Key Takeaways

• One of Venn’s four core pillars is to be holistic, meaning that our approach to portfolio analytics attempts to explain the large majority of return variation. 

• In addition to providing a parsimonious, orthogonal, and actionable way to analyse investments, Venn’s Two Sigma Factor Lens appeared to explain more than 90% of the total variation in representative institutional portfolios. 

If an investment returned that same exact amount every month, the only relevant question is likely: “What is the return?” In reality, investment returns go up and down which invites the opportunity for deeper analysis.

One’s first instinct may be to understand what is driving portfolio variation. While this is a great place to start, equally important is what percentage of total portfolio variation one can explain.¹

This is why one of Venn’s four core pillars is to be holistic, meaning that our approach to portfolio analytics attempts to explain the large majority of return variation. To achieve this goal, Venn’s Two Sigma Factor Lens uses a “less-is-more” approach, using 18 or fewer statistically independent risk factors that span across asset classes. 

Measuring the holisticness of a less-is-more approach in portfolio analytics
To test how holistic our Factor Lens has been, we constructed five representative institutional portfolios for five different investor types.² In Venn, we decomposed these portfolios into common risk factors, but also separated risk into what our Factor Lens can and cannot explain. The risk unexplainable by our Factor Lens is called residual.

As a result, we are able to paint a clear picture as to what percentage of portfolio variation our common risk factors are capturing among some of the largest investors in the world. As shown below, we believe that Venn’s Two Sigma Factor Lens not only explained the majority of the risk but in all cases, it explained more than 90% of the total variation in these representative institutional portfolios. 

Spoiler: Equity, Interest Rates and Foreign Currency Factors explained a large majority of risk in these portfolios over the 5-year period ending Dec 31, 2021.

Percentage of risk explained by Venn’s Two Sigma Factor Lens for institutional investor types

Source: Venn by Two Sigma. See the appendix for more information on the construction of representative institutional portfolios. The period analysed was the trailing 5-year period ending Dec 31, 2021. Portfolios were rebalanced quarterly.

Are there other ways to explain this much risk?
The short answer as to whether other approaches can explain this much risk is: yes. In fact, in 2018 we found an asset class framework to have similar explanatory power. However, as we mentioned earlier, there is more to Venn than just being holistic.

More specifically, Venn uses 18 or fewer factors to reduce confusion from overlapping risks among highly correlated asset classes. In addition, the independent nature of our factors may lead to better and more precise decision-making. As we have shown, these characteristics can be achieved while still explaining over 90% of the risk in institutional portfolios.

Put another way, when we think about Venn’s four fundamental pillars (see below), holistic is unique. It is not so much that Venn is holistic, but rather, that Venn is holistic while also providing a parsimonious, orthogonal, and actionable way to conduct portfolio analytics. 

Venn’s four fundamental pillars:

• Holistic, by capturing the large majority of cross-sectional and time-series risks for typical institutional portfolios. 
   
• Parsimonious, by using as few factors as possible.
   
• Orthogonal, with each risk factor capturing a statistically uncorrelated risk across assets.
   
• Actionable, such that desired changes to factor exposure can be readily translated into asset allocation changes.

Stay tuned for our fourth piece on the last fundamental pillar of Venn.

Appendix
The authors referenced various sources for these institutional portfolio allocations. Those original source allocations can be found in the footnote of the table.

They then used their best judgment to consolidate and match source allocations with existing publicly available indexes. The results of that exercise are shown in the table below, detailing each portfolio's weight to different representative indexes. 

To create the first chart displayed, each portfolio was created in Venn using the matching index and corresponding weight to represent each institutional investor type. These portfolios were rebalanced quarterly and analysed in Venn’s risk decomposition module over the five years ending 31st of December, 2021. Venn breaks down each portfolio’s risk into common risk factors, but also reveals what percentage of risk cannot be captured by our Two Sigma Factor Lens, also known as residual. 

100% minus this residual gives us the percentage of total risk that Venn’s Two Sigma Factor Lens is able to explain in these representative institutional portfolios. The percentage of risk explained by Venn’s Two Sigma Factor Lens, or 100% minus the residual, was featured in the first chart.

Representative institutional portfolios 

Sources: Public Pensions: “Investment”, NASRA, Endowments and Foundations: “NACUBO-TIAA Study of Endowments”, NACUBO (2021), Sovereign Wealth Funds: “Global Traditional Asset Allocation”, Statista Research Department (May 23, 2022), Family Office: “Global Family Office Report 2022”, UBS (2022), Insurers: “Capital Markets Special Report”, NAIC (2021).

References
¹ This concept is similar to the statistical concept of R squared, which is the per cent of variance of a dependent variable explained by explanatory variables

² See appendix for details on construction

Read the original article here.