Risk Tolerance Questionnaires
Risk Tolerance Questionnaires
A questionnaire asking "How would you react to a 30% loss?" predicts almost nothing. Asking "What did you do in 2008?" predicts everything.
Key takeaways
- Risk tolerance questionnaires measure aspirational tolerance, not actual tolerance revealed under stress.
- Typical questionnaires over-estimate tolerance by 15–30%; investors who score "aggressive" often want to reduce risk after 20% crashes.
- The best predictor of future behaviour is past behaviour; a questionnaire about your 2008 experience is more accurate than a hypothetical.
- Most standardised questionnaires cluster investors into 5–7 risk buckets, which over-simplify reality.
- A better approach: stress-test your willingness with specific, historical scenarios.
How standard questionnaires work
A typical risk tolerance questionnaire (used by most UK advisers) has 10–20 questions, scored on a points system. Common questions:
- If your £100,000 portfolio fell to £80,000 in three months, you would: (A) Sell everything to cut losses [0 points] (B) Sell half to reduce exposure [2 points] (C) Hold and wait for recovery [4 points] (D) Buy more to average down [6 points]
- Your investment experience is: (A) None [0 points] (B) 1–5 years [2 points] (C) 5–10 years [4 points] (D) 10+ years [6 points]
- Your income is: (A) Unstable or part-time [0 points] (B) Stable employed [2 points] (C) Stable, with bonuses [4 points] (D) Multiple income streams [6 points]
- You describe yourself as: (A) Very risk-averse [0 points] (B) Cautious [2 points] (C) Moderate [4 points] (D) Comfortable with risk [6 points]
Score 0–10: Conservative Score 11–20: Moderate Score 21–30: Aggressive
The portfolio is allocated accordingly: conservative = 20/80, moderate = 50/50, aggressive = 80/20.
Why these questionnaires fail to predict actual behaviour
The problem is fundamental: a questionnaire measures what you think you'd do, not what you'd actually do.
Research example 1: Vanguard surveyed 401(k) investors in 2007 (before the 2008 crash), asking them how they'd react to a 30% portfolio decline. 80% said they'd "stay the course or buy more." Actual behaviour in 2009: 15% of those same investors actually stayed the course or bought more; the other 65% reduced equity exposure or sold out.
Research example 2: A 2021 study by Morningstar tracked investor-stated risk tolerance against actual behaviour during the 2020 COVID crash and the 2022 bear market. Investors who scored "aggressive" on a questionnaire were 40% more likely to want to shift to "conservative" during downturns than their questionnaire score suggested.
Research example 3: Schwab analysed data from clients who had completed risk tolerance questionnaires. Those who scored in the "aggressive" bucket (80/20 or more) were, on average, overweighting bonds by 15% after a 20% market decline—despite the questionnaire predicting they'd hold or buy.
The gap between stated tolerance and actual behaviour is enormous.
Why the gap exists
Several psychological biases explain why questionnaires are unreliable:
Aspirational answering: When asked hypothetically, most people want to see themselves as rational and brave. A banker might answer "I'd buy more" (the smart financial move), but when the crash actually arrives and they see their net worth fall 30%, fear overrides aspiration.
Recency bias: Investors who completed a questionnaire in 2021 (after years of gain) or 2023 (after strong recovery) over-estimate tolerance. Investors who completed one in March 2020 or October 2022 (during crashes) under-estimate tolerance.
False confidence: First-time investors or those who've only known bull markets don't have experiential calibration. A 28-year-old who's invested for five years in a rising market might score "aggressive" because they've never seen a crash. When one arrives, reality hits.
Social desirability: Nobody wants to admit they're anxious or risk-averse. Self-reporting contains an inherent bias toward the "right" answer (brave, rational).
Failure to imagine: It's one thing to read "your portfolio falls 30%" on a questionnaire. It's another to see your £100,000 become £70,000, watch it daily, field calls from panicked friends, and read headlines about recession. The emotional reality vastly exceeds the hypothetical.
What questionnaires actually measure
Despite their flaws, questionnaires do measure something:
- Demographic risk capacity (age, income, time horizon) — actually fairly predictive.
- Self-perception of risk attitude — only weakly predictive of actual behaviour.
- Stated preferences — what you say you'd do, not what you'd actually do.
A questionnaire is useful for separating 25-year-olds from 70-year-olds. It's useless for distinguishing between two 50-year-olds, both employed, both with £500,000, one of whom will panic-sell in a crash and one who won't.
Better methods: behaviour-based assessment
A more reliable approach asks about actual past behaviour:
1. Historical crash questions:
In 2008, the FTSE 100 fell 48%. If you were invested, did you: (A) Sell, (B) Hold, (C) Buy, (D) Wasn't invested?
If they sold, willingness is low (actual behaviour, not hypothetical). If they held and later bought, willingness is high.
2. Recent volatility questions:
In September 2022, when the pound fell sharply and gilt yields spiked, what did you do: (A) Check your portfolio weekly, (B) Check daily or sold some holdings, (C) Didn't check, trusted the plan?
(A) shows moderate anxiety. (B) shows high anxiety. (C) shows strong conviction or detachment.
3. Temptation questions:
During a market downturn, do friends or media often make you question your allocation, or do you stick with your plan?
Susceptibility to "fire sirens" (media panic) predicts susceptibility to panic-selling.
4. Recovery questions:
After the 2020 crash (which recovered by August 2020), did you stay invested the whole time, or did you go to cash and re-enter later?
Those who re-entered later typically have lower willingness; they timed the market (unsuccessfully) rather than staying invested.
5. Spending behaviour under stress:
When other areas of your life are stressful (job uncertainty, relationship strain), are you more likely to check investments obsessively or less likely?
Stress-driven checking predicts stress-driven selling. Investors who check more during bad times tend to make bad decisions.
These questions are far more predictive than hypothetical scenario questions, because they're grounded in actual behaviour.
The questionnaire-to-allocation trap
Most UK advisers use a standardised questionnaire that sorts investors into, say, five risk buckets:
- Conservative: 20/80
- Balanced Conservative: 40/60
- Balanced: 50/50
- Balanced Aggressive: 70/30
- Aggressive: 80/20
The problem: this oversimplifies. An investor might need 60/40 (based on time horizon and spending needs), but the questionnaire only offers discrete buckets. The adviser rounds to 50/50 or 70/30, creating a mismatch between the recommended allocation and what the investor actually needs or can tolerate.
A better approach: use the questionnaire as input, not gospel. Calculate the allocation based on:
- Time horizon and ability (empirically derived, not questionnaire-based)
- Actual willingness (based on past behaviour, not hypothetical answers)
- Financial need (return required to meet goals)
Then, build a customised allocation that may fall between the standard buckets.
The honesty check
Before finalizing an allocation based on a questionnaire, ask yourself:
-
Would I be comfortable checking my portfolio during a bear market? If "no" or "only under duress," your tolerance is lower than your questionnaire suggests.
-
If I held this allocation and saw a 30% crash, would I want to talk to my adviser about reducing it? If "yes," your tolerance is lower.
-
Have I lived through a significant drawdown? If no, I should probably rate my tolerance conservatively (assume I'm lower than I think).
-
What would I actually do if the market fell 30% and didn't recover for two years? Not what you'd like to do, but what you'd actually do. That's your real tolerance.
The questionnaire flowchart
Related concepts
Next
Questionnaires are flawed, but they're not useless. The real test of risk tolerance is simpler and more direct: can you sleep at night? The next article explores the sleep-at-night test—a straightforward way to calibrate your willingness without questionnaires or hypotheticals.