Moment-matched Returns

Compute the closest - in the sense defined in the first reference - asset returns to original asset returns whose sample mean vector and (biased) sample covariance matrix exactly match a target mean vector and a target covariance matrix.

Due to the nature of the endpoint, subsequent calls with the same input data might result in different output data under specific numerical situations.

References

• Lin, J., Lermusiaux, P.F.J. Minimum-correction second-moment-matching: theory, algorithms and applications. Numer. Math. 147, 611–650 (2021)