Generating Input Distributions for Explaining Portfolio Optimization Pipelines
For financial analysts and portfolio managers, this provides a new way to explain complex predict-then-optimize pipelines, though the approach is demonstrated only on illustrative examples without quantitative performance metrics.
The paper introduces a predict-optimize-explain framework that uses gradient-based sample generation to interpret portfolio models by identifying macroeconomic conditions leading to specific portfolio outcomes, demonstrating its utility on four what-if questions to reveal behavioral differences between decision pipelines.
We propose a predict-optimize-explain framework that uses gradient-based sample generation to interpret various portfolio models by identifying macroeconomic conditions that induce specified portfolio outcomes. Unlike traditional feature-importance methods, this approach directly probes decision pipelines (predictive models coupled with portfolio optimization) by constructing economically meaningful what-if questions. We focus on four such questions: under what macroeconomic conditions a predict-then-optimize pipeline closes or reverses its return gap with a predict-and-optimize pipeline; what conditions lead a pipeline to diversify rather than concentrate its allocation; when a pipeline trained on calm markets overtakes one trained through crises; and what conditions would let a pipeline match a benchmark return. These examples illustrate how our framework uncovers key behavioral differences between various decision pipelines. Beyond these cases, the proposed framework is flexible and can support a wide range of probing questions tailored to specific portfolio objectives. Our findings highlight the value of integrating prediction, optimization, and explanation to produce more robust and transparent portfolio strategies.