Alpha and active ETFs
An active ETF implements an actively managed strategy. Evaluate its stated objective first, then compare returns, risk exposure, and costs on a consistent basis.
Change one assumption
Does beating the benchmark mean positive alpha?
All returns share a horizon and convention; the comparison benchmark and CAPM market are separate inputs. Initially, benchmark excess is 3 pp and Jensen alpha is 0.
Benchmark excess return (pp)
CAPM-required return (%) and Jensen alpha (pp)
This illustrates model sensitivity, not a return promise or investment recommendation.
Start with the active strategy's objective
An index-tracking ETF primarily follows a specified index; an active ETF adjusts its portfolio through a strategy and managerial decisions. Active management may seek benchmark outperformance, income, lower volatility, or another objective. Not every active ETF has beating the market as its sole aim.
Before comparing performance, identify the investment universe, constraints, and an appropriate benchmark. Benchmarks from different regions or asset classes can make the same return look very different. Comparison horizons must also match.
Three percentage points ahead, yet zero alpha?
Suppose the ETF returns 12% and its comparison benchmark returns 9%. Their difference is a benchmark excess return of 3 percentage points. This describes relative performance before adjusting for the model-required compensation for higher market exposure.
With a 3% risk-free rate, 9% market return, and beta 1.5 over the same horizon, CAPM requires 12%, leaving Jensen alpha of 0. Changing only beta to 2 raises the required return to 15% and makes alpha −3 percentage points, even though the ETF still beats its comparison benchmark.
Tracking error and alpha answer different questions
Tracking error is the standard deviation of active returns—the ETF return minus its comparison benchmark return—over multiple periods. It measures variability in relative performance. A single 3-percentage-point gap is not tracking error, and high tracking error does not imply high alpha.
Historical Jensen alpha requires a time series of excess returns at a consistent frequency and attention to statistical uncertainty. An ordinary-return STAT regression intercept cannot simply be relabeled Jensen alpha; omitting the risk-free rate changes its meaning.
Compare costs, horizons, and evidence
Use dividend-inclusive total returns, a common currency, and the same horizon, then state whether returns are gross or net of fees. If published NAV returns already reflect fund expenses, do not deduct those expenses again. Disclose whether trading costs, taxes, and personal charges are included.
A single-period illustration explains arithmetic, not persistent performance. Short samples, survivorship bias, favorable benchmark selection, and omitted risk factors can flatter alpha. Even longer, comparable evidence cannot guarantee future results.
Show formal definition and formula
In a single-period illustration, Jensen alpha equals realized return minus the CAPM-required return. In historical regression it is the intercept of asset excess returns regressed on market excess returns. Both depend on benchmark, model, and horizon and do not directly identify pure managerial skill.
- Ri / Rm
- Same-period asset / market benchmark returns.
- Rf
- Same-period risk-free rate.
- β
- Asset sensitivity to the specified market systematic risk.
- α
- Return beyond the CAPM requirement, distinct from a simple ETF benchmark gap.
A common misconception
Beating a benchmark does not necessarily mean positive risk-adjusted alpha. Short-period positive alpha may reflect luck, omitted risk factors, or model choice; it neither proves selection skill nor guarantees future returns.
The active versus passive ETF case separates a fictional alpha calculation from fixed FBCG and IWF annual NAV returns, so a benchmark gap is not mistaken for proof of skill.
Frequently asked questions
Must the comparison benchmark equal the CAPM market?
No. A strategy's performance benchmark and a model's market proxy serve different purposes. Explain each choice rather than silently treating them as identical.
Are active ETFs always better?
There is no such general conclusion. Consider objectives, risk, costs, and comparable performance evidence rather than relying on the active/passive label or one period of positive alpha.