum of coefficients (weights) should equal to 1 in the replication regression. If not, then the remaining weight is allocated to the risk-free asset.

  • Short-selling constraints: usually, the replicated assets cannot take short positions.

  • Volatility adjustment: please refer to the paper linked above for more details on this option.

Last updated