The portfolio manager should use incremental VaR to compute the change in VaR in this scenario because VaR changes in a non-linear fashion. Incremental VaR is the change in VaR from the addition of a new position in a portfolio. Since it applies to an entire position, it is generally larger than marginal VaR and may include nonlinear relationships. For this reason, incremental VaR can be viewed as the more accurate VaR measure especially when the anticipated positions are large in magnitude and size. In the question, the manager intends to change the composition of his portfolio. That implicitly means the changes would be significant, and the change in VaR would be best measured using the incremental VaR.

Choice A is incorrect. The portfolio manager's assertion that the change in VaR due to the addition of a new asset can be measured equally well by either marginal or incremental VaR is not accurate. Marginal and incremental VaR measure different aspects of risk and are used in different contexts.

Choice B is incorrect. The statement suggests that marginal VaR should be used because it assumes changes in VaR are linear, which is not always true. Marginal VaR measures the effect on overall portfolio risk when a small amount of an asset is added to the portfolio, assuming all other factors remain constant. However, this does not necessarily mean that changes in VaR will always follow a linear pattern.

Choice D is incorrect. This choice incorrectly suggests that marginal VaR should be used when changes in Value at Risk (VaR) occur non-linearly. In reality, incremental Value at Risk (IVaR), rather than marginal Value at Risk (MVaR), would be more appropriate for measuring non-linear changes as it considers the change in total portfolio risk when an entire position or asset class is added or removed from the portfolio.