Benchmark: what to measure your result against, honestly

Your result should be compared with the portfolio you could have held with no effort at all — and compared on a total return basis, over the same period, in the same currency, with the same deposits and withdrawals, net of costs. If your result is above that yardstick, the effort has paid off. If it is below or roughly equal, you have paid in time and nerves for something that is sold ready-made. Anything else in a comparison is curve-fitting.

What a benchmark actually does

A benchmark is not "the average market result" and not "how much the professionals make". It is the alternative to your decisions: a portfolio available to you without picking securities and without picking entry points. That alternative is usually the total return index of your own market or an exchange-traded fund tracking it — you can see what is available in the funds section.

The point of the construct is that it answers the question "was it worth it?". Without it, any result looks meaningful: a gain is pleasing, a loss is blamed on the market. A benchmark takes away both excuses at once.

A comparable period

The mistake that spoils a comparison more often than any other is mismatched time spans. Investors count their own result from the moment things started going well, and take the index from the start of the year. Or they compare a portfolio assembled in the spring with the index for the full calendar year.

The rule is simple: the start date and the end date must be identical for both figures, and the span must be chosen before you know the result. The same logic applies to company reporting — this is covered in detail in the piece on comparable periods.

A comparable composition

A portfolio of equities cannot be measured against a bond index, and vice versa. If your portfolio is mixed, the benchmark must be mixed too — built from an equity index and a debt index in the same proportions in which you hold stocks and bonds.

Idle cash is a separate case. It is part of your result, so it must be part of the yardstick as well: otherwise you are comparing the invested part of your portfolio with a fully invested index, and winning in downturns out of thin air.

Total return, not just price

A price index ignores dividends, and comparing against it systematically flatters a dividend portfolio. The yardstick should be a total return index — the kind that adds the payouts back in.

Count your own side the same way: payouts received are part of the result regardless of whether you reinvested them or withdrew them. The nearest payouts on the market:

And the full list is in the dividends section.

For bonds, the role of total return is played by yield to maturity, not by the coupon and not by the price — why that measure in particular is explained in the piece on yield to maturity.

Deposits and withdrawals

If you have been adding money, the simple difference between "what it was" and "what it is now" is no longer a return, and it cannot be compared with an index. A deposit ahead of a rally inflates the result, a deposit ahead of a decline understates it — even though you made no decisions at all.

Comparison with a benchmark calls for a time-weighted return: it measures the quality of the decisions, not how lucky the dates of your deposits were. A money-weighted return answers a different question — what your wallet actually received. The mechanics of calculating both are in the piece on how to calculate portfolio return honestly.

Costs and taxes

An index does not pay brokerage commission, the spread or custody fees, and you do. That is why your own result is taken after all costs — otherwise you are comparing clean theory with messy practice, in your own favour.

Taxes are better kept on a separate line: they depend on the account regime and the holding period, not on the quality of the decisions. The tax base is determined by the provision 13%.

The result is not the whole story

A benchmark can be beaten simply by taking more risk: concentration in a few securities, leverage, a bet on one sector. In a good period that produces outperformance; in a bad period, a fall deeper than the market's.

That is why the result is reported alongside the depth of the drawdown and how strongly the portfolio fluctuated. A portfolio that beat the yardstick with a noticeably deeper drawdown beat it not through skill but through a willingness to endure.

And a separate point: the valuation multiples of an individual security — say, no data for SBER — have nothing to do with comparing a portfolio against a benchmark. They are a tool for choosing a security, not for assessing a result; the sequence of such checks is in the piece on how to compare two companies.

The checking procedure

Step 1 — fix the span: a start date, an end date, and no shifting after the fact.

Step 2 — build the yardstick to match your composition: the shares of equities, debt and cash as in your portfolio, and the indices on a total return basis.

Step 3 — calculate your own time-weighted return, after costs.

Step 4 — put the drawdown of your portfolio next to the drawdown of the yardstick over the same span.

Step 5 — repeat on the same basis. A short span says almost nothing about the quality of the decisions: over it, chance outweighs skill.

What a benchmark will not show

It will not tell you whether your outperformance will be repeated, and it will not separate skill from luck — that takes a long history of comparisons on an unchanged basis. It does not take your circumstances into account: your horizon, your need for cash, your tolerance for drawdown. And it says nothing about the future — neither yours nor its own.

A model portfolio is a convenient way to check the calculations without risking money: how it works is described in the piece on the demo account.