Alpha & Jensen's Alpha

Definition

Alpha is the portion of an investment's return that its risk exposure does *not* explain. It is the leftover — the value a fund's strategy added (or subtracted) after you account for what the market itself handed out. A positive alpha means the fund delivered more than its risk warranted; a negative alpha means you took the risk but did not get paid for it.

The word gets used in two related but different ways, and telling them apart matters:

• Simple alpha is just the fund's return minus its benchmark's return. If a fund earns 9% while the S&P 500 earns 10%, its simple alpha is −1%. This is easy to compute but crude, because it ignores how much market risk the fund actually took.
• Jensen's alpha — named for economist Michael Jensen — is the fund's return minus what the Capital Asset Pricing Model (CAPM) says a fund with that level of market risk *should* have returned. Instead of comparing the fund to the raw benchmark, it compares the fund to a risk-matched expectation built from the fund's beta.

Jensen's alpha is the version quoted on most fund fact sheets and analytics dashboards, and it is the more useful one. A fund that takes only 70% of the market's risk should not be expected to match 100% of the market's return — Jensen's alpha builds that fairness in, while simple alpha does not.

Throughout this article, all returns are illustrative round numbers, not actual fund results. And alpha should always be computed from total return — price change plus reinvested distributions — which matters doubly for income funds, where distributions are most of the story.

The Jensen's Alpha Formula

The formula looks dense on a fact sheet but unpacks into plain English:

Jensen's alpha = Rp − [Rf + β × (Rm − Rf)]

  Rp = the fund's or portfolio's total return over the period
  Rf = the risk-free rate (typically a short-term U.S. Treasury bill yield)
  β  = the fund's beta versus the benchmark
  Rm = the benchmark's ("market") return over the same period

Reading it term by term:

• Rp — the fund's return. Total return, including reinvested distributions. For a dividend or covered-call fund, using price return alone would throw away most of what the fund actually paid you.
• Rf — the risk-free rate. What you could have earned with essentially no risk, usually a short-term Treasury bill yield. It is the floor every risky investment must clear to be worth holding.
• β — beta. How sensitive the fund is to moves in the benchmark. A beta of 0.7 means the fund tends to capture about 70% of the market's swings, up and down. If beta is new to you, read the beta article first — Jensen's alpha is built directly on top of it.
• Rm − Rf — the market's excess return. How much the benchmark earned *above* the risk-free rate. This is the reward the market paid for bearing full market risk.
• The bracket, [Rf + β × (Rm − Rf)] — the CAPM expected return. Start at the risk-free floor, then add the fund's *share* of the market's excess return, scaled by its beta. A 0.7-beta fund is entitled to the risk-free rate plus 70% of the market's risk premium — no more, no less.

Alpha is simply the fund's actual return minus that bracket. Whatever is left over is return the fund's risk exposure cannot explain — the part credited to (or blamed on) the strategy itself.

Why It Matters

Income investors face a comparison problem that growth investors mostly do not. Dividend-growth ETFs and especially covered-call ETFs tend to run betas well below 1.0 — the option overlay and the tilt toward mature, cash-rich companies both dampen market sensitivity. That makes raw-return comparisons against the S&P 500 unfair in *both* directions:

• In a bull market, a 0.7-beta income fund will almost certainly trail the index. Headlines will call it a laggard. But trailing a benchmark while taking 30% less market risk is not necessarily underperformance — it may be exactly what the fund was built to do, done well.
• In a flat or falling market, the same fund will likely beat the index. Headlines will call it a winner. But beating the market while taking far less risk is the *expected* outcome for a low-beta fund in a downturn — it may reflect no skill at all, just the mechanical result of lower sensitivity.

Alpha is the metric that referees both cases. It asks the only question that is fair to a low-beta strategy: given the risk this fund actually took, did it deliver more or less than that risk explains? A covered-call fund with positive alpha added value on top of its risk profile. One with negative alpha destroyed value even if its headline return looked respectable.

Alpha also complements — not replaces — the Sharpe ratio. The Sharpe ratio measures return per unit of *total* volatility; alpha measures return above a *market-risk-based* expectation. A fund can look good on one and mediocre on the other, which is itself useful information about where its risk comes from.

Example

All numbers here are illustrative. Suppose a dividend-focused fund posts these results over a year:

• Fund return (Rp): 9%
• Risk-free rate (Rf): 4%
• Fund beta (β): 0.7
• Market return (Rm): 10%

Step through the formula:

1. Market excess return: Rm − Rf = 10% − 4% = 6%
2. The fund's share of it: β × 6% = 0.7 × 6% = 4.2%
3. CAPM expected return: Rf + 4.2% = 4% + 4.2% = 8.2%
4. Jensen's alpha: Rp − 8.2% = 9% − 8.2% = +0.8%

Notice the twist: this fund trailed the market by a full percentage point (9% versus 10%), so its simple alpha is −1%. Yet its Jensen's alpha is positive. A 0.7-beta fund was only "entitled" to 8.2% given the risk it took, and it delivered 9%. The strategy added value — the raw comparison just could not see it.

Now run the reverse case. A high-octane fund with a beta of 1.3 returns 11%, beating the market's 10%:

1. Expected return: 4% + 1.3 × (10% − 4%) = 4% + 7.8% = 11.8%
2. Jensen's alpha: 11% − 11.8% = −0.8%

This fund *beat the market* and still earned a negative alpha. Carrying 1.3 times the market's risk, it should have returned 11.8% in a year that strong — it showed up with less than its risk exposure alone would have produced. An investor comparing raw returns would congratulate it; alpha quietly points out you were not fully paid for the ride.

Put three illustrative funds side by side (Rf = 4%, Rm = 10%) and watch the rankings flip:

Takeaway: ranked by raw return, the aggressive fund wins and the dividend fund finishes last. Ranked by alpha, the order fully reverses. Same year, same numbers — the only difference is whether you account for risk.

A broad index fund like VOO sits near the middle by construction: it *is* the benchmark, so its beta is about 1.0 and its alpha hovers near zero (slightly negative after fees). Low-beta income strategies — a dividend-growth fund like SCHD or a covered-call fund like QQQI — are exactly the funds for which the raw-return column and the alpha column tell different stories, which is why alpha belongs in an income investor's toolkit.

What Alpha Cannot Tell You

Alpha is powerful but fragile. Three honest caveats before you lean on it:

• It depends entirely on the benchmark and the period. Measure a Nasdaq-100 covered-call fund against the S&P 500 and you get one alpha; against the Nasdaq-100 you get another. Neither is "wrong," but they are not comparable. Likewise, an alpha computed over three calm years can evaporate — or reverse — over five turbulent ones. Always check which benchmark and window produced the number.
• Low R² makes alpha nearly meaningless. Alpha borrows its logic from beta, and beta is only trustworthy when the benchmark actually explains the fund's movements. R-squared measures that explanatory power: with an R² near 0.9, alpha is a credible read on skill; near 0.3, the "expected return" in the formula is built on a relationship that barely exists, and the alpha on top of it is mostly noise. Read alpha and R² together, always.
• Fees eat alpha directly. Alpha is quoted after expenses, and management fees subtract from it one-for-one. A strategy that generates 0.6% of gross alpha but charges a 0.75% expense ratio hands its investors negative alpha every year. This is why persistent positive alpha is rare and why low fees are the most reliable "alpha" most investors will ever capture.

Common Mistakes

• Judging a low-beta income fund by raw returns. Trailing the S&P 500 in a rally is what a 0.7-beta fund is *supposed* to do. The fair question is whether it beat its risk-matched expectation — that is what Jensen's alpha measures.
• Crediting a bear-market win as skill. A low-beta fund beating the index in a selloff may be pure mechanics, not management. Check the alpha, not the headline.
• Confusing simple alpha with Jensen's alpha. "Fund minus benchmark" ignores risk entirely. The two can carry opposite signs for the same fund in the same year, as the example above shows.
• Quoting alpha without its benchmark, period, and R². An alpha figure with no context is close to meaningless; the same fund can show different alphas from different providers for these exact reasons.
• Treating past alpha as a forecast. Alpha is backward-looking, and positive alpha is notoriously hard to sustain. Academic evidence finds most funds' long-run alpha is near zero or negative after fees.
• Using price returns instead of total returns. For income funds, omitting distributions from Rp guts the numerator and manufactures fake negative alpha. Use total return, always.

FAQ

What is a good alpha?

Any positive alpha, sustained over a meaningful period and measured against an appropriate benchmark, is genuinely good — after fees, most funds deliver alpha near zero or slightly negative, so even +0.5% to +1% per year is a strong result if it persists. Be more skeptical of very large alphas (say, +5% or more): they usually signal a mismatched benchmark, a short or unusual measurement window, or a low R-squared that makes the number unreliable, rather than extraordinary skill.

What is the difference between alpha and beta?

Beta measures *sensitivity* — how much a fund tends to move when the market moves. Alpha measures *value added* — the return left over after beta-driven market exposure is accounted for. Beta describes how bumpy the ride is relative to the market; alpha judges whether the driver earned more than the road conditions explain. The two come from the same regression, which is why fact sheets quote them side by side.

Can alpha be negative if a fund beats the market?

Yes. A high-beta fund that beats the market can still post negative Jensen's alpha if it returned less than its risk warranted. In the example above, a 1.3-beta fund returning 11% against a 10% market carried an expected return of 11.8% — so it beat the index by 1% and still logged −0.8% alpha. The mirror image is equally true: a low-beta fund can trail the market and earn *positive* alpha.

Is simple alpha the same as Jensen's alpha?

No. Simple alpha is just fund return minus benchmark return, with no adjustment for risk. Jensen's alpha subtracts a CAPM-based expected return that scales with the fund's beta, so it compares the fund against a *risk-matched* bar rather than the raw index. For funds with betas near 1.0 the two are similar; for low-beta dividend and covered-call funds they can differ substantially — and even point in opposite directions.

Why do different websites show different alphas for the same fund?

Because alpha is only defined relative to a benchmark, a time window, a data frequency, and a risk-free rate — and providers choose these differently. One site may regress three years of monthly returns against the S&P 500 while another uses five years against a total-market index. Neither is wrong, but the outputs are not comparable. When two alphas disagree, check the benchmark and period first, and look at the R² to see how much faith either number deserves.

Does a high dividend yield mean high alpha?

No — the two are unrelated. Yield describes how a fund *distributes* its return, not whether that return exceeded a risk-adjusted expectation. A high-yield fund whose share price steadily erodes can post deeply negative alpha, while a modest-yield dividend-growth fund can generate positive alpha for years. Judge income funds on total return, alpha, and risk measures like the Sharpe ratio — never on the distribution rate alone.