I dont think there is something odd about the article. The author just makes his point that the correlation of two variables i.e. instruments that compose his portfolio change over time which in turn means that the portfolio manager is responsible for adjusting the weights of variables A and B...