Risk tolerance: what a questionnaire really measures (and what it doesn't)
Purely educational content. It is not personalised advice, a recommendation or a suitability assessment. The Risk Lab mentioned here is a teaching exercise: it collects and stores no data and produces no recommendations.
Why you're asked to fill in a questionnaire
Anyone who has opened a brokerage account has met it: before you invest, banks and advisers must ask about goals, horizon, experience and how you react to losses. It isn't decorative bureaucracy: European investment-services rules (the directive known as MiFID II) require intermediaries to assess the suitability of what they propose for the person in front of them. The underlying idea is simple and serious: an investment can be excellent in the abstract and wrong for a specific person, if it forces them to sell at the worst moment.
The three dimensions (and the one questionnaires measure worst)
A sound profiling process looks at three different things:
- Risk tolerance (willingness): how much loss you can psychologically bear without abandoning the plan. A personal trait, not a balance-sheet item.
- Risk capacity (ability): how much loss your finances can absorb without compromising essential goals — it depends on income, reserves, debts, expenses.
- Time horizon: how long you can leave the capital untouched. It's what turns a temporary fall into a permanent problem, or the other way round.
The weak spot is the first one: tolerance is declared in calm markets, in front of a form, and lived in falling markets, in front of a red balance. The two often disagree.
The number linking declared and lived: drawdown
The most concrete measure of "how much it hurts" is the maximum drawdown (MDD): the distance between the portfolio's highest point and the following trough. It has two properties every investor should know. First, it is asymmetric: after −20% you need +25% to break even; after −50% you need +100%. Second, it has a duration: what matters is not only how far you fall, but how many years you stay below the previous peak. Other metrics are covered in how to measure an investment.
Declared vs lived: three dates to remember
- 2008: global equities lost roughly a third within the year; from the 2007 peak to the 2009 trough, more than half. Those who had declared "I can bear −15%" discovered what seeing double that feels like.
- March 2020: about −30% in a month, followed by a very fast recovery. Anyone looking only at year-end data never saw it — anyone invested certainly lived it.
- 2022: the year that surprised the cautious. Bonds fell too (about −17% for the euro aggregate): the traditional shelter didn't shelter, and defensive portfolios suffered as much as balanced ones, or more.
Try the logic from the inside: Risk Lab
To make all this tangible we built Risk Lab, a free educational exercise that replicates the logic of profiling: seven questions produce a score and a teaching profile-type (from Defensive to Aggressive, each with an illustrative equity/bond mix); the eighth question sets your tolerance threshold; the exercise then compares that threshold with 25 years of stylised history of the mix: maximum drawdown, years to reclaim the peak, final value of €10,000.
The interesting part is when the two numbers clash: if the mix's history exceeded the declared threshold, the exercise says so — with the figures — and shows, for comparison, the profile-type that stayed within the threshold, and its cost in foregone growth. With the most cautious thresholds (−5%, −15%) you'll discover a lesson worth the whole exercise: over those 25 years no illustrative mix, not even the most defensive, stayed within the threshold — mostly because of 2022. Absolute safety is not on the market's menu.
What Risk Lab is NOT (important)
- Not a suitability questionnaire: it does not consider your overall wealth, your real goals or your capacity to bear losses.
- No recommendations: profile-types and mixes are teaching categories on stylised series, not investment proposals.
- Nothing is stored: answers and results stay in your browser.
- Data are annual closes: real intra-year falls are deeper than shown (March 2020 doesn't appear at all) — the exercise says so explicitly.
Real decisions require a full assessment with an authorised provider: the real questionnaire, with all three dimensions.
Educational content, not a recommendation or advice. Past returns are not indicative of future ones. The series mentioned are stylised, gross of costs, taxes and inflation.
