Tracking Error under Exclusion vs. Optimisation

We compared these two approaches in terms of carbon intensity and tracking error, using the MSCI© World Index as the benchmark. All simulations are as of the end of February 2019. In the first approach we excluded the three most carbon-intensive sectors from the universe and reallocated their weights proportionally to the remainder of the portfolio. In the second approach we used a portfolio Optimiser and risk model to minimise the carbon footprint, and constrained the sector weights to be in line with those of the MSCI© benchmark. 

The result of the sector Exclusion simulation is shown by the green square in Chart 11. In the simulation, the Exclusion method achieved a 75% reduction in carbon intensity compared to the benchmark represented by an orange triangle, but with a relatively high tracking error of 1.12%.

To assess the carbon Optimisation approach, we constructed a series of portfolio simulations with varying combinations of carbon intensity and tracking errors, shown as blue X's. The Optimisation portfolio within this series for which the carbon intensity is similar to that of the Exclusion approach is marked by the blue square: a 75% reduction in carbon intensity but with a significantly lower tracking error of 0.27%.

The Optimised approach reduces risks arising from sector biases. Although the portfolio remains invested in high-carbon sectors, it does so by focusing on the most efficient, or lowest-carbon-emitting, companies within each sector. This contributes to the market signals sent by investors and supports investor campaigns to promote carbon responsibility within high-carbon sectors.

Donwload the Research Paper 

This Research Paper is dedicated for professional investors only