Risk Tolerance and Risk Capacity Are Different Things
5 steps · one page
In short
Risk tolerance is how much variability a person is willing to live with. Risk capacity is how much they can absorb without something breaking.
Scope. MarketClue operates no risk questionnaire, assigns no risk profile, and does not tell any reader how much risk they should take. This article explains the distinction between two quantities that are routinely merged, and what the arithmetic of holding periods does and does not settle. All figures use the illustrative teaching parameters fixed on this pillar's hub and are not forecasts. Why a stated tolerance fails to survive a decline is behavioural and belongs to Pillar 32.
They are frequently treated as one quantity and they are not even the same kind of thing. Tolerance is a preference — it belongs to the person, it is not right or wrong, and nobody can tell someone else what theirs should be. Capacity is a set of facts — obligations, horizon, income stability, what else is available — and it can be wrong, in the sense that a person can believe they have more of it than they do.
Worked example
Why merging them is the error rather than a shortcut. The four possible combinations behave completely differently. High tolerance with high capacity and low with low are internally consistent. The other two are where the difficulty lives. High tolerance with low capacity is a person comfortable with variability they cannot actually absorb — the appetite is real, the ability is not, and nothing about the comfort changes what happens if the money is needed. Low tolerance with high capacity is a person who could absorb far more than they are willing to, which costs them nothing except foregone return and is nobody else's business. A single "risk profile" number cannot represent two independent dimensions, and any instrument that returns one has collapsed a preference and a set of facts into a figure that describes neither.
What the horizon arithmetic actually shows
The most-quoted claim about capacity is that a long horizon reduces risk. On the hub parameters — equities at a 9.0% expected return with 16.0% variability — here is what a longer horizon does.
| Holding period | Chance of ending below the starting sum | 5th-percentile outcome on $100,000 | Median outcome |
|---|---|---|---|
| 1 year | 28.7% | $82,611 | $109,000 |
| 3 years | 19.1% | $79,810 | $127,000 |
| 5 years | 13.4% | $81,953 | $147,000 |
| 10 years | 6.2% | $94,665 | $214,000 |
| 20 years | 1.6% | $144,792 | $454,000 |
| 30 years | 0.4% | $238,199 | $961,000 |
Worked example — the probability falls and the spread widens, and only one of those is usually mentioned. The chance of finishing below where you started drops from 28.7% at one year to 1.6% at twenty. That is the basis for the claim, and on these parameters it is correct. What the same table also shows is that the range of outcomes explodes: the median rises from $109,000 to $454,000 while the fifth percentile rises much more slowly, so the distance between a good outcome and a poor one grows enormously with time. And the poor outcome does not improve monotonically: the fifth-percentile result is worse at three years ($79,810) than at one ($82,611), because early variability has time to compound against you before the expected return catches up. A crucial caveat: this table depends heavily on the assumed 9.0% return against 16.0% variability. A lower assumed return moves every figure in the first column upward, and the claim that time reduces risk is only as good as the expected return being assumed. MarketClue does not forecast returns and therefore does not assert that any horizon makes anything safe. (Simulated on annual returns from the stated mean and standard deviation; the percentiles vary by a few hundred dollars from run to run and the pattern does not.)
Why this settles less than it appears to
The arithmetic above is about a sum left alone for a fixed period. Capacity is about whether it can be left alone.
A thirty-year horizon is only a thirty-year horizon if the money is genuinely not needed for thirty years. If circumstances can force a sale in year four — job loss, an obligation falling due, a change in circumstances — then the relevant column is the four-year one, and the intention to hold for thirty is not a fact about capacity. Capacity is determined by the shortest period over which the money might actually be required, not the longest period over which it might be held.
The second reason is that the probability of a poor outcome is not the same as the cost of one. A 1.6% chance of ending below the starting sum is a small probability attached to an unspecified consequence. For one person that outcome is a disappointment; for another it is the difference between a plan working and not working. Capacity is about the consequence, and no probability contains it.
What each one actually depends on
Capacity is assessable from facts: when the money is needed and how firmly; what other resources exist; how stable the income is; what obligations are fixed; and whether other people depend on the outcome. None of these is a matter of opinion, and all of them can change.
Tolerance is a preference and is harder to establish than it looks, because the reliable measurement is behavioural rather than stated. A tolerance expressed in calm conditions and a tolerance revealed during a decline are frequently different — which is a matter for Pillar 32 and is the reason this portal treats a questionnaire answered in the abstract as weak evidence about anything.
Worked example
What follows, without advising anyone. The binding constraint is the lower of the two. A person with high tolerance and low capacity is constrained by capacity, because willingness does not create the ability to wait. Capacity can be improved and tolerance largely cannot — an emergency reserve, a stable income, a longer genuine horizon all raise capacity, whereas a preference is close to fixed. And the two move independently: capacity changes with circumstances, and a person can find their capacity has fallen without noticing, since nothing announces it. MarketClue assigns nobody a risk profile and does not describe any reader as conservative, moderate or aggressive.
Frequently asked
8 questions
What is the difference between tolerance and capacity?
Tolerance is how much variability a person is willing to live with — a preference, which cannot be right or wrong. Capacity is how much they can absorb without something breaking — a set of facts, which can be misjudged.
Why does merging them cause problems?
Because the four combinations behave differently, and two of them are internally inconsistent. High tolerance with low capacity is an appetite for variability that cannot actually be absorbed; low tolerance with high capacity costs only foregone return. A single profile number cannot represent two independent dimensions.
Does a longer horizon reduce risk?
On the hub parameters the chance of finishing below the starting sum falls from 28.7% at one year to 1.6% at twenty. But the range of outcomes widens enormously over the same period, and the result depends heavily on the assumed 9.0% return — a lower assumption raises every probability of loss.
Does the poor outcome always improve with time?
No. The fifth-percentile result is worse at three years than at one, because early variability has time to compound against you before the expected return catches up.
What determines a real horizon?
The shortest period over which the money might actually be required, not the longest over which it might be held. If circumstances could force a sale in year four, the four-year figures are the relevant ones however long the intention.
Is a small probability of loss the same as low risk?
No. A probability says nothing about the consequence. The same outcome can be a disappointment for one person and the failure of a plan for another, and capacity is about the consequence.
Can either be improved?
Capacity can — an emergency reserve, stable income or a genuinely longer horizon all raise it. Tolerance is a preference and is close to fixed. So the binding constraint is usually capacity, and it can fall without anything announcing it.
Why does MarketClue not have a risk questionnaire?
Because capacity depends on obligations and circumstances the platform does not know, tolerance is a preference it has no standing to assess, and a stated tolerance measured in calm conditions is weak evidence about behaviour in a decline.
References
- Tobin (1958) — Liquidity Preference as Behavior Towards Risk, Review of Economic Studies 25(2) (preference and the risk-free split) —
- Samuelson (1969) — Lifetime Portfolio Selection by Dynamic Stochastic Programming, Review of Economics and Statistics 51(3) (the argument against "time diversification" as usually stated) —
- Investor.gov (SEC) — Risk Tolerance —
Educational and informational only — not investment advice, a recommendation, or an offer to buy or sell any security. Investing involves risk, including the possible loss of principal. Worked examples use fictional companies and figures.