What is a benchmark actually for?

A return on its own is a number, not a verdict. "+14% last year" tells you what happened to the money and nothing about whether that was impressive, ordinary, or disappointing. The missing half of the sentence is always the same: compared to what?

A benchmark is the answer. It is the alternative that was genuinely available for the same money, over the same period, at roughly the same kind of risk. Its whole job is to separate two things that look identical from the inside:

• the market went up, and you were in it;
• the portfolio did something the market didn't.

Without a benchmark those are indistinguishable. With one, they separate cleanly.

The arithmetic is trivial; the choice is not

Active return = portfolio return − benchmark return

That's it. A portfolio returns +14%, its benchmark returns +18%, so active return is −4 percentage points. The portfolio made money and still fell short of what was freely available for the same exposure. Flip it: +14% against a benchmark that did +5% is +9 points of active return, a genuinely different year with the exact same headline.

Note the units. Returns are in percent; the difference between two returns is in percentage points. Saying a fund "beat the index by 4%" when it beat it by 4 points is sloppy, and sloppiness in this direction is rarely accidental.

Three properties that make a benchmark usable

1. Investable. You could actually have held it, cheaply, for the whole period. A hypothetical basket nobody can buy is a debating trick, not a benchmark.
2. Specified in advance. Chosen before the period, not selected afterwards from a menu of indices once the results are known. A benchmark picked in hindsight measures nothing.
3. Representative. It holds the same kind of thing the portfolio holds — same asset class, same region, same currency, same broad risk. Comparing a small-cap equity fund to a cash rate is not a comparison, it's a costume.

Every professional performance report in the world rests on these three. Most misleading performance claims fail at least one of them, and the next two lessons are about exactly how.

A benchmark is a measuring stick, not a target

One clarification worth making early. Comparing to a benchmark tells you where a result sat relative to the available alternative. It does not tell you the portfolio was well built, that the manager has skill, or that anything should be bought or sold. It is a measurement, and the rest of this course is about how much a measurement can honestly carry.

In the data

An index and a fund tracking it are quoted side by side, in the same way:

The first number is a level in points; the second is a price in dollars that someone paid for a unit. Nothing in how they are shown tells you which could have been held, so investability is a fact about an instrument, established separately, and never something the existence of a series demonstrates.

Try it now

1. Two lines over the same year are below: a broad price index first, then an investable fund tracking it. Measure each from the first bar to the last and write the two percentages down.

2. Subtract to get the difference in percentage points. Write one neutral sentence describing the gap — not whether it was "good", just what it was, and where roughly a percentage point a year of it comes from.
3. Now the part that matters more than the number: one of those two series could have been held and one could not. Nothing on either chart says which. An index is a measurement; a fund is a thing you can own, minus fees, and the difference between them is the whole reason benchmarks are argued about.