What R-Squared Actually Tells You About a Portfolio

What R-Squared Actually Measures (And Why the Definition Matters)

R-squared measures the proportion of variance in a security's returns explained by movements in a benchmark index. Expressed as a value between 0 and 100, it answers one specific question: how much of this investment's behavior is the market's doing? That's it. Not whether the investment is good. Not whether the manager is skilled. Just the explanatory power of the benchmark.

The CFA Institute formally distinguishes R-squared from the Pearson correlation coefficient: R-squared equals the square of the correlation coefficient and measures explained variance, while correlation simply measures directional relationship. A correlation of 0.90 between a fund and its benchmark produces an R-squared of 81, not 90. That distinction matters when you're evaluating whether a fund's behavior is truly independent of market movements or just appears that way.

This is where most retail-oriented explanations stop. For a portfolio with meaningful complexity, that's the wrong place to stop.

How to Interpret R-Squared Values Between 0 and 100

Morningstar defines R-squared on a scale of 0 to 100, with values near 100 indicating a fund behaves very similarly to its benchmark index. The practical interpretation breaks into three ranges:

A fund with R-squared above 95 relative to its benchmark is widely considered a closet indexer. It tracks the index so closely that active management fees, typically 0.75% to 1.25% annually, represent pure cost drag with no incremental diversification or alpha potential.

For a $5M+ portfolio with $3M in actively managed equity, a 1% fee drag equals $30,000 annually in unnecessary costs. Screening your advisor's recommended funds by R-squared is one of the most direct ways to audit whether active management fees are earning their keep.

The SEC has published guidance confirming that even small differences in annual expense ratios compound significantly over time. An R-squared screen is the first filter, not the last.

What Is the Difference Between R-Squared and Beta in Portfolio Analysis?

R-squared and beta and market volatility are related but answer different questions. Confusing them is a common and costly mistake.

Beta measures the magnitude of a security's movement relative to the benchmark. R-squared measures how reliably that relationship holds. A fund with a beta of 1.2 and an R-squared of 45 moves dramatically when the market moves, but only about 45% of its variance is explained by the market. The rest is noise, or idiosyncratic risk.

Beta without R-squared context is misleading. If a fund shows a beta of 0.7 but an R-squared of 35, the beta figure is statistically unreliable. The benchmark explains only 35% of the fund's variance, so the beta calculation is built on a weak foundation. Your private banker should be presenting both numbers together.

For context on how standard deviation as a volatility measure interacts with these metrics, the relationship between tracking error and R-squared is particularly useful when evaluating whether a manager's active positions are adding or destroying value.

What Is a Good R-Squared Value for a Mutual Fund or ETF?

There is no universally "good" R-squared. The right value depends entirely on what you're paying for.

For a passive index fund, R-squared should be 98 or above relative to its stated benchmark. Anything lower suggests tracking inefficiency, often caused by cash drag, sampling methodology, or securities lending practices. When comparing S&P 500 index funds, R-squared differences of even 1–2 points can reveal meaningful operational differences in how closely the fund replicates the index.

For an actively managed fund, the question inverts. Vanguard research demonstrates that actively managed funds with high benchmark correlation rarely outperform their index counterparts net of fees over long time horizons. If you're paying active management fees, you should expect R-squared below 80, ideally below 70, as evidence that the manager is actually making differentiated bets.

Cremers and Petajisto's foundational NBER research introduced Active Share as a complement to R-squared, finding that funds with high Active Share (low R-squared to benchmark) and low tracking error significantly outperformed their benchmarks after fees. The combination matters: high Active Share with high tracking error often just means concentrated, volatile bets. High Active Share with controlled tracking error suggests disciplined, differentiated management.

The practical screen: if a fund charges more than 0.50% annually and shows R-squared above 90 to its benchmark, the burden of proof for keeping it is high. Ask your advisor to show you the fund's alpha generation over a full market cycle, not just a bull run.

What Does R-Squared Tell You About Active Fund Manager Skill vs. Benchmark Hugging?

R-squared is a necessary but insufficient test of manager skill. This distinction is frequently missed.

A fund can have a high R-squared (85–95) and still generate positive alpha if the manager consistently outperforms the benchmark in the same direction. Conversely, a low R-squared fund is not automatically a skilled active manager. It may simply be taking idiosyncratic risks that haven't been penalized yet.

The Journal of Finance research on mutual fund performance established that funds with low R-squared relative to a benchmark have historically shown greater potential for both outperformance and underperformance. Low R-squared widens the distribution of outcomes. It does not shift the distribution upward.

The correct framework for evaluating manager skill uses R-squared as a filter, not a conclusion:

1. Screen for closet indexers: Eliminate any active fund with R-squared above 90 to its benchmark. You're paying for differentiation that doesn't exist.
2. Validate beta with R-squared: Only interpret beta as meaningful when R-squared exceeds 70. Below that threshold, the benchmark relationship is too weak to rely on.
3. Pair with alpha analysis: For funds that pass the R-squared screen, evaluate whether the manager has generated statistically significant alpha over a full market cycle (typically 7–10 years minimum).
4. Check Active Share: R-squared and Active Share together give you a cleaner picture. High Active Share with R-squared below 75 suggests genuine differentiation.

For FatFIRE investors working with RIAs or family offices, this framework is a direct audit tool. An advisor presenting low-R-squared funds as inherently superior without demonstrating alpha generation over a full cycle is selling you a story, not a strategy.

What Is the Relationship Between R-Squared and Tracking Error in Index Funds?

Tracking error and R-squared measure related but distinct things. Tracking error is the standard deviation of the difference between a fund's returns and its benchmark returns. R-squared measures how much of the fund's variance the benchmark explains. A fund can have low tracking error (consistent small deviations) and still have a high R-squared. These metrics are complementary, not redundant.

For passive index funds, the relationship is straightforward: high R-squared should accompany low tracking error. A fund with R-squared of 99 and annualized tracking error of 0.05% is doing exactly what it promises. A fund with R-squared of 95 and tracking error of 0.40% is introducing unexplained variance, worth investigating before assuming it's benign.

Morningstar calculates R-squared using 36 months of trailing returns data by default. This creates a meaningful limitation: R-squared values are highly sensitive to the time period selected and can shift materially during regime changes. The same fund can show R-squared of 75 over a 3-year window and 92 over a 10-year window.

During the 2020–2022 period, many funds showed artificially low R-squared values due to elevated volatility and dispersion across sectors. Investors who interpreted that as genuine market independence were misreading a statistical artifact as a portfolio characteristic.

The practical implication: always request R-squared calculations across multiple time horizons (3-year, 5-year, full-cycle) before concluding that a strategy offers genuine benchmark independence. Rolling returns for long-term analysis provide useful context for evaluating whether R-squared stability holds across different market regimes.

How High-Net-Worth Investors Should Use R-Squared to Evaluate Concentrated Stock Positions

This is where R-squared becomes directly actionable for FatFIRE portfolios, and where most generic investment content never goes.

For concentrated positions, common among individuals who built wealth through equity compensation or founder shares, R-squared between a single stock and the S&P 500 informs tax-loss harvesting strategy in a precise way. A stock with R-squared above 85 to the index can often be replaced with a sector ETF during a wash-sale window without materially changing portfolio market exposure, while preserving the tax loss.

The mechanics: if you hold a large unrealized loss in a large-cap technology stock with R-squared of 88 to the S&P 500 and 94 to the Nasdaq-100, selling and replacing with a broad technology ETF maintains nearly identical market exposure. The IRS wash-sale rule prohibits repurchasing a "substantially identical" security within 30 days, but a diversified sector ETF is not substantially identical to a single stock. The R-squared analysis gives you quantitative evidence that the economic exposure is preserved.

For a $10M concentrated position, the tax efficiency of a well-executed harvest can be worth hundreds of thousands of dollars in a single year. R-squared is the analytical foundation for that decision.

The same logic applies to correlation matrices in portfolio analysis when evaluating whether adding a new position actually reduces portfolio concentration or simply adds a different name with nearly identical market behavior.

How R-Squared Is Used to Evaluate Hedge Funds and Alternative Investments

The application of R-squared to alternatives requires a significant caveat that most allocation frameworks ignore.

Research published in the Journal of Portfolio Management notes that private equity, hedge funds, and real assets often report artificially low R-squared values relative to public equity benchmarks due to infrequent mark-to-market pricing, not genuine diversification. Private equity funds that report quarterly NAVs based on internal valuations show smoothed return series that, by construction, correlate weakly with daily-priced public markets. The low R-squared is a statistical artifact of the valuation methodology.

When private equity returns are unsmoothed using methods like the Geltner adjustment, R-squared to public markets rises substantially. The apparent diversification benefit shrinks.

This matters for FatFIRE portfolios allocating 20–30% to alternatives. The reported R-squared of 20–40 for many private equity and real estate funds overstates their diversification benefit, particularly during liquidity crises when correlations across asset classes historically spike toward 1.0. The 2008–2009 period demonstrated this clearly: assets that appeared uncorrelated in normal markets moved together when liquidity dried up.

The practical framework for alternatives:

• Request unsmoothed return series or ask your manager how they handle valuation lag
• Evaluate R-squared across multiple time horizons, not just the most recent 36 months
• Treat reported R-squared below 40 for illiquid assets with skepticism unless you understand the valuation methodology
• Use private equity performance metrics alongside R-squared to get a complete picture of how alternatives actually behave in your portfolio

For hedge funds, R-squared to equity benchmarks varies enormously by strategy. Long/short equity funds often show R-squared of 60–80 to the S&P 500. Global macro strategies may show R-squared below 20. Neither number tells you whether the strategy is worth the fee structure without pairing it with alpha analysis and the Sharpe ratio for risk-adjusted returns.

R-Squared in the Active vs. Passive Management Debate

The active vs. passive debate is often framed around fees and long-term performance data. R-squared adds a more precise dimension: it tells you whether an active fund is actually active.

Vanguard's research is direct on this point: high-R-squared active funds rarely outperform their index counterparts net of fees over long time horizons. The math is straightforward. If a fund with R-squared of 93 charges 1.00% annually and its benchmark is available for 0.03%, the fund needs to generate 0.97% of annual alpha just to break even. Over a 20-year horizon, that's a compounding headwind that very few managers consistently clear.

The standard 60/40 guidance and most retail-oriented fund selection frameworks don't account for this. Someone holding a $5M actively managed equity allocation needs to run this analysis explicitly, not assume their advisor has done it.

Comparing major stock indices also matters here: a fund's R-squared is only meaningful relative to the correct benchmark. A small-cap value fund with R-squared of 70 to the S&P 500 may have R-squared of 92 to the Russell 2000 Value index. Always verify that the benchmark used for R-squared calculation is actually the appropriate comparison.

R-Squared Limitations Every Sophisticated Investor Should Know

R-squared is a useful tool with real constraints. Using it without understanding those constraints produces bad decisions.

It doesn't measure performance quality. A fund with R-squared of 95 to the S&P 500 could be up 30% or down 20% in a given year. R-squared tells you the source of returns, not their magnitude or direction. Pair it with realistic investment return expectations and market risk premium dynamics to get the full picture.

It is time-period sensitive. Morningstar's default 36-month calculation window means R-squared values computed during high-volatility periods may look artificially low. A fund that appears to offer genuine market independence based on 2020–2022 data may show much higher R-squared over a 10-year window. Always request multiple time horizons.

It doesn't capture tail behavior. R-squared is calculated using ordinary least squares regression, which weights all observations equally. It doesn't tell you how a fund behaves during market dislocations, which is precisely when correlation structure matters most for preservation of capital.

It is benchmark-dependent. The same security will show different R-squared values against different benchmarks. A global equity fund measured against the S&P 500 will show lower R-squared than when measured against the MSCI World. The choice of benchmark is a judgment call, and it can be manipulated to make a fund appear more or less correlated than it actually is.

Alpha and R-squared are mathematically independent. High R-squared does not mean no alpha. Low R-squared does not mean positive alpha. These are separate dimensions of fund behavior, and conflating them is a common error in advisor presentations.

Historical PE ratio trends and valuation context also affect how R-squared should be interpreted across different market regimes. A fund's correlation to the benchmark may behave differently in high-valuation environments than in periods of mean reversion.

References

• Morningstar -- "Morningstar Glossary: R-Squared"
• Morningstar -- "Morningstar Direct: Fund Statistics Methodology" (2023)
• Vanguard -- "The Case for Low-Cost Index-Fund Investing" (2023)
• CFA Institute -- "CFA Program Curriculum: Quantitative Methods" (2024)
• SEC -- "Investor Bulletin: How Fees and Expenses Affect Your Investment Portfolio" (2014)
• NBER (Cremers and Petajisto) -- "Active Share and Mutual Fund Performance" (2009)
• Journal of Finance -- "Mutual Fund Performance and the Incentive to Generate Alpha" (2004)
• Journal of Portfolio Management -- "Alternative Investments and Portfolio Construction for High-Net-Worth Investors" (2021)