Evaluating performance with attribution based on the investment decision process

Asset owners, including pension funds, insurance companies, and sovereign wealth funds, are operating in an increasingly complex investment landscape.

This complexity is driven by more dynamic asset allocation, heightened governance expectations, a mix of both internal and external managers and overlay programs to manage risks on fund level. In addition, increased merger and acquisition activity among funds in markets such as the UK and Australia has resulted in professional investment organizations managing a broad and very diverse range of asset classes and instruments.

Against this backdrop, evaluating total fund performance is no easy task. To enable asset owners to gain a comprehensive, fund-wide view, a framework is needed to capture not just the portfolio structure, but the underlying decisions resulting in such a structure of individual portfolios. It requires a top-down decision-based attribution model, that helps to explain multi-asset processes, overlay strategies, and the link between policy and implementation. Such a model can identify the strong and weak points in the investment process, which in turn is useful information to improve future performance.

Why decision-based attribution contributes to better investment decisions

An asset owner’s investment process follows a top-down hierarchy of investment decisions, with each decision contributing to the overall profit or loss on fund level. In such a top-down framework, the measured value of higher-level (earlier) decisions should not be affected by (operational) decisions made at lower levels. As such, the value added by the decision to allocate capital to active management should be evaluated independently of the alpha generated by any specific manager.

By understanding the excess gain of a fund through the contributions of investment decisions, asset owners can identify key factors that drive performance. This is done by modeling the full investment decision process, from strategic and tactical decisions to overlay strategies, manager allocation and security-level choices. This Investment Decision Process (IDP) approach allows for the measurement of each decision’s contribution to the total fund performance.

How does attribution based on IDP differ from standard approaches?

Standard attribution approaches, such as well-known equity attribution models, explain performance from a portfolio structure perspective whereas IDP enables performance to be explained from a decision-making perspective. It quantifies the impact of every decision - from high level strategic asset allocation and currency policy to implementation choices like duration positioning or stock selection. It not only captures explicit decisions but also helps to understand implicit sources of risk and return like benchmark mismatches and asset allocation drifts.

It connects the dots between strategic intentions, their implementation, and the final outcomes - something standard attribution models can’t fully capture. This makes it especially powerful for multi asset funds, fiduciary managers, overlay strategies, and any investment process where performance is driven by a layered series of decisions. By going beyond the allocation and selection effects typically examined in standard attribution, decision attribution expands the understanding of why a fund outperformed or underperformed by directly linking results to the investment decisions that shaped the portfolio.

The IDP approach - Measuring value through modelling decision layers

To measure the value of each decision, specific benchmarks for each of these decision layers are imperative. Each decision serves as a reference for the next, with the value added measured by the difference in results between the layers, reflecting the actual asset management process.

In the IDP framework, benchmarks naturally follow from these decision layers. Each decision serves as a reference for the subsequent one, with the value added measured by the difference in results between layers, precisely reflecting the asset management process. Under the IDP approach, the added values from all decisions will always aggregate to the overall profit or loss of the investment strategy.

These insights offer a clearer, more intuitive explanation of performance, which resonates with stakeholders managing complex portfolios across a broader range of asset classes, and supports a stronger feedback loop for future decisions.

Overcoming additional performance measurement challenges with IDP

As the IDP approach mimics the actual investment decision process, it can overcome several common challenges faced by investment performance professionals, by:

• Measuring the unintentional allocation effect of illiquid assets by adding a decision layer that reflects the difference between intended and actual allocated weights.
• Avoiding issues arising from benchmark inconsistency and risk-adjusted performance measures, as decision layers are evaluated independently of one another
• Accurate modeling of multi-asset class portfolios, recognizing that different asset classes may often employ different investment strategies.
• Reducing pitfalls in overlays, absolute return, hedge funds, and leveraged investments caused by poorly defined weights and returns