Statistical Learning for Individualized Asset Allocation

We establish a high-dimensional statistical learning framework for individualized asset allocation. Our proposed methodology addresses continuous-action decision-making with large number of characteristics. develop discretization approach to model the effect continuous actions and allow frequency be diverge observations. estimate value function using penalized regression our generalized penalties that are imposed on linear transformations coefficients. show Discretization Regression fOlded concaVe penalty Effect discontinuity (DROVE) enjoys desirable theoretical properties allows inference optimal associated decision-making. Empirically, is exercised Health Retirement Study data in finding The results strategy improves financial well-being population. Supplementary materials this article available online.