Hidden Risks of Overdiversification: When Less is More

The purpose of this research is to determine the optimal number of core real estate funds that investors should hold to achieve effective portfolio diversification while balancing tracking error and information ratio. By analyzing 10,000 simulated portfolios of U.K. core real estate funds, we provide actionable insights to help institutional investors streamline fund selection, enhance portfolio efficiency, and align with performance objectives. Key findings include:

• A portfolio of three to five funds achieves sufficient diversification, reducing TE to ~2% relative to the benchmark.
• Marginal diversification benefits diminish sharply beyond seven funds and plateau by 10 funds.
• Smaller portfolios allow better risk-adjusted returns by focusing on top-performing funds, as IR converges toward zero in larger portfolios.
• The ability for a core fund portfolio to outperform its benchmark quickly diminishes as the number of funds held increases.
• Hypothetical modeling suggests these findings extend to U.S. core funds, with idiosyncratic risk reduced by 80% in a six-fund portfolio.

The following steps outline the methodology used:

1. Data collection
We collected quarterly total return data for 40 U.K. private market core diversified real estate funds, spanning from Q1 2006 to Q4 2023. Although additional fund performance history is available, we felt the number of funds covered by the index were too few prior to 2006. In the first quarter of 2006, the index contained a reasonably robust sample of 28 funds. This dataset represents a broad cross-section of the U.K. core fund universe and captures varying performance profiles over nearly two decades. The UK All Balanced Open-Ended Property Fund Index served as the benchmark, representing an equal-weighted average of balanced open-ended property funds in the U.K.

2. Portfolio Simulation
We simulated a total of 10,000 random portfolios comprised of x-number of randomly chosen funds (from x=1 to x=20), For each randomly generated portfolio, we calculated performance metrics against the equal weighted UK All Balanced Open-Ended Property Fund Index.

We chose a random number of funds (from 1-20) and constructed an equally weighted portfolio; then calculated the tracking error, information ratio, and relative performance against the benchmark. We did this 500 times for one-fund portfolios, two-fund portfolios, three-fund portfolios, until 20-fund portfolios.

Next, we examined the distribution of performance metrics in each portfolio-size group and assessed at what point we can achieve certain investment criteria.

We calculated these measures for each portfolio, across the entire range of portfolio-sizes.

• We then presented the results as boxplots, showing the distribution of performance metrics across the range of portfolio sizes. To interpret a boxplot there are a few key traits:
  • the median line within the box,
  • the length of the box (representing the interquartile range—IQR)
  • the length of the whiskers
  • any data points plotted outside the whiskers (outliers)
• Together, we gain an understanding of the central tendency, spread and potential skewness of the data distribution. We can compare the boxplots of different groups to see how their distributions differ in terms of center and spread.

3. Performance Metrics
For each portfolio, we calculated tracking error as the standard deviation of excess returns relative to the benchmark. TE measures how closely a portfolio tracks the benchmark and serves as a key indicator of relative risk.

The information ratio for each portfolio was calculated by dividing the average excess return by its tracking error. This metric evaluates the risk-adjusted return relative to the benchmark, helping to determine whether increased diversification enhances or dilutes potential outperformance.

4. Analysis process
By analyzing the distributions of TE and IR across the 10,000 simulated portfolios, we assessed the relationship between portfolio size and diversification benefits. This approach allowed us to identify thresholds where additional funds no longer significantly reduce TE or improve IR. Based on these distributions, the number of funds required to achieve specific TE or IR targets were determined. The results provide actionable insights for investors looking to balance diversification and performance.

5. U.S. market hypothetical modeling
To extend the analysis to U.S. core real estate funds, we modeled hypothetical portfolios based on assumed correlations of 0.7 between funds, consistent with U.S. Open-end Diversified Core Equity (ODCE) fund characteristics. The results suggest that idiosyncratic risk diminishes by approximately 80% with a six-fund portfolio, reinforcing the relevance of our findings across markets.

Assumptions and limitations
Portfolios were constructed with equal weights to simplify the analysis and focus on fund count as the primary variable. While real-world portfolios may apply different weighting strategies, this assumption isolates the core question of optimal fund number. The study relies on U.K. fund data, which benefits from public availability. While similar data for U.S. funds is unavailable, the findings are expected to hold due to comparable market dynamics.

The results highlight the inefficiencies of overdiversification in core real estate fund portfolios and provide actionable insights for institutional investors. By examining the relationship between the number of core funds held in a portfolio, tracking error and information ratio, we arrive at the following findings and implications:

1. Diminishing benefits of diversification: A concentrated portfolio of three to five funds achieves sufficient diversification, reducing tracking error to ~2% relative to the benchmark.

Chart 1: Tracking error distribution by number of randomly selected funds, %

Marginal diversification benefits diminish sharply beyond seven funds and plateau entirely by 10, offering little relative risk-reduction benefit. Larger portfolios dilute IR as out- and underperformers converge to the average, while incurring higher costs, such as manager fees and transaction expenses. For investors able to identify outperformers, a smaller portfolio of funds is more efficient and effective.

Chart 2: Marginal change in tracking error as funds are added

An analysis of the relationship between information ratio and performance across a distribution of funds did not reveal any benefits of building larger portfolios (see Chart 3). This is intuitive, since the average information ratio should be close to zero.

Chart 3: Information ratio distribution by number of randomly selected funds

Beyond reducing risk, larger portfolios also significantly diminish the likelihood of generating excess returns, as demonstrated in the probability analysis in Chart 4 below. As more funds are added, the probability of outperforming the benchmark declines sharply across all outperformance thresholds. This is particularly evident when we include an outperformance hurdle. If the target is to outperform the benchmark by 50bps, a randomly selected three-fund portfolio achieved that mark 35% of the time, while a 10-fund portfolio did so in less than 20% of cases. The trend is even more pronounced for higher return targets—only 1% of 10-fund portfolios exceeded the benchmark by 100bps. The consistency of this pattern across different levels of outperformance reinforces the trade-off between diversification and return potential.

Chart 4: Portfolio outperformance, %

2. U.S. market applicability: Although this study is based on U.K. core fund data, hypothetical modeling suggests similar results for U.S. core real estate funds. With an assumed fund correlation of 0.7, idiosyncratic risk in a U.S. portfolio diminishes by approximately 80% with only a six-fund portfolio. This supports the validity of the findings across different real estate markets.

Chart 5: Fund specific risk, %

Institutional investors can achieve sufficient diversification and a reasonable tracking error with just three to five funds. This concentrated approach greatly improves the likelihood of outperforming the benchmark and avoids the pitfalls of over-diversification.

These findings, though based on U.K. core fund data, likely extend to U.S. and other real estate markets. Hypothetical modeling of U.S. ODCE funds reinforces this, showing significant risk reduction with smaller, concentrated portfolios. While this study did not analyze whether fund outperformance persists over time, this relationship merits separate research.

Key recommendations:

• Avoid overdiversification: Limit fund count to reduce costs and minimize risks associated with underperforming funds. As more funds are added to a portfolio, the ability to outperform the benchmark quickly diminishes.
• Balance risk and return: Use TE and IR as guiding metrics to align portfolio construction with investment goals and risk appetite.

Investors are encouraged to reassess their portfolios, reduce redundancy, and adopt a more concentrated strategy to optimize performance and manage risks effectively.