- Analyzing 94 actual portfolios from 2008 and 2019, we find asset selection accounted for anywhere between 60% and 78% of tracking error between portfolios and benchmarks over any given year.
- During past market disruptions, such as the 2008 global financial crisis, variation in both allocation and selection increased, but the increase for selection was more pronounced.
- With the COVID-19 pandemic causing unprecedented disruption, attribution analysis may be a useful tool for helping investors understand and explain what drove returns in their real estate portfolios.
Real estate, along with politics and sports, must be one of the world’s favorite obsessions. There is often plenty written about the wider trends shaping real estate markets and no shortage of opinions. In the current crisis this may be especially true. But real estate is a heterogeneous asset class: Every property is unique, and no two portfolios will provide the same returns.
Similar statements apply to individual real estate portfolios, which can be shaped by a number of specific factors that will influence performance relative to the broader market. The use of attribution analysis may help investors understand and communicate their relative performance, separating the impact of allocation differences from the impact of asset selection within those allocations, and providing the means to position performance within the context of wider market trends. Our analysis of historical allocation and selection scores from actual portfolios suggests that asset selection may have been a particularly important driver of relative performance, especially in times of crisis. With the COVID-19 pandemic expected to continue causing considerable disruption to real estate markets, attribution analysis may be a useful tool.
Attributing Relative Portfolio Performance
Asset-level attribution analysis can help investors understand the reasons for a portfolio’s outperformance or underperformance versus an index. It breaks down the relative return into structure- and property-specific scores, allowing the influences of submarket allocations and asset selection to be distinguished.
Specifically, attribution analysis distinguishes that part of the relative return derived from the portfolio’s weightings in strong or weak sectors of the market (allocation), from that part derived from the performance of the assets in the portfolio within each segment of the market (selection).
Using data from 94 real portfolios we explored how allocation and selection scores contributed to relative returns between 2008 and 2019. Over the entire analysis period, selection accounted for 70% of the annual tracking error between portfolios and benchmarks, with its contribution ranging from a low of 60% in 2018 to a high of 78% in 2016.
Selection Accounted for 60% to 78% of Observed Tracking Error in Any Given Year
During periods of market stress like the 2008 global financial crisis, as the spread in relative returns between portfolios increased, we observed more variation in both allocation and selection scores. However, the increase tended to be greater for selection, meaning that it tended to account for more tracking error in these years.
Why the Lower Contribution from Allocation?
One potential explanation for the lower contribution we observed from allocation could be that real portfolios are built and managed while referencing benchmark weightings that may constrain the allocation scores. Another explanation may lie in the construction of benchmark segmentations themselves. Traditionally, segmentations have been built along property type and geography lines, but underlying these groupings may be additional systematic drivers of return that could be used to construct alternative segmentations, which, if implemented, could potentially increase how much tracking error is explained by allocation.
Putting Portfolios into a Broader Market Context
As strategies become more sophisticated and investors demand more insight into the drivers of risk and return in their portfolios, attribution analysis may provide valuable insights. Our analysis showed that asset selection played an important role in driving the relative returns of real estate portfolios from 2008 to 2019, particularly during periods of disruption. With the COVID-19 pandemic causing unprecedented disruption, attribution analysis may be a useful tool for understanding and communicating the idiosyncratic aspects of portfolios.