Portfolio managers can’t know every potential interplay among holdings through fundamental research alone. Our Portfolio Analytics and Risk Team provides key insights.
Portfolio managers prioritize security selection and portfolio construction. The former relies on an intimate knowledge of individual security valuation and fundamentals that may lead to powerful insights into future performance. The latter must be aided by mathematical analysis, as portfolio managers can’t know every potential interplay among portfolio holdings through fundamental research alone. This is where quantitative portfolio analytics come into play, providing multiple cross-sectional perspectives on performance and risk.
Quantitative insights must inform and enable action to support portfolio construction effectively. The definition of actionable may vary across strategies and through changing market conditions, underscoring the importance of continuously refining the analytics suite. Joint development effort between fundamental and quantitative professionals is critical to ensure careful consideration of all relevant factors for well-informed portfolio construction decisions.
Our collaborative approach involves a development cycle where quantitative analysts work closely with portfolio managers to generate new analytics ideas. We design and build analytics prototypes that are shared and tested via PowerBI while continuously receiving feedback from the investment team. If a new report proves valuable, it transitions to the Data and IT groups for production implementation. The existing report is seamlessly updated throughout this process, ensuring uninterrupted front-end usability.
For a quantitative portfolio analytic to be relevant, portfolio managers must understand its purpose and methodology. This understanding transforms information into insight, turning something interesting into something actionable. By bridging the gap between fundamental and quantitative domains of expertise through collaboration, teams improve the understanding of each other’s language and points of view. As a result, we increase the relevance of the sophisticated analyses and the completeness of the information the PMs consider for their portfolio construction decisions.
Analytics impacting portfolio construction and executive decisions can range from simple asset-weighted breakdowns of MPT (Modern Portfolio Theory) stats to complex factor-based risk management platforms. Regardless of the level of sophistication, every analytic must pass the ultimate usability test and become an intuitive tool for quantitative analysts and portfolio managers. Here is an incomplete list of self-serve reports that have successfully passed this test, followed by detailed case studies of the first three:
- Time series of drill-down stress test outcomes
- Factor-based active risk breakout and performance attribution
- Portfolio basket vs. Universe basket composition and performance
- Portfolio vs. Benchmark region and sector weights and fundamentals
- Unified price target and ESG ratings tracker
- Factor drift headwind and projections
- Best-worst case by factor stress test outcomes
- Drill-in and multi-period MPT (Modern Portfolio Theory) stats
- Daily fund flows and trading activity
Time Series Stress Test Results
Stress test analysis based on the foresight of the investment team is a crucial part of our risk management process. It supplements standard measures like ex-ante tracking error, value at risk, and expected tail loss. The risk team collaborates with portfolio managers to build relevant stress test scenarios. Ultimately, portfolio managers decide whether to act on the insights provided by stress test outcomes.
It can be challenging to connect the overall outcome of stress tests to the individual components of a portfolio and the interactions among these components. To help address this, we developed a self-service report that presents a heatmap grid of the contributions to stress test outcomes across regions and sectors, alongside another heatmap indicating over- and underweights in the portfolio versus the benchmark by the same sector/region decomposition. Since relative (among strategies and across time) magnitudes of the stress test outcomes are informative, we add a third dimension to the heatmap — an ability to see how the individual contributions change over time.
This detailed breakout allows investment professionals to connect the dots from the factor-based stress test result to more familiar views of portfolio positioning through the granular contributions from sector and region pieces. It also allows them to see how these outcomes have changed through time in the context of portfolio adjustments and evolving market conditions. Access to the fundamental interpretation of a quant result provides a more comprehensive assessment of the potential risks to the portfolio. Ultimately, it enables better-informed decisions regarding any necessary adjustments to portfolio construction.
Breaking Out Factor-Based Active Risks
A series of factor-based self-service reports facilitates our active risk budget process (see the article “Three Dos and Don’ts for Managing Risk in Active Portfolios” for details on our active risk budget process). These reports break down active risk (ex-ante tracking error) and factor exposures relative to the benchmark into individual and group risk factors. The risk team utilizes these reports as a quantitative lens for portfolio construction, and fundamental portfolio managers find them valuable for summarizing their positioning around critical strategy-specific guardrails during portfolio reviews.
While these provide vital input for the risk team’s analyses, the outcomes from factor-centric reporting alone may not be enough for a fundamental portfolio manager to act on. To see these analytics through a fundamental lens, we further deconstruct the active risk from econometric or statistical factors into contributions from geographic, sector, and industry positioning over time.
Yet another view into the changes in active risk components is to see whether portfolio exposure (weight) to a specific factor (industry, sector) has changed or whether it was the standalone riskiness (volatility) of the said factor or industry that changed. We automated tracking of these exposures, weights, and volatilities, along with drill-downs to contributions from individual names. On a more ad-hoc basis, we perform trade simulations that model the impact on a portfolio’s risk profile when specific securities – already or not yet held in the portfolio – are subtracted or added.
Portfolio Basket vs. Universe Basket Composition and Performance
Basketing the portfolio into groups of stocks that perform differently over time has long been essential for achieving fundamentally diversified portfolios at Thornburg. When considering Thornburg’s baskets, evaluating them in the context of market/benchmark baskets is crucial. The composition of market baskets changes over time, which requires consistent monitoring to construct appropriately sized and diversified baskets in the portfolio.
While the names held in Thornburg portfolios are thoroughly researched and assigned to appropriate baskets by portfolio managers and analysts, the same meticulous fundamental analysis process cannot be repeated for each name in the entire investible universe. To address this, we created a real-time report that parses the broader equity universe into Thornburg’s traditional basket framework. We accomplish this through a proprietary algorithm based on the same econometric factors in our risk management process.
Aligned with Thornburg’s approach to enhancing the investment process with analytical tools, this dashboard tool was designed collaboratively with portfolio managers, thus leveraging fundamental and quantitative expertise. The dashboard uses proprietary scoring to automatically assign investible stocks to baskets based on their relative ranking in factor exposures. The resulting assignments provide early warnings when a name in the portfolio may have shifted from one basket to another due to recent valuation or company fundamentals changes. Most importantly, this tool facilitates basket construction by helping portfolio managers think about and navigate the shifting basket compositions of the portfolio benchmarks.