Variance Moyenne Estimation (MVE) is a fundamental concept in finance and statistics used to assess the risk and return profile of investment portfolios. Developed by Harry Markowitz in the 1950s, MVE provides a structured way to select assets that maximize expected returns for a given level of risk.
L'idée centrale derrière la MVE est d'analyser la mean (average rendement attendu) et variance (measure of risk) of different assets. By combining these two metrics, investors can create a diversified portfolio that balances potential gains with acceptable levels of risk.
In practical terms, MVE involves calculating the expected return of individual assets and their correlations with one another. This allows investors to understand how assets might perform together, as some assets may offset risks associated with others. The goal is to construct an ‘efficient frontier’ of optimal portfolios that provide the highest expected return for a given level of risk.
Mathematically, the expected return of a portfolio is calculated as the weighted sum of the expected returns of its individual assets, while the variance is derived from the weights of the assets, their individual variances, and their covariances. This complex interplay forms the basis of Modern Portfolio Theory (MPT).
Bien que la MVE soit largement utilisée, elle présente des limites, notamment dans sa dépendance aux données historiques pour prédire les rendements et risques futurs, ce qui peut ne pas toujours être fiable. De plus, elle suppose que les investisseurs sont rationnels et que les marchés sont efficients, ce qui peut entraîner des écarts dans les applications réelles.