Additional portfolio selection weights method: LASSO, qp

Source: R/additional_weights.R

Default using lasso after quadratic programming. As a approximation to constrained risk minimization

weights_qp(Rt, constr = 1)

weights_lasso(Rt, N = NCOL(Rt[[1]]), qp_lasso = TRUE,
  qp_weights = NULL)

Arguments

Rt

Lists of return rt
constr

If numeric, take as the imposed constraint. If NULL, no constraint. Otherwise compute proper constraint using 10-fold cross-validation
qp_lasso

if TRUE, taking no-short-sale portfolio from quadratic programming as y in lasso
qp_weights

can supply weights of o-short-sale portfolio from quadratic programming

Value

List of weights for each period

Functions

  • weights_qp: portfolio selection quadratic programming for multiperiod data set

See also

weights_lm