Risk tolerance and ambiguity tolerance are not the same thing

Consider two choices. In the first, an investor knows that a portfolio has historically experienced losses of a given range. In the second, the investor cannot work out what a new investment does, how reliable its impact data are, or which assumptions sit behind the forecast.

Both involve uncertainty. The first is closer to risk: possible outcomes can be described with at least some probability or distribution. The second contains ambiguity: the probabilities, model or relevant information are themselves unclear.

What risk tolerance tries to capture

Risk tolerance generally describes willingness to accept uncertain financial outcomes, especially potential loss, in pursuit of a return. In a regulated advice context it sits alongside separate considerations such as capacity for loss, objectives, time horizon, knowledge and experience.

Survey measures of risk willingness can be informative. Dohmen and colleagues combined a large representative survey with an incentivised field experiment and found that a general willingness-to-take-risks question was a useful all-round predictor of several risky behaviours. That is evidence that a well-designed self-report item can carry real information.

It is not evidence that one item captures every form of uncertainty, or that willingness is fixed across contexts.

What changes when probabilities are unclear

Ambiguity tolerance concerns comfort when relevant outcomes or probabilities cannot be confidently specified. For an investor, ambiguity can come from an unfamiliar product, conflicting information, an uncertain future, opaque terminology or doubt about whether a stated sustainability outcome is meaningful.

Dimmock, Kouwenberg, Mitchell and Peijnenburg measured ambiguity aversion using specially designed questions in a representative US household survey. They found associations with stock-market non-participation, lower allocations to equities, foreign-stock ownership, own-company stock and under-diversification. During the financial crisis, ambiguity-averse respondents were more likely to sell stocks.

Why the distinction matters in product design

If a firm interprets every hesitation as low risk tolerance, its response will often be wrong.

More risk disclosure may not solve an ambiguity problem

A longer disclosure can add information while making the decision environment harder to parse. If the investor cannot identify which facts matter, volume increases ambiguity rather than reducing it.

A lower-risk product may not solve a trust problem

An investor may accept market volatility yet distrust the provider’s claims or the quality of impact reporting. Product substitution does not address that uncertainty.

Confidence can mask ambiguity

A person who reports high comfort with risk may still choose the most familiar option when probabilities are unclear. Conversely, someone cautious about loss may make sound decisions when the evidence and decision process are transparent.

How to assess the concepts separately

Useful research gives each construct its own job:

1. Ask about willingness to accept financial loss or variability without mixing the question with ability to absorb that loss.
2. Present decisions with different kinds of uncertainty, including situations where probabilities or relevant information are incomplete.
3. Measure knowledge and confidence separately, because uncertainty caused by unfamiliarity may look like a risk preference.
4. Keep context. A response to a hypothetical scenario is evidence about that scenario, not a complete prediction of a future trade.

u impact uses ambiguity tolerance as one axis in its nine-profile model. Standardised scenarios contribute a stronger signal than an individual self-report item, while objective knowledge and other answers remain part of the profile. The rule is deterministic and reviewed as new projects add aggregate evidence.

Match the response to the uncertainty

Different patterns suggest different tests:

• Known risk feels unacceptable: clarify loss, horizon and capacity; do not persuade somebody past an appropriate boundary.
• The situation feels unknowable: show assumptions, comparable examples, ranges and what remains uncertain.
• Too many options create uncertainty: test a shorter, explained choice set rather than adding more filters.
• Trust is the missing input: make provenance, fees, conflicts, limitations and evidence visible.

This does not mean designing a different product for every profile. Often the most effective change is to make the same decision environment more legible.