I have the following problem: typical mean variance
minimize w_long(S)w_long.T + w_shortSw_short.T - lambda*mu
I am having issue formulating this constraint in cvxopt lets say we have 10 securities (n)
w_long>=0
w_short<=0
sum(w_long)=1
sum(w_short)<=0
A*(w_long) = b
A*(w_short)<= b*sum(w_short)
my A is a matrix of 3 x 10 where each security is put in one of 3 sectors. my b is an equally weighted matrix of 3x1. I am having trouble represented the A matrix S is your identity matrix
I know its easy to write this using cvxpy but I am interested in the A matrix for the input to cvxopt qp solver
From what I have read I need to introduce slack variables, if anyone can point to some source or give me guidance on how to set this simple example up that would be greatly appreciated
