How to benchmark your portfolio (beating the S&P 500 is the wrong question)

To benchmark a portfolio properly, compare it against a yardstick that matches its own asset mix — a blended benchmark — and measure the return in a way that neutralises the timing of your deposits and withdrawals. A single equity index like the S&P 500 is the wrong yardstick for most private portfolios, because it answers a question you did not ask: how did 100% US large-cap equity do, rather than how did my mix of stocks, bonds and cash do. This article explains why the popular question “did I beat the market?” misleads, how to build a benchmark that fits your allocation, why a well-timed deposit can make a naive comparison look far better or worse than reality, and a four-step protocol you can run from a spreadsheet.

Why “did I beat the S&P?” is the wrong question

The S&P 500 is a specific bet: roughly five hundred large US companies, all equity, all one currency, no bonds and no cash. If your portfolio holds international stocks, bonds, a cash buffer, or anything outside US large-cap, then measuring it against that index compares two different things. In a year when equities rise sharply, a diversified portfolio with a bond allocation will almost mechanically trail an all-equity index — not because the strategy went wrong, but because it was never trying to be 100% equity in the first place. In a year when equities fall, the same diversification runs the other way.

The trap is that the comparison feels rigorous while being meaningless. A gap against the S&P 500 tells you almost nothing about whether your portfolio did what it was built to do, because most of that gap is explained by a single fact: you were not fully invested in US large-cap equity, and you never intended to be. A benchmark is only informative when it represents the alternative you would actually accept — the same risk, the same asset classes, run passively. Anything else measures your choice of asset allocation, which you already made on purpose, rather than the quality of what you did within it.

Choosing a benchmark that matches your allocation

The fix is a blended benchmark: a weighted combination of broad indices that mirrors your own asset mix. If you run a 70/30 split between stocks and bonds, the fair yardstick is not the S&P 500 but 70% of a broad equity index plus 30% of a broad bond index — for example a global developed-market equity index paired with a global aggregate bond index. Now the comparison is honest: both you and the benchmark carry the same broad exposure, so any difference between them is about your decisions within each asset class, not about the allocation itself.

Two details make the blend fair rather than convenient. First, the weights should be your target allocation, not whatever the market happened to drift to — otherwise you are grading yourself against your own drift. Second, the blend should be rebalanced on the same cadence you rebalance, because a fixed 70/30 benchmark that is never rebalanced slowly becomes something else as the equity sleeve grows. Match the mix, match the rebalancing, and the benchmark becomes the passive version of your own plan — the thing you could have bought instead of doing any work at all.

Computing the blend's return is just a weighted average of its components' returns. With clearly hypothetical inputs, the arithmetic is:

That is the whole formula: multiply each component's return by its weight and add them up. The numbers above are placeholders to show the method — the real inputs are whatever the chosen indices actually returned over your measurement window, which you would take from the index providers themselves rather than from an article.

Time-weighting: why deposits distort a naive comparison

Even with the right benchmark, a second problem hides in how you measure your own return. An index assumes a single lump of money invested for the whole period. Your account does not work that way — you add money, sometimes withdraw it, and the timing of those flows changes the raw euro result without changing the quality of the strategy at all. Compare a flow-affected number against an index and you are, again, comparing two different things.

A clean illustration. Suppose an account starts the year at €10,000. In the first half nothing much happens — call it a 0% half. Then, right before a rally, you deposit another €10,000, taking the balance to €20,000. In the second half the market rises +10%, so the account finishes at €22,000. Your total profit is €2,000. A back-of-the-envelope “profit over what I started with” reads €2,000 ÷ €10,000 = +20% — which looks like a strong year against any single-digit benchmark.

But the strategy did not earn 20%. The time-weighted return chains the sub-period returns and ignores the deposit's timing: (1 + 0%) × (1 + 10%) − 1 = +10%. The extra euros of profit came from putting more money in right before the rise, not from the strategy doing better. A money-weighted return (an internal rate of return) would land somewhere between the two — it credits the well-timed deposit — but it answers a different question again: “how did my euros do, timing included?” rather than “how did my strategy do?” Only the time-weighted number is comparable to an index, because only it strips out the cash-flow timing that the index never had. (All figures here are illustrative.)

This is the same distinction the German companion piece works through in a DACH context — the difference between a time-weighted return (TWROR) and a money-weighted one (IRR) — when it explains how to compare a portfolio against the MSCI World. The practical takeaway is identical in any language: to compare against a benchmark, use a time-weighted return, and let your tool do the sub-period chaining for you.

Risk-adjustment in plain terms

Two portfolios can arrive at the same return by very different roads. Imagine two accounts that both finish a period up the same amount — say the same +6% — but one got there in a near-straight line while the other plunged to a −25% drawdown mid-way before recovering. Same destination, wildly different journey. A return-only comparison treats them as identical; anyone who actually lived through the second one knows they were not. This is why a raw gap in percentage points is only half the story.

Risk-adjustment puts the two on comparable footing by asking how much return you earned per unit of risk taken. In plain terms: reward divided by turbulence. Measures like the Sharpe ratio (return above cash, divided by volatility) or a simple look at the worst drawdown formalise the idea that a smoother +6% is a different result from a white-knuckle +6%. You do not need the formula to use the concept — when you read a gap against your blended benchmark, check the two paths' drawdowns side by side before you conclude anything. A portfolio that trailed its benchmark by a couple of percentage points while taking far less risk has not necessarily done worse; it has done differently, and the difference is the point.

A four-step benchmarking protocol

Putting it together, an honest benchmarking pass needs no special software — just your position data, a spreadsheet, and the public index returns. Four steps:

1. Write down your target allocation. The asset-class weights your plan aims for — for instance 70% equities, 30% bonds — not the drifted weights of the moment. This is what the benchmark has to mirror.
2. Build the matching blend. Pick one broad, low-cost index per asset class and combine them at your target weights. Compute the blend's return as the weighted average shown above, rebalanced on the same cadence you rebalance.
3. Measure your own return, time-weighted. Chain your sub-period returns so deposits and withdrawals do not distort the figure. Almost every portfolio tool can report a time-weighted return; if you do it by hand, break the period at each cash flow and multiply the sub-period factors.
4. Read the gap — then read the risk. Subtract the benchmark's time-weighted return from your own to get the gap in percentage points, and put the two drawdown paths next to each other. The gap is a fact to explain, not a grade.

What a performance gap does — and does not — tell you

Run the protocol and you get a single, honest number: a gap of some percentage points between your portfolio and its matched blend. That gap is genuinely useful — it isolates what your decisions within your allocation added or subtracted, once the allocation itself and the cash-flow timing are held constant. A persistent gap in either direction is worth understanding: it might trace to fees, to a tilt toward or away from certain sectors, to cash sitting idle, or to trading that helped or hurt.

What the gap does not do is deliver a verdict. One period is a small sample, and a few percentage points over a single year can be noise rather than skill. A gap also says nothing on its own about risk — a portfolio that ran a couple of points ahead of its benchmark by taking on much larger drawdowns has bought that lead with turbulence you may not want to repeat. And a gap is silent about whether the benchmark itself was fair: get the blend wrong and you will “explain” a gap that only exists because the yardstick was mismatched. The number is a starting point for questions, never the end of them.

Automating the comparison

None of this requires a bank connection — it requires position data and index returns, both of which you can export and download. The manual version is a spreadsheet: your holdings and their time-weighted return in one place, the blended benchmark computed as a weighted average in another, and the gap in a single cell. The value of automating it is not sophistication but consistency — the same blend, the same time-weighting, recomputed every period without re-deriving it by hand.

Acutic's performance page already tracks how the platform's published scores have behaved over time, in the open, and the product reads a portfolio imported by CSV — no account linking — and reports its composition. An integrated view that measures an imported portfolio against a blended benchmark of your own target weights is on the roadmap rather than something the product ships today; the method above is the one it would automate, and the one you can run yourself in a spreadsheet in the meantime. Whatever tool computes it, the output stays a fact — a time-weighted gap in percentage points against a matched blend — and what to do about that fact remains the reader's decision, not the tool's.

If your question runs the other way — not how the whole portfolio compares to a benchmark but how heavily it leans on a single name — the companion piece on portfolio concentration risk measures that side of the same portfolio.

Further reading: the Acutic performance page shows how the published scores have behaved over time, and the product page shows the CSV import and portfolio view in detail. Create free account to import a portfolio and run the numbers yourself.

Acutic provides investment research and educational analysis under MAR Art. 20 / § 85 WpHG. Acutic does not provide investment advice (Anlageberatung per § 1 Abs. 1a S. 2 Nr. 1a KWG / Art. 4(1)(4) MiFID II), portfolio management, or any other licensed investment service. No content in this article constitutes a personal recommendation.