Portfolio construction stands at a crossroads. Traditional methods rely on mathematics removed from the actual mechanics of markets, treating investments as abstract tokens rather than entities embedded in complex systems. David Wood perceives this disconnect and works to bridge it. As Chief Quantitative Strategist and Co-Chief Technology Officer at Brooklyn Investment Group and Senior Managing Director at Nuveen, following the 2025 acquisition, Wood designs systems, where artificial intelligence doesn’t merely predict patterns but reasons through financial reality.
His doctoral work at the University of Chicago explored spherical space forms in statistics and econometrics, examining how mathematical structures can better represent financial phenomena. This theoretical foundation now manifests in practical systems that respect what he describes as “balance-sheet mechanics, market microstructure, risk transmission, tax and custody constraints”. Rather than building algorithms that chase historical correlations, Wood constructs models that internalize the structural dynamics governing how capital actually moves through markets.
Theory Collides With Market Reality
The path from abstract mathematics to functioning financial systems requires more than computational power. Wood’s decade-long academic journey encompassed economics, mathematics, quantum field theory, and string theory before culminating in his 2018 Ph.D. in Financial Econometrics. This interdisciplinary foundation enables him to translate theoretical concepts into algorithms that operate within real-world constraints. During his tenure as a quantitative researcher at Credit Suisse’s Quantitative Investment Strategies group, he developed systematic multi-asset strategies and researched regime models, trend following, and machine learning detection of jump points.
At Brooklyn Investment Group, Wood leads development of what the firm describes as “financially native” AI models—systems trained to understand markets through the lens of actual operational mechanics rather than statistical abstraction. This philosophy extends beyond traditional quantitative methods. Where conventional models might optimize for risk-adjusted returns in isolation, Wood’s frameworks account for how trading affects liquidity, how tax regulations influence investor behavior, and how market microstructure impacts execution.
“I’m constantly exploring ideas that connect abstract reasoning with real-world systems, drawing inspiration from how theory and practice inform each other,” Wood explains, describing his approach to continuous learning. His apartment houses thousands of books spanning philosophy, mathematics, finance, and engineering—a personal library that reflects his belief that breakthrough solutions often emerge at the intersections of disciplines.
The multi-asset custom direct indexing platform Wood helped the architect demonstrate this principle in action. Brooklyn’s technology enables personalized portfolio construction at scale, executing automated tax-loss harvesting across individual securities rather than fund-level instruments. The system doesn’t simply identify tax-saving opportunities; it reasons through replacement securities that maintain risk exposure while navigating wash sale rules and portfolio constraints. When implemented across hundreds of positions, this requires sophisticated optimization that balances competing objectives: tracking benchmark performance, harvesting losses, managing turnover costs, and respecting each client’s unique constraints.
Reconceiving Algorithms As Financial Agents
Traditional quantitative strategies treat portfolio construction as an optimization problem divorced from market participation. Wood’s work reframes the challenge: systems must function as participants within market ecosystems, understanding how their actions influence prices and how regulatory structures shape possibilities. This perspective leads to different algorithmic architectures. His collaboration with Professor Ruey Tsay on the MTS toolkit for multivariate time series analysis in R exemplifies this approach. The toolkit doesn’t merely fit statistical models to data; it provides instruments for analyzing how relationships between assets evolve across time, accounting for structural breaks and regime changes that invalidate static assumptions.
Research demonstrates that daily tax-loss harvesting yields approximately 30 basis points of additional annualized tax savings compared to monthly approaches. Yet capturing these gains requires technology that monitors market movements in real-time, evaluates replacement securities across various risk dimensions, and executes trades with minimal market impact. Wood’s systems at Brooklyn incorporate these capabilities while maintaining portfolio characteristics that align with benchmarks—a technical achievement that demands both mathematical sophistication and practical market understanding.
The firm’s partnership with Nasdaq resulted in the creation of the Nasdaq-Brooklyn ADR Index, which incorporates liquidity-adjustment factors for American Depositary Receipts. This seemingly technical innovation addresses a fundamental challenge: ADR liquidity often diverges significantly from the underlying company’s market capitalization, creating unexpected trading costs. By scaling index weights based on actual liquidity rather than market value alone, the methodology reduces transaction costs for separately managed accounts pursuing international exposure. This exemplifies Wood’s approach—recognizing where market reality deviates from theoretical assumptions and engineering solutions that navigate those frictions rather than ignoring them.
Reshaping Industry Architecture
Direct indexing through separately managed accounts represents one of the fastest-growing segments in asset management, reaching $864 billion by the end of 2024, with a 43% compound annual growth rate since 2020. Wood’s technological contributions help drive this expansion by making sophisticated portfolio customization economically viable on a large scale. Where personalized portfolio management once required intensive manual oversight, restricting availability to ultra-high-net-worth clients, automated systems now deliver comparable customization across broader investor populations.
His work extends across multiple asset classes. Brooklyn’s platform constructs unified solutions that pair tax-advantaged equity and fixed income strategies, enabling comprehensive portfolio management within a single account. This multi-asset capability addresses practical client needs while presenting substantial technical complexity, as different asset classes exhibit distinct liquidity profiles, regulatory treatments, and risk characteristics that require specialized modeling. Coordination across these domains demands systems that reason through cross-asset implications rather than optimizing components in isolation.
The firm’s partnership with Archer, a technology-enabled service provider, makes Brooklyn’s AI-powered platform available to asset managers seeking turnkey direct indexing capabilities. This distribution model amplifies impact beyond Brooklyn’s direct client relationships, potentially influencing how hundreds of advisory firms structure portfolios. Wood’s technological architecture becomes an infrastructure that enables industry-wide evolution toward personalized, tax-managed strategies. His role thus transcends individual firm contributions—the systems he builds reshape competitive dynamics and client expectations across the wealth management industry.
Implications For Quantitative Finance’s Trajectory
Wood’s emphasis on “financially native” reasoning signals a broader evolution in quantitative methods. Early quantitative finance imported techniques from physics and statistics, achieving powerful results but sometimes overlooking domain-specific structures. Machine learning amplified this pattern, with models demonstrating predictive success while remaining agnostic to financial fundamentals. The next generation of systems, exemplified by Wood, synthesizes computational sophistication with a deep understanding of finance.
This matters because markets evolve. Strategies that exploit static relationships eventually face diminishing returns as competition intensifies or the market structure changes. Systems reasoning through underlying mechanics rather than surface correlations may prove more adaptable. When Wood’s models internalize how market makers manage inventory or how tax-loss harvesting alters investor behavior, they operate from first principles rather than relying solely on empirical patterns. This grounding potentially enables more robust performance across market regimes.
His interdisciplinary background positions him to explore questions at the forefront of quantitative finance. How should AI systems incorporate regulatory frameworks that constrain but don’t determine outcomes? Can algorithms account for market reflexivity—the phenomenon where participants’ models influence the reality they attempt to predict? These challenges require not just more sophisticated mathematics but different conceptual frameworks, exactly the domain where Wood’s philosophical and theoretical training proves relevant.
