Whether a plan is carried through to the end depends on what the user can bear. Risk tolerance has three faces: capacity, the loss a person can objectively bear without harming their livelihood and goals; willingness, the risk they are subjectively willing to take; and behaviour, what they actually do when a decline comes. The three often disagree, and the plan can only be set by the lowest of them. Capacity changes with age and income, willingness changes as markets rise and fall, and behaviour can be seen only after a decline has been lived through. Any measurement of risk tolerance is therefore a statistical judgment: it describes how investors similar in the past have usually behaved, and cannot assert how any one person will behave.
Two people fill in the same questionnaire, and both choose "willing to take relatively high risk". One is thirty, with a stable income and no debts, and will not need the money for twenty years. The other is fifty-eight; this money is their entire savings, and they retire in two years. For the first, willingness matches capacity; for the second, willingness far exceeds capacity. Even if two people had the same capacity and the same willingness, they might react quite differently to a sharp fall. Investment practice therefore divides risk tolerance into two aspects, capacity and willingness (Klement, 2015), and research in behavioural finance has added a third: actual behaviour.
A person's total wealth includes, besides the financial assets in their accounts, the present value of their future earnings: human capital. Young people have few financial assets but a great deal of human capital; with age, human capital is gradually converted into financial assets, until by retirement it is largely used up. Research by Bodie, Merton and Samuelson (1992) and by Viceira (2001) shows that if earnings are stable and only loosely related to the stock market, human capital behaves much like a bond, and a young person's financial assets can bear somewhat more equity risk; if earnings themselves rise and fall with the economy, for example when they depend on bonuses or sales commissions, human capital behaves more like a stock, and financial assets should be more conservative. How much risk a person can bear depends on the composition of their total wealth, not only on the figure in their account.
A common way to measure willingness to take risk is to have people choose between a series of certain payoffs and risky alternatives. Holt and Laury (2002) designed a set of paired lotteries: as the probability of the high payoff rises row by row, the row at which a person switches from the safe option to the risky one reflects their degree of risk aversion. Yet willingness measured by questionnaire is not stable. Guiso, Sapienza and Zingales (2018), surveying clients of an Italian bank, found that clients' risk aversion rose markedly after the 2008 financial crisis, even among clients who had suffered no losses. Willingness shifts with market moves and recent experience: people overestimate themselves after a rise and underestimate themselves after a fall.
Measuring willingness to take risk with a set of lotteries
In each row, choose between the safe option A and the risky option B. Going down the rows, B's chance of the high prize rises step by step. Tap the row at which you switch to B.
A plan may say "can bear a 30% decline", but when a decline actually comes, some people sell on the way down. Hoffmann, Post and Pennings (2013) tracked the perceptions and trading of a group of Dutch individual investors during the financial crisis: their return expectations and risk tolerance fell, and their risk perception rose, at the worst point of the crisis, and then gradually recovered; changes in trading and risk taking moved with these perceptions. Selling during a decline often means turning a paper loss into a realized one and missing the recovery that follows. For a product, behaviour is the hardest face to measure in advance, yet it is the one that ultimately decides what the user gets.
When capacity, willingness and behaviour disagree, the plan can only be set by the lowest of them. When willingness exceeds capacity, the allocation should not follow willingness even if the user wants it to, because a single decline could endanger their livelihood and goals; this is the starting point of suitability rules. When capacity exceeds willingness, one can explain the long-term benefits of taking more risk, but should not force it, because a plan beyond the user's willingness is likely to be abandoned in a decline. And the actual tolerance shown in the behavioural record is the most direct test of the other two.
When the three faces disagree, the lowest governs
Adjust capacity, willingness and behaviour separately to see the risk level at which the plan can be carried out and how the product should respond.
Combining questionnaire answers, assets and behavioural records into a risk grade or an investor profile makes communication easier and helps match products. But a profile is in essence a statistical description: it states how "investors who answered similarly have usually behaved in the past", and the differences within one category are often large. Klement's (2015) review notes that common risk questionnaires explain only a limited part of actual allocations to risky assets. A profile should therefore be phrased in terms of a statistical group, for example "among investors similar to this in the past, about one third sold during a 30% decline" (illustrative figure), rather than as a conclusion about the individual, for example "you can bear a 30% decline". The first offers a point of reference; the second makes the judgment on the user's behalf, and makes a promise on the institution's behalf that it cannot keep.
Risk tolerance should be understood through each of its three faces, acted on according to the lowest, and reviewed over time. When it is described, it should be described only as a statistical judgment.
Beyond the questionnaire, take in assets, income, horizon and actual trading records from past declines.
Ask about the stability of income and plans for using the money, and let age and the nature of income shape the recommended risk level.
When willingness exceeds capacity, explain why and set limits; reassess after a sharp rise or fall and as the goal approaches.
Say "how similar investors have usually behaved in the past", show the differences within the group, and do not assert how much any individual can bear.