This website presents academic research and analytical tools for informational purposes. It does not constitute financial advice. Results may vary. This website presents academic research and analytical tools for informational purposes. It does not constitute financial advice. Results may vary.

Ask yourself

What if your model is wrong about tomorrow?

Most frameworks assume you know the distribution of returns. Our team does not. We build multiperiod ambiguity aversion tools for investors who respect uncertainty rather than ignore it.
See our approach

2500+ Research citations

Our research program sits at the intersection of robust statistics, dynamic programming, and the psychology of uncertainty. We publish on recursive multiple-prior utility, robust Bayesian updating, and time-consistent ambiguity averse preferences across multiperiod investment horizons. The goal is not to predict markets but to build decision frameworks that survive when assumptions break.

Dynamic ambiguity in returns

We model asset return ambiguity as a dynamic process that evolves with information arrival. Unlike static robust approaches, our frameworks allow the ambiguity set to shrink or expand based on empirical evidence, preserving appropriate caution without unnecessary conservatism.

Learning under ambiguity

How should investors update beliefs when they distrust their models? Our work on robust Bayesian learning shows how to filter signals without overreacting to noise, maintaining calibrated uncertainty across multiple investment horizons and regime shifts.

Robust multiperiod rules

Single-period robust optimization often fails when extended dynamically. We develop time-consistent multiperiod rules using recursive utility formulations, ensuring that today's planned behavior remains optimal when tomorrow arrives and new information emerges.

Empirical applications

Theory without evidence is speculation. We test our frameworks against historical regimes, simulated stress scenarios, and adversarial model perturbations, documenting where ambiguity aversion adds value and where simpler approaches suffice.

Committee communication

A robust portfolio that cannot be explained is unused. We study how to communicate ambiguity bounds, worst-case outcomes, and model uncertainty to investment committees who demand clarity without false precision.

Research milestones

7 milestones — Sort by year

Year Event Area Details
Recursive ambiguity foundations Theory Initial framework combining recursive multiple-prior utility with dynamic portfolio choice, establishing axiomatic foundations for multiperiod ambiguity aversion in asset allocation.
Robust Bayesian filters Methods Development of filtering methods that update ambiguity sets without overreaction, enabling real-time adaptation to regime changes while preserving appropriate model skepticism.
Time-consistency proofs Theory Formal proofs establishing dynamic consistency conditions for multiperiod ambiguity averse preferences, resolving long-standing tensions between robustness and rational planning.
Historical regime testing Empirical Comprehensive out-of-sample validation across major market regimes, demonstrating worst-case performance improvements relative to mean-variance benchmarks under model misspecification.
Open-source toolkit release Tools Publication of robust Bayesian filtering code and multiperiod optimization templates, enabling independent researchers to replicate and extend our methodological contributions.
Institutional adoption study Applied Collaborative research with asset management firms documenting practical implementation challenges, committee communication strategies, and maintenance protocols for robust frameworks.
Conference keynote series Dissemination Invited presentations at premier quantitative finance conferences on multiperiod robust optimization, recursive ambiguity, and the future of model-aware portfolio construction.

Published work and collaborations

Selected papers, working papers, and collaborative research projects with leading finance departments and quantitative research groups.

Recursive ambiguity in dynamic choice
Published paper

Recursive ambiguity in dynamic choice

This paper establishes the axiomatic foundations for recursive multiple-prior utility in multiperiod portfolio settings. We prove dynamic consistency conditions and derive closed-form solutions for ambiguity-averse investors facing unknown return distributions. The framework nests classical expected utility as a special case while providing robustness against model misspecification.

Recursive utility Dynamic programming
Robust Bayesian filtering toolkit
Open-source tool

Robust Bayesian filtering toolkit

An open-source implementation of robust Bayesian filters for financial time series. The toolkit enables practitioners to construct ambiguity sets, update beliefs conservatively, and integrate filtering outputs into multiperiod optimization workflows. Includes documentation, examples, and stress-testing utilities for independent validation.
Bayesian inference Time-series filtering

Recognition and academic standing

Our research program has received citations, awards, and invitations from academic institutions, professional societies, and quantitative finance communities where robust methodology matters.

Distinguished research citation

Foundational paper on recursive ambiguity aversion cited as methodological cornerstone in survey of robust portfolio theory.

Best methodology award

Recognized for innovative robust Bayesian filtering approach that bridges theoretical rigor with practical implementation.

Keynote invitation

Invited to present multiperiod robust optimization framework at premier international quantitative finance conference.

Open science recognition

Acknowledged for transparent publication of code, data, and replication materials supporting reproducible research standards.