Is Longevity Risk an Estate Planning or Retirement Planning Problem?

America has become sophisticated at planning for how long people may live. We model returns, withdrawal rates, taxes, Social Security and longevity down to decimal points. But we still define longevity risk too narrowly.

It is usually described as the risk of outliving your money. But longevity risk does not fit neatly inside either a retirement or estate plan. A change in health can increase spending, reduce your ability, or a family member's ability, to keep earning, force assets to become liquid and reshape what remains for the next generation.

When I was 16, my aunt, who was like a second mother to me, was diagnosed with terminal colon cancer. For the next two and a half years, my mom became her primary caregiver. Family members stepped away from work and school, bringing lost income, exhaustion and financial strain. We understood cancer was a health crisis. We had no idea it could also become a labor, retirement and family crisis. Longevity is not simply "more retirement." Health changes how we use wealth.

The Hidden Liability Is Often Labor

The cost of care is not always a check written to a nursing home. Sometimes the bill is a spouse becoming a full-time caregiver, a daughter reducing her hours or someone leaving the workforce during peak earning years. When professional care is not available or affordable, the care does not disappear. Someone provides it. America therefore has not only an unmodeled LTC expense inside millions of retirement plans, but an unmodeled transfer of wealth and labor between generations. A plan designed to leave assets to children can quietly become one in which those same children provide thousands of hours of unpaid care, risking their own ability to build wealth.

By your 40s and 50s, longevity planning can become a two-generation problem. Parents may need help precisely when you are raising children, advancing your career and accumulating assets for retirement. Advisors should ask not only, "What happens if you need care?" but, "What happens if one of your parents needs you?"

Try the "Need Care Tomorrow" Test

One of the most useful exercises I have seen an advisor do with a client is surprisingly simple: pretend you need care tomorrow. Who gets the first call? Where do you want care? Who finds the caregiver? Which account pays? Which assets were intended for your care, your spouse or the next generation? If cognitive decline is involved, who can actually carry out the plan? A family can have more than enough money and still be profoundly unprepared.

What often blows up a retirement plan is not a diagnosis but the cascade around lost independence: paid care, coordination, quickly liquidated assets, reduced work for a spouse or child, and a plan its creator may no longer be able to execute. A dementia diagnosis can be devastating, but what often breaks the plan is everything it requires the family to do.

The Problem With Planning Around the "Average" Person

After my aunt died, I studied Genetics and Data Science at UC Berkeley and spent much of my career building technology. Years later, researching long-term care, I wanted to know whether my family's experience was unusual. It wasn't. I kept hearing the same story: families spent decades accumulating assets for retirement, yet had almost no model for one of its largest potential expenses. When they did plan, they were often given population averages.

But no family actually experiences the average. A 58-year-old's health history, geography, finances, family structure and preference to remain at home can produce a dramatically different care journey from another 58-year-old. That became the premise behind Waterlily.

The simplest way I explain what we do is this: most retirement planning forecasts how much money you may have. We help forecast one of the biggest things you may need that money for if your health changes, then help you decide what to do about it.

We built models drawing on more than 500 million data points to estimate an individual's potential care journey: whether they may need care, when it could begin, how long it could last, how needs might progress, local costs and how much may otherwise fall on family caregivers. But the useful part is not simply the probability. It is being able to rehearse the financial event before it happens. What if care begins at 79? Lasts two years versus seven? What if you want to remain at home? How much family help would that require? What happens to the portfolio and your spouse? If we can personalize the risk, we can personalize the decision.

Personalization Changes Behavior

One of my favorite examples involved an advisor whose client Waterlily predicted had a 97% likelihood of eventually needing long-term care. You might assume the 97% changed the client's mind. It didn't. Even a startling probability can remain abstract. What changed the conversation was showing what sat behind it: when care might begin, how long it might last, what it could cost and what those scenarios could do to the financial plan and family.

After reviewing their full personalized results in Waterlily, the client's response was essentially: "Okay. How much do you want to budget?" Numbers get someone's attention, but simulation can change behavior. The goal is not to scare everyone into buying insurance. Some families can self-fund, some already have protection and others may transfer only part of the risk. The goal is an intentional decision understood in the client's own terms.

The Longevity Conversation Should Come Before the Product Conversation

Historically, long-term care conversations have often started with a product: "Do you have long-term care insurance?" I think the better conversation starts with the family. If you lose independence, where do you want to live? Who do you expect to help? Does that person know? What would providing care do to their life? What assets are you willing to spend, preserve or pass on? What needs to remain for a spouse? If you cannot make those decisions yourself, does the person acting on your behalf understand those priorities? Only then should you ask which financial, insurance and estate-planning tools belong in the plan.

AI can help because no human can reasonably analyze millions of combinations of health, care settings, family capacity, assets and insurance designs for every client. The machine can calculate possible futures. A person still has to decide which future they care about protecting.

"I'll Just Work Longer" Is a Terrible Assumption

Another longevity assumption deserves scrutiny: "I'll just work longer." Delaying retirement is a backup plan for many Americans if the numbers do not work at 65. But it assumes your health, and your family's health, gives you permission to keep working.

A parent who needs care at 62, a spouse who becomes ill at 64 or your own loss of health can turn planned earning years into unexpected withdrawals. Health cannot sit outside the retirement model because it can change expenses, earning capacity and timing at once. Yet planning information remains fragmented: your physician understands pieces of your health, your advisor your assets, an insurer certain risks, and your family your preferences and what they realistically can or cannot do.

Taking it home, literally

Very rarely does anyone put those things together. Yet that is what longevity planning requires. So is longevity risk an estate-planning problem or a retirement-planning problem? It is both, because the same health event can determine how much you spend during your lifetime, who has to help you and what wealth ultimately remains afterward. The next evolution of planning will require modeling health, wealth and family capacity together rather than as separate systems. We cannot tell someone exactly what life will look like at 83. The value is preserving optionality.

Planning earlier can determine whether you remain at home longer, whether your spouse has support, whether your children become default caregivers and how many choices remain when health changes. When my aunt became sick, one emergency created the next. We could not control her diagnosis, but I can now see how much might have been different if we had understood the journey earlier. Building Waterlily has changed my definition of control.

Control does not require certainty. Sometimes it simply means having enough information, early enough, to still have choices and improve our chances of creating a better future for our family and the next generation of our aging population.

About the Author

Lily Vittayarukskul is the founder and CEO of Waterlily, an AI-powered long-term care planning company on track to reach hundreds of thousands of families by the end of 2026. A 2x World Economic Forum speaker, she has been featured in The New York Times and USA Today and was named to both the 2026 Forbes 30 Under 30 list and WealthManagement.com’s 2026 Ten to Watch. She studied Genetics and Data Science at UC Berkeley.