Leveraging quantitative tools for portfolio construction

Portfolio construction is an inherent aspect of portfolio management and a critical consideration for institutional investors. In our view, creating well-balanced, diversified portfolios that align with investment objectives, risk tolerance, and the desired time horizon is essential in achieving strong risk-adjusted returns. 

At Manulife Investment Management, we do this by taking a multidimensional approach to portfolio construction. Our managers have well-defined processes, crafting portfolios that align with their investment styles and strategy objectives. In addition, they benefit from the analysis and tools provided by our investment risk and construction solutions (IRCS) team. In this paper, we explore these tools and their applications in addressing diversification, position sizing, and style consistency. 

How quantitative tools are used in portfolio construction

Our tools cover a wide range of considerations, from traditional risk factor decomposition to more innovative metrics such as our proprietary fundamental correlation and true alpha estimates. The IRCS team’s comprehensive support enables our managers to construct portfolios that aren’t only well diversified and appropriately sized but also consistent with their investment style. Our analysis focuses on three distinct risk categories: diversification, position sizing, and style consistency. 

1 Diversification

When a portfolio grows overly concentrated in certain assets, sectors, or geographic regions, it becomes more vulnerable to specific market events or economic downturns affecting those areas. Insufficient diversification can lead to higher volatility and larger losses if the concentrated investments perform poorly. 

We employ a range of traditional approaches to identify and manage diversification risk, including active weight analysis, factor analysis, active risk decomposition, correlation comparison, and concentration analysis. The IRCS team dissects risk-taking into specific components such as security risk contributions, sector contributions, and fundamental exposures. This comprehensive approach ensures a thorough understanding of diversification within our portfolios.

Moreover, we employ additional tools, such as our fundamental correlation metrics, clustering techniques, dendrogram analysis, and diversification tables, to provide a quantitative approach to maintaining a balanced portfolio. These tools enable us to address the challenge of diversification and enhance the robustness of our portfolio construction process. 

In addition to price correlation, we measure the similarity of names based on their fundamental characteristics. We present the fundamental correlation between assets in a table format, where the color coding provides a quick visual cue.

Dendrograms offer a clustered visual representation of these similarity measures. They provide an intuitive way to understand and analyze the relationships between different assets based on their fundamental correlations. When names are very similar, it’s advisable to examine their total weight in the portfolio. This helps to assess the concentration risk and ensure a balanced portfolio construction. In addition, dendrograms can quickly identify names that have different exposures to the average portfolios. While these are usually diversifiers in a portfolio, they need to be evaluated further to ensure alignment with the portfolio manager’s style of investment.

Diversification tables: enhancing portfolio balance

Diversification is a key tenet of risk management, and our diversification tables provide a practical solution for managing overweight and underweight positions within a portfolio. By offering a list of assets with similar characteristics, this tool enables managers to explore alternative investments that can enhance diversification and reduce concentration risk. Whether addressing significant overweight or seeking to mitigate risk in underweight positions, the diversification tables facilitate a more informed decision-making process, ultimately aiming to create a more resilient portfolio.

Managing downside risk

One of the fundamental pillars of active portfolio management is the ability to protect against downside risk. Recognizing the significant impact that large drawdowns can have on portfolios, we’ve developed a specialized tool designed to identify assets that tend to move together during downside markets. The key advantage of this method is its focus on co-movements that occur during market downturns, which are often the most detrimental to portfolio performance.

Drop dependence: measuring asset movement in a market downturn

Unlike traditional price correlation measures, which are linear and treat upside and downside movements equally, our tool specifically targets downside co-movements. This approach allows us to better manage and mitigate risks associated with adverse market conditions without penalizing beneficial upside co-movements. By incorporating this tool into our decision-making process, we enhance our ability to safeguard investments and maintain stability during volatile periods. 

2 Position sizing

Allocating inappropriate amounts of capital to individual investments, either too much or too little, can lead to an imbalance in the risk/return profile of the portfolio. 

The Brinson attribution model is a widely recognized method that breaks down a portfolio’s performance relative to a single benchmark into selection and allocation effects; however, we employ an innovative approach that extends beyond this traditional framework. Our method involves comparing our portfolio not only against a single benchmark, but also against a broad set of potential alternatives with similar risk levels. This allows us to calculate our selection and sizing efficacy more comprehensively.

By evaluating our selection decisions against a range of possibilities, we gain insight into how our choices perform relative to many potential alternatives rather than just one benchmark. Additionally, we assess our sizing decisions by comparing them with various alternatives to determine if they’ve provided superior risk-adjusted returns. This multifaceted analysis helps ensure that our investment strategies are robust and in the best shape to aim for strong outcomes.

True alpha: unveiling manager conviction

One of the cornerstone tools in our arsenal is the true alpha score, a metric that quantifies a manager’s conviction in specific investments. This tool addresses the challenge of identifying and quantifying idiosyncratic conviction within a portfolio. By analyzing active weight, estimated volatility, and differences from a manager’s typical names, true alpha provides a nuanced view of where a manager’s strongest beliefs lie. This insight is crucial for portfolio managers seeking to align their sizing with their highest-conviction ideas, ultimately leading to more informed investment decisions.

3 Style consistency versus drift  

Style drift refers to the deviation of a portfolio’s investment strategy from its stated style or asset allocation over time. This can pose significant risks to investors who rely on a strategy’s stated style to align with their own investment goals and risk tolerance. Style drift can also lead to unexpected changes in the risk/return profile and can complicate the process of maintaining a balanced and diversified portfolio, as the portfolio’s actual holdings may no longer fit the investor’s strategic asset allocation plan. 

We employ a range of tools to help identify style drift, including style analysis, which uses grids to visualize and compare current portfolio holdings against historical style categorizations.

Holdings-based analysis examines the characteristics of the securities held in a portfolio, such as market capitalization and fundamental characteristics, to detect shifts in investment style. Factor exposure analysis assesses the portfolio’s exposure to different risk factors, such as size, value, and growth, to detect changes in investment style. Additionally, tracking error and active share metrics measure the volatility of the portfolio’s returns relative to a benchmark and the percentage of the portfolio that differs from the benchmark, respectively. 

Furthermore, we employ our fundamental correlation, cluster analysis, and dendrogram tools to help managers ensure that their portfolios remain aligned with stated objectives while minimizing unintended deviations.

Cluster charts help identify similar pairs or clusters, as well as names that act differently, serving as diversifiers or a flag on potential deviations from the process.

Additional metrics: trading patterns and performance analysis

Further to the measures outlined above, the company uses several other tools to help identify and manage risks within our portfolios; trading analysis, for example, allows us to examine our buy and sell decisions to help identify recurring patterns. Do managers typically purchase assets when their valuations become more attractive, or do they tend to sell following a price decline? This analysis is then contrasted with each team’s investment process to ensure alignment and adherence to our established strategies. By doing so, we can ensure that our trading activities consistently reflect our investment principles and objectives.

We also use many standard tools to help evaluate the sources of a portfolio’s returns relative to a benchmark. These tools typically include returns-based attribution, which analyzes the portfolio’s performance by comparing it against a benchmark and attributing differences to asset allocation or security selection. Factor-based attribution assesses the contribution of various risk factors, such as market, size, and value, to the portfolio’s returns. Additionally, multiperiod attribution allows for the analysis of performance over multiple time periods, providing insight into the consistency and sustainability of investment strategies. These tools collectively help managers understand the drivers of performance and make informed decisions.

Multidimensional risk management

Effective portfolio construction is critical to achieving strong risk-adjusted returns. At Manulife Investment Management, we employ a comprehensive approach that integrates quantitative analysis with the insight of our fundamental managers. Our managers use in-depth research to construct well-balanced portfolios while also leveraging the quantitative tools and solutions developed by our IRCS team.

This approach aims to ensure that our portfolios aren’t only well diversified and aligned with investment objectives but also resilient to market changes. By combining quantitative insight with fundamental expertise, we strive to deliver superior investment outcomes to meet the diverse needs of our clients.