Sharpe, Sortino, Beta and Alpha, without the jargon
Four numbers that get quoted constantly and explained rarely. Each asks a narrow question — and each is easy to misread as something grander.
Every fund page carries a row of risk ratios. They are useful and they are narrow, and most of the trouble comes from reading them as more than they are.
Volatility, first
Before any of the four: volatility is simply how much the returns scattered around their own average. High volatility means the ride was bumpy. It does not mean the destination was worse.
Everything below is built on it.
Sharpe ratio — was the bumpiness paid for?
Take the return, subtract what a risk-free investment would have paid, divide by volatility.
It asks: for every unit of scatter you sat through, how much extra return did you get? Higher means better-paid volatility.
The catch: Sharpe punishes upward surprises exactly as hard as downward ones. A fund that occasionally leaps is marked down for it, though nobody has ever complained about an unexpected gain.
The other catch, and this one matters more: a Sharpe ratio means nothing without knowing the window and the risk-free rate used. A three-year return divided by twenty years of volatility is a number, but not a meaningful one. Any site quoting Sharpe should tell you both. Ours does.
Sortino — the same idea, downside only
Sortino is Sharpe with the upside removed from the denominator. Only downward moves count as risk.
It answers the question most people actually meant: was I paid for the falls? For a fund that rises in jumps, Sortino is usually the fairer number of the two.
Beta — how much of the market's movement showed up
Beta compares a fund's moves against its benchmark. Beta 1.0 means it moved roughly with the index. 1.3 means it amplified — up more and down more. 0.7 means it dampened.
Beta is not quality. It is not good or bad. It tells you what to expect on the days the market moves hard, which is a different thing from whether the fund is any good.
And it is only as honest as the benchmark. Beta against the wrong index is arithmetic performed on an irrelevant comparison. A sector fund measured against a broad index, or a hybrid fund measured against a pure equity one, will produce a confident number that means nothing.
Alpha — what was not explained by the market
Alpha is the return left over after accounting for what the benchmark did and how much of its movement the fund took on. Positive alpha means the fund did better than its market exposure alone explains.
It is the one people want most, and the one to treat most carefully:
- It inherits every problem beta has with the wrong benchmark, and adds its own.
- Over a short window it is statistically indistinguishable from luck. An alpha of −3% over 57 months usually cannot be told apart from zero. The number is printed without an error bar, and the error bar is often wider than the number.
- It is backward-looking. It describes what happened, not a skill that must continue.
How to use the four together
They are best read as a description of the ride, not a score. Something like: this fund moved more than its index, most of that movement was upward, and what you were paid for the falls was reasonable.
What they cannot do is rank funds, and we would be wary of anyone who uses them that way. Turning four numbers into one score requires weights, and weights are an opinion about what matters — which is exactly the judgement nobody can make on your behalf.
On a fund page here, each ratio is shown with the window it was computed over and the benchmark it was measured against, because without those two facts the ratio is decoration.
Analysis, not advice. These are descriptions of the past, not predictions, and no single number identifies a good investment.