For centuries, people have used actuarial tables to figure out how long they are likely to live. Now artificial intelligence is taking on this task – and its answers could well be of interest to economists and money managers.
The recently released Death Clock, an AI-powered app for longevity, has proven to be a hit with paying customers; it has been downloaded about 125,000 times since its launch in July, according to market research firm Sensor Tower.
The AI was trained on a dataset of more than 1,200 life expectancy studies with around 53 million participants. It uses information about diet, exercise, stress levels and sleep to predict a likely date of death. The results are a “pretty significant” improvement over standard lifespan expectations, says the developer, Brent Franson.
Despite its somewhat morbid tone – it features a ‘loving farewell’ death anniversary card with the Grim Reaper – Death Clock is catching on among people who are trying to live a healthier life. It ranks highly in the Health and Fitness category of apps. But the technology potentially has a wider range of applications.
Life expectancy is key to a range of financial and economic calculations by governments, businesses and individuals – from retirement income needs, to policy coverage in life insurance and pension funds, and financial planning.
In the US – which has lagged behind other developed economies in recent years in the life expectancy of its citizens – the Social Security Administration has its own mortality rate table, which appears in the annual financial report to trustees.
The government agency currently predicts that an 85-year-old man in the US has a 10% chance of dying within a year, and living an average of 5.6 years. But such averages can be significantly off, says Franson, and the new algorithms can yield a more customized measurement: a customized death clock.
The importance of such findings in economics is evident from the publication – just this past month or so – of two articles on the subject by the National Bureau of Economic Research.
‘Take advantage of the benefits’
One, entitled ‘On the Limits of Chronological Age’, examines the various ways in which the aging process affects physiological capabilities. It finds that many aspects of economic behavior, such as willingness to join the labor force, may not be well captured by people’s calendar ages – even though policies such as statutory pensions are typically based on this.
By continuing to rely on chronological age as a measure of how well people can function, societies may ultimately fail to “fully reap the benefits of longer lifespans,” the Harvard and London Business School researchers conclude.
Another working paper examined ‘value per statistical life’ or VSL – an insensitive-sounding metric used for cost-benefit analyzes in areas such as pollution regulation or industrial accident compensation. It is usually calculated based on pay for workers in high-risk jobs.
The researchers behind the NBER study “The value of Statistical Life for Seniors” relied on a different data set: the propensity of older Americans to spend money on medical services that reduce mortality risk. They found an average VSL at age 67 of just under $2 million for people who reported their health as “excellent,” compared to $600,000 for those in “good” health.
When it comes to personal finance, better measures of life expectancy will have profound implications for people saving for retirement, said Ryan Zabrowski, a financial planner at investment advisory firm Krilogy.
“A major concern for the elderly, our retirees, is the survival of their money,” says Zabrowski, who addresses the issue in his forthcoming book “Time Ahead.”
‘Out the Window’
Decisions such as how much to save and how quickly to withdraw assets are often based on generic and unreliable life expectancy averages. AI-driven testing that could potentially reduce this uncertainty is largely unheard of now, but likely won’t be such an unusual idea in the future.
Furthermore, AI technology itself, together with advances in medicine, has the potential to increase life expectancy – and with it the risk of running out of savings. Zabrowski thinks one consequence is clear: Longer retirements will mean savers will need higher-yield investments for their old age, pushing them to invest more in stocks instead of fixed income.
“The conventional method of measuring stock demand will be thrown out the window,” he writes in his forthcoming book. As people expect to live longer, there will be a “massive escalation in demand for stocks.”
There are already numerous technologies available – such as heart rate monitors and maximum oxygen consumption meters from wearables – that, when combined with new AI-powered devices, have the potential to reduce uncertainty around personal mortality.
Of course there will always be limits. In addition to completely unpredictable variables, such as accidents or even pandemics, there are numerous intangible assets.
Longevity gap
For example, it is often thought that loneliness reduces life expectancy. Gratitude can increase this. A Harvard study found that women who reported feeling the most grateful had a 9% lower risk of dying within three years than those who reported feeling the least.
Then there is the issue of inequality. Money is important for life expectancy. Multiple studies – including Nobel Prize winner Angus Deaton’s work on “Deaths of Despair” – have shown a clear divide between rich and poor Americans.
Research published by the American Medical Association shows that the lifespan gap between the richest and poorest 1% at age 40 is nearly 15 years for men and 10 years for women.
For Death Clock users, who must pay $40 a year to subscribe, the app suggests lifestyle changes that can keep mortality at bay — along with a second-by-second countdown of estimated time remaining.
“There is probably no more important date in your life than the day you die,” says Franson.