BlackRock AI Labs

We address portfolio allocation with illiquid assets using a convex optimization-based model predictive control policy, achieving performance close to fully liquid asset scenarios despite delays and uncertainties.

We describe a light-weight yet performant system for hyper-parameter optimization that approximately minimizes an overall scalar cost function that is obtained by combining multiple performance objectives using a target-priority-limit scalarizer.

The continuous nature of today's stock markets places a premium on speed, which has in turn given rise to a latency arms race. In this talk, we discuss how financial market design can potentially mitigate some of these issues.

Using a lifecycle framework with Epstein-Zin utility and a mixed-integer optimization approach, we compute the optimal age to claim Social Security benefits. We summarize the decision criteria via the ratio of wealth to the primary insurance amount.

Mean–variance portfolio optimization problems often involve separable nonconvex terms, including penalties on capital gains, integer share constraints, and minimum position and trade sizes. We propose a fast heuristic algorithm for this problem.

We formulate a tax-aware portfolio construction problem that explicitly accounts for tax liabilities from long- and short-term capital gains, while maintaining the standard objectives of return, risk, and transaction costs.

We consider an investment process that includes a number of features, each of which can be active or inactive. We argue for attribution of performance to features based on Shapley value and propose efficient computation methods.

We revisit Merton’s seminal 1970 formulation (and solution) of the consumption and investment decisions of an individual investor. We present a formulation of Merton’s problem as a deterministic convex optimal control problem.