Decoding alpha and beta: manager skill versus market risk

A fund returned 18% last year. Its benchmark returned 14%. Did the manager add four percentage points of value?

Almost certainly not — and separating the part that came from taking more market risk from the part that came from skill is exactly what alpha and beta exist to do.

Beta: how much market you are holding

Beta measures sensitivity to the benchmark. A beta of 1.0 means the fund has historically moved roughly one-for-one with the index. 1.2 means it tends to move 20% more in both directions. 0.8 means it dampens both.

Beta is not a quality measure. It is a dial for how much market exposure you have taken, and you can raise your return in a rising market simply by turning it up. That is not skill; it is leverage on the index, and it costs you symmetrically when the market falls.

This is the first thing to check when a fund beats its benchmark in a bull year. A beta of 1.3 in a 14% market roughly explains an 18% return on its own.

Alpha: what is left over

Alpha is the return above what the fund’s beta would predict, given how the market did. It is the residual after market exposure has been accounted for — the part attributable to the manager rather than the tide.

This site computes Jensen’s alpha, annualised, against the fund’s own benchmark using CAPM. A positive alpha means the fund did better than its risk exposure alone would justify; negative means worse.

Two disciplines when reading it:

• Against the correct benchmark. A mid-cap fund measured against a large-cap index will show flattering alpha that is really a cap-size effect. Check what it is being compared with.
• Over a long window. One year of alpha is noise. Three to five years, ideally confirmed with rolling returns, starts to be evidence.

R²: whether alpha and beta mean anything at all

This is the check almost everyone skips, and without it the other two numbers can be meaningless.

R² tells you how much of the fund’s movement the benchmark explains. It runs from 0 to 1 (often shown as a percentage).

• Above ~0.85 — the benchmark is a good description of the fund. Alpha and beta are trustworthy.
• Below ~0.70 — the fund is doing something the benchmark does not capture. Beta is describing a relationship that barely exists, and alpha computed from it is close to meaningless.

Always read R² first. A spectacular alpha with an R² of 0.5 is a statistical artifact, not a finding.

R² has a second use: an index fund should have R² essentially at 1.00 against its own index. Meaningfully below means it is not doing its one job.

The closet-indexing test

Put the three together and you get the most practically useful thing they do.

A fund with beta near 1.0, R² above 0.95, and alpha near zero is delivering the index — while charging you an active fee for it. That is a closet index fund, and it is the worst value proposition available: index performance at four or five times the cost.

The expense ratio then becomes a guaranteed annual loss against simply buying the index.

Pitfalls to avoid

• Reading alpha without R². The single most common misuse.
• Confusing high beta with skill. More market exposure is not more ability, and it will be repaid in the next drawdown.
• Comparing alpha across categories. Alpha is benchmark-relative. Two funds measured against different indices are not comparable.
• Treating one year as evidence. Alpha is noisy and mean-reverting.
• Assuming low beta means safe. A low-beta fund can still be concentrated, illiquid or exposed to a risk the benchmark does not contain — that is precisely what a low R² is telling you.

Key takeaway

Beta is how much market you took. Alpha is what you got beyond it. R² tells you whether either number is trustworthy. Read them in that order — R², then beta, then alpha — over multi-year windows and against the correct benchmark. Their most valuable use is not finding the next star manager; it is catching the fund charging active fees to deliver the index. Every formula and window used here is on the methodology page.