Chapter 9: Risk: capacity, tolerance, and need
$346,500. That is what $500,000 becomes when a 70/30 portfolio meets the worst year in the record since 1926, a fall of 30.7%. Whether that number reads to you as survivable or sickening is one of the most personal facts in this guide, and no questionnaire score can stand in for finding out. The industry compresses the whole subject into one question, "what is your risk tolerance," and the compression is the problem, because three different questions hide inside it. What can your life absorb. What can your sleep absorb. What does your goal require. They have different answers, different tools measure them, and your mix from chapter 2 has to pass all three.
This chapter gives each question a name and a number. Risk capacity is what your life can absorb: how many years until the money is needed, how steady the paycheck behind it is, who else depends on it. Risk tolerance is what your sleep can absorb: the loss you can watch without grabbing the wheel. Risk need is what your goal requires: the return that gets you from here to the number with the savings you can actually make. Capacity is arithmetic, tolerance is a rehearsal, and need is a comparison against honest forecasts.
Capacity: what your life can absorb
Capacity comes off your chapter 1 lines almost mechanically. Money with a date more than ten years out has high capacity, because every bad year in the Vanguard record had a recovery inside a decade-plus horizon's patience, even the worst ones. Money due inside five years has low capacity no matter how brave its owner feels, which is why chapter 6 parked it. Between the dates sit the modifiers: a stable paycheck raises capacity, because new contributions keep arriving through a crash and buy at the lows; an income tied to the same economy as your portfolio lowers it; dependents and fixed obligations lower it again.
Notice that courage appears nowhere in that paragraph. Capacity is the gate that does not care how you feel. Elaine's $700,000 funds next month's groceries, so its capacity is capped by the calendar regardless of her calm, while a 24-year-old's retirement account has decades of capacity even on the days its owner has none.
Need: what your goal requires
Need is the required return, and it gets measured against forecasts that must be labeled as what they are. Vanguard's VCMM run of March 31, 2026 puts US stocks at 4.9–6.9% annualized over ten years, international stocks at 5.0–7.0%, US bonds at 4.2–5.2%, and cash at 2.9–3.9%. BlackRock's mid-2026 capital market assumptions put US stocks near 8.5% at the center, a full 3 points above Vanguard's range. Both houses are smart, both publish, and the disagreement is the lesson: these are model assumptions, so plans use ranges and never single numbers.
A worked example shows the gate in action. Suppose a goal needs $500,000 in 15 years, starting from $150,000 with $500 a month in savings. Reaching it requires about 6% a year. A 50/50 mix pencils out near 5.3% a year at the midpoints of Vanguard's ranges, model assumptions again, which leaves the plan short. That gap has exactly three honest fixes: save more (about $800 a month closes it at the same mix), move the date, or hold more stocks, and the third fix is only available if the bigger mix also passes the other two gates. The dishonest fix is the popular one: keep the plan and assume the higher forecast.
Use the ranges the way engineers use them: test the plan against the bottom. A plan that works at 4.9% is sturdy, a plan that needs 6.9% is exposed, and a plan that only works at BlackRock's 8.5% central case is a hope wearing arithmetic. The test also runs in reverse. When the bottom of the range already covers your need, you hold surplus, and chapter 2's table says the calm way to spend surplus is a smaller worst year, not a reach for more.
Tolerance: the rehearsal
Tolerance is measured in dollars, never in percentages, because crashes arrive in dollars. The rehearsal below starts with $500,000 invested at 70/30 and replays three real storms against it. The worst year in the record since 1926, a fall of 30.7%, takes it to $346,500. A year shaped like 2008, at 24.3% down, leaves $378,360. A year shaped like 2022, at 16.6% down, leaves $417,100.
Set the widget to your own balance and your chapter 2 mix, then read the three endings slowly and notice which one crosses your "I would sell" line. That line is the binding constraint of your whole plan. Every discipline this guide teaches, the rebalancing of chapter 10, the glide of chapter 11, assumes you stay invested through the storm, so a mix that breaches your sell line in rehearsal will breach it in life, when the headlines are screaming and the rehearsal's calm is gone. The honest output of this section is one dollar figure: the worst single-year loss you can watch without grabbing the wheel.
Quinn ran the rehearsal's $500,000 default first and felt nothing, the way you feel nothing about a stranger's x-ray. Then she did the arithmetic on her own money. Her 2065 fund and her small brokerage account sit near 90/10, a mix whose worst year in the Vanguard record reads about −39% (1926–2024). Against her current $4,000, that is a $1,560 dent, a stretch of overtime shifts, genuinely fine. Then she multiplied forward. At $400,000, somewhere in her fifties, the same percentage is $156,000, several years of her contributions gone in one statement. The same mix, the same investor, and two completely different feelings. Her note afterward: capacity will grow with every paycheck, but tolerance has to be re-tested as the balance grows, so she wrote her worst-year line in dollars and a rule to re-run the rehearsal every time her balance doubles.
What happens when the three disagree?
The lowest gate wins, and which gate is lowest decides what you are allowed to fix. When need exceeds tolerance, the example above short of its $500,000, the temptation is to retake the questionnaire until it blesses a bigger mix. Change the goal or the savings rate instead, never the questionnaire, because re-answering a quiz changes nothing about how you will behave at 2 a.m. in the next 2022. When tolerance exceeds capacity, a brave investor with money due in three years, capacity wins, because the calendar does not negotiate. And when capacity exceeds tolerance, a long horizon attached to a nervous stomach, respect the stomach: take the smaller mix with the higher savings rate, since chapter 6's behavior gap (investors earned 7.0% in funds returning 8.2% over the ten years to 2024) is the documented price of holding a mix you cannot actually hold.
The triad also explains a pattern you will meet in the wild: the questionnaire that asks fourteen versions of "how would you feel" and zero versions of "when is the money due" or "what return does the goal need." A score built from feelings measures one gate of three. Bring the other two yourself, the dates from chapter 1 and the required return from this chapter, and staple them to whatever score the quiz prints before letting it near your mix.
Set your mix to pass all three gates and let the lowest one bind. When need is the gate that fails, change the goal or the savings rate, never the questionnaire, and write your worst-year line in dollars, because the storm will arrive in dollars.
Where people go wrong
Retaking the quiz until it agrees. A questionnaire that gets re-answered after every market move is a mood ring. The rehearsal's dollar figure is the durable version, and it only changes when your life does.
Testing tolerance in percentages. "Down 30%" passes inspection on every sunny day. $153,500 gone from $500,000 is the same fact wearing its real clothes, which is why the rehearsal and your plan line both speak in dollars.
Letting capacity impersonate tolerance. A 30-year horizon makes a 90/10 mix defensible on paper and does nothing for the investor who sells it in the first crash. Capacity says what you may hold. Only tolerance says what you will.
Testing once, at one balance. Quinn's discovery generalizes: a mix chosen at $4,000 gets inherited by $400,000 unless someone re-tests it. Rehearse again at every doubling, every big life change, and every time the worst-year dollars stop feeling theoretical.
Line nine is the shortest and hardest one: "My worst-year dollar line: $____." Quinn's reads: "My worst-year dollar line: $1,600 today, re-rehearsed every time the balance doubles." If your chapter 2 mix would breach the line, the mix moves, or the goal does.
Key takeaways
- "Risk tolerance" hides three questions: capacity (what your life can absorb), tolerance (what your sleep can absorb), and need (what your goal requires). Different tools measure each, and your mix must pass all three.
- Capacity is arithmetic from chapter 1's dates and your income, and it does not care how brave you feel; the calendar caps courage.
- Need is required return measured against labeled model assumptions: Vanguard's March 2026 VCMM puts US stocks at 4.9–6.9% over ten years while BlackRock's mid-2026 assumptions sit near 8.5% central, and the 3-point disagreement is why plans use ranges.
- Tolerance is rehearsed in dollars: $500,000 at 70/30 becomes $346,500 in the worst year since 1926 (a 30.7% fall), $378,360 in a 2008-shaped year (24.3% down), and $417,100 in a 2022-shaped year (16.6% down).
- When need exceeds tolerance, fix the goal or the savings rate, never the questionnaire; a quiz re-answered until it agrees changes nothing about 2 a.m. in the next crash.
- Tolerance is balance-dependent: a loss that feels fine at $4,000 can be unbearable at $400,000, so re-run the rehearsal as the balance grows, the way Quinn now does at every doubling.
Sources: Vanguard return forecasts (VCMM) · Vanguard model allocation data · Morningstar Mind the Gap · Finvest Cash & Bonds Guide · BlackRock's capital market assumptions, cited by name.