Tracking Error Calculation: Measuring Active Portfolio Risk

Key Takeaways

• Tracking error is annualized by multiplying the periodic standard deviation of active returns by the square root of periods per year, sqrt(252) for daily, sqrt(12) for monthly.
• Enhanced-index funds typically target tracking error under 1 percent; high-conviction concentrated managers may budget 6 to 10 percent annually.
• A common mistake is treating high tracking error as a skill signal, without a corresponding active return in the information ratio, it is just uncompensated risk.
• GIPS standards require a three-year annualized ex-post standard deviation calculated from monthly returns, providing a consistent basis for comparing funds.

What It Is

Tracking error (also called active risk or tracking risk) is a single number that summarizes the dispersion of active returns. A fund that follows the S&P 500 within plus or minus 0.1 percent per year has a tiny tracking error. A concentrated long-only stock picker that deviates by 5 to 8 percent per year has a large one. Neither value is intrinsically good or bad. The right tracking error depends on the mandate.

Tracking error is also the denominator of the information ratio, which measures active return per unit of active risk. A high information ratio with modest tracking error is often more valuable than a high absolute alpha with wild swings.

The Intuition

Active return alone tells you how much a manager beat or missed the benchmark. It does not tell you whether the result came from one or two big bets or from many small consistent ones. Tracking error fills that gap. Two managers can both deliver 2 percent active return over a year. One might have done it with steady 0.2 percent monthly outperformance (low tracking error). The other might have done it with wild swings (high tracking error). The first manager is demonstrating repeatable skill. The second may be taking on risk that has not yet bitten.

Allocators pair the tracking error with the active return to decide whether the strategy fits a mandate. An enhanced-index fund is expected to have tracking error under 1 percent per year. A high-conviction concentrated fund might budget 6 to 10 percent. A market-neutral fund, typically benchmarked to cash, can run tracking error of any size because it has no index it is trying to follow.

How It Works

Let Rp_t be the portfolio return in period t and Rb_t be the benchmark return. Define the active return AR_t = Rp_t - Rb_t. Tracking error is the sample standard deviation of active returns:

TE = sqrt( sum over t of (AR_t - mean(AR))^2 / (n - 1) )

Where:

AR_t    = active return in period t
mean(AR) = average active return over the sample
n       = number of periods

Periodic tracking error is annualized by multiplying by the square root of the number of periods per year. Daily TE is multiplied by sqrt(252), weekly by sqrt(52), monthly by sqrt(12). The GIPS standards require a three-year annualized ex-post standard deviation of returns for composites and benchmarks (a related but different figure), calculated from monthly returns.

The information ratio is:

IR = mean(AR) / TE

Both active return and tracking error should use the same return type (gross or net of fees) and the same frequency. Mixing them produces nonsense.

An ex-ante tracking error is also common: a risk model (Barra, Axioma, or a Fama-French-style factor model) forecasts tracking error from current portfolio holdings and factor covariances. Ex-ante and ex-post tracking errors often differ, and the gap is itself diagnostic.

Worked Example

A US large-cap fund returns 10.8%, 7.2%, 12.5%, and 9.1% over four quarters, against an S&P 500 benchmark returning 10.0%, 7.5%, 11.8%, and 8.9%.

Active returns: +0.8%, -0.3%, +0.7%, +0.2% Mean active return: (0.8 - 0.3 + 0.7 + 0.2) / 4 = 0.35%

Squared deviations from the mean: (0.8 - 0.35)^2 = 0.2025 (-0.3 - 0.35)^2 = 0.4225 (0.7 - 0.35)^2 = 0.1225 (0.2 - 0.35)^2 = 0.0225

Sum = 0.7700. Divide by (n-1) = 3: 0.2567. Square root: approximately 0.506% per quarter.

Annualized tracking error: 0.506% x sqrt(4) = approximately 1.01%. Annualized active return: 0.35% x 4 = 1.40%. Information ratio: 1.40 / 1.01 = 1.39. That is a strong information ratio, with the caveat that four quarters is a very short sample. A reliable tracking error estimate needs at least 36 monthly observations (three years) per the GIPS convention.

Common Mistakes

1. Using arithmetic active return on compounded returns. For multi-period results, tracking error is the standard deviation of period-by-period active returns. It is not computed from the compound return gap. Mixing the two leads to wrong annualization.

2. Mismatching benchmarks or frequencies. If the portfolio is monthly net-of-fee and the benchmark is daily gross, the active return series is polluted. Return type and frequency must match.

3. Ignoring liquidity and non-synchronous pricing. Portfolios with illiquid holdings (real estate, private credit, certain bonds) show artificially low tracking error because their prices are stale. Smoothed returns understate the true risk.

4. Treating tracking error as a skill metric. A high tracking error is neither good nor bad on its own. Without a corresponding active return, it is just risk. The information ratio is the right skill gauge.

5. Annualizing short samples. One year of monthly data gives a noisy tracking error. Reporting it as though it were stable is misleading. The GIPS standards require three years of monthly observations for a reason.

Frequently Asked Questions

Q: What is tracking error calculation in simple terms? It is the standard deviation of the difference between a portfolio's return and its benchmark's return in each period. A small tracking error means the portfolio moves closely with the benchmark. A large tracking error means the manager is taking significant bets away from the benchmark.

Q: How does tracking error affect investment decisions? It anchors the information ratio, the primary measure of active management efficiency. A manager delivering 2 percent annual active return with a 1 percent tracking error (IR of 2.0) is adding value more consistently than one delivering the same return with 8 percent tracking error (IR of 0.25).

Q: What is a real-world example of tracking error calculation? A US large-cap fund's quarterly active returns are +0.8%, -0.3%, +0.7%, and +0.2%. The standard deviation of those four values annualizes to about 1.01 percent. Combined with a 1.40 percent annualized active return, the information ratio is 1.39, strong for a large-cap mandate.

Q: How can investors use tracking error to evaluate whether a fund is truly active? Compare the tracking error to the management fee. A fund charging 80 basis points with a 0.2 percent tracking error is effectively an expensive index fund. Expect high-conviction active managers to show at least 3 to 5 percent tracking error to justify their fee.

Q: How is tracking error calculation different from standard deviation of returns? Standard deviation measures the total volatility of the portfolio itself. Tracking error measures only the volatility of the difference between the portfolio and its benchmark. A fund can have low standard deviation and high tracking error if it takes concentrated bets in low-volatility sectors, or high standard deviation and low tracking error if it simply leverages the index.

Sources

1. AnalystPrep. "Sources of Active Risk, Tracking Risk and Information Ratio (CFA Level 2)." https://analystprep.com/study-notes/cfa-level-2/explain-sources-of-active-risk-and-interpret-tracking-risk-and-the-information-ratio/
2. AnalystPrep. "Tracking Error (CFA Level 3)." https://analystprep.com/study-notes/cfa-level-iii/tracking-error/
3. CFA Institute. "Risk-Adjusted Performance Measures: A Case Study." https://rpc.cfainstitute.org/sites/default/files/-/media/documents/code/gips/case-study-risk-adjusted-performance-measures.pdf
4. CFA Institute. "GIPS Standards Tools and Resources." https://rpc.cfainstitute.org/gips-standards/tools-and-resources

Disclaimer

This article is educational content only and is not financial advice. Nothing here is a recommendation to buy, sell, or hold any security. Consult a licensed advisor before making investment decisions.