CONTRIBUTORS
Adrian Chan, CFA
Head of Quantitative Investments
Brett Risser
Head of Quantitative Equity Portfolio Management
Jacqueline Kenney, CFA
Head of Solutions Research
Lisa Wang, CFA
Head of EMEA Investment Strategy
Subash Pillai
Head of APAC Solutions
Executive summary
- Institutional investors increasingly rely on multiple active managers with repeatable processes and strong returns. Yet portfolio-level alpha is not always a perfect proxy for underlying security-level insight.
- Combining portfolios often dilutes conviction, introduces unintended risk concentrations and can create an overall portfolio that too closely resembles a benchmark index.
- We introduce a framework we name “Capturing Conviction” which addresses this challenge by identifying the return expectations implicitly embedded in a manager’s holdings (which we dub “implied alphas”)—exposing security-level conviction independent of broader portfolio structure.
- Rather than interpreting raw active weights as conviction, the approach derives the hidden return assumptions that rationalize the portfolio within modern portfolio theory.
- The resulting implied alpha vector captures each manager’s stock-selection insight while adjusting for volatility, correlation and portfolio constraints.
- These signals are optimization-ready and can be orthogonalized to remove unintended systematic exposures such as beta, sector or style tilts.
- Compared to traditional regression-based stripping or naïve active-weight replication, Capturing Conviction provides a forward-looking, economically grounded and scalable approach to portfolio design.
- By extracting implied security return expectations from multiple managers, allocators can aggregate these views into a unified signal and construct a new, integrated portfolio of best ideas, enabling them to separate stock-selection insight from portfolio construction decisions.
- This reconstructed portfolio can be aligned to a chosen benchmark, calibrated to a desired tracking error and constrained to meet institutional risk objectives.
- The framework supports enhanced indexing, multi-manager best-ideas portfolios, “quantamental” blending and improved manager oversight.
The challenge facing institutional investors today is not simply identifying manager skill—it is deploying that skill efficiently within a broader portfolio context. Traditional approaches to combining active managers often embed structural biases and dilute conviction. As a result, portfolios can become more complex without becoming more differentiated.
The Capturing Conviction framework reframes this problem at its source. By reverse-engineering the expected returns embedded within fundamental portfolios, it translates qualitative conviction into a quantitative, risk-adjusted signal grounded in modern portfolio theory. This transformation allows allocators to extract the informational content of active portfolios while disentangling it from benchmark structure, factor tilts and idiosyncratic construction choices.
Most importantly, Capturing Conviction enables a structural shift in portfolio design. Rather than stacking managers and accepting portfolios as fixed building blocks, institutions can reconstruct an integrated portfolio of best ideas across managers—aligned to their benchmark, calibrated to their risk budget and governed within a transparent optimization framework. Institutions can preserve conviction and enhance control.
In an environment where alpha is increasingly scarce and risk budgets are finite, the ability to maximize active return per unit of risk is paramount. By bridging fundamental insight and quantitative rigor, Capturing Conviction offers a scalable, governance-friendly pathway to more efficient active portfolio construction. For allocators seeking to move beyond manager selection toward true portfolio-level skill extraction, it provides a coherent and economically grounded solution
GLOSSARY
Quantamental describes a combination of quantitative and fundamental approaches.
Active return: Difference between a portfolio’s return and the return of its benchmark
Tracking error: Standard deviation of a portfolio’s active returns versus its benchmark (measures the volatility of benchmark-relative performance)
Information ratio: Active return per unit of tracking error.
Transfer coefficient: Correlation between signal and resulting active weights under mean-variance optimization.
Correlation of active returns with multi-manager alpha: Correlation between reconstructed and original multi-manager outperformance/underperformance versus a benchmark, measured using the time-series correlation of active returns.