Grad Students Apply Probability Studies
To Long-Term Care Insurance Coverage
February 7, 2000
Research by two Ph.D. candidates in actuarial science is expected to help insurance companies establish fair prices for long-term care coverage and help regulators determine adequate reserves for the industry.
The project by Chayanna Siripirom and Guy Rasoanaivo is being directed by Charles Vinsonhaler, professor of mathematics and head of the mathematics department, and Jay Vadiveloo, head of the actuary department at Aetna Financial Services and an adjunct professor of mathematics at UConn.
"Long-term care is one of the hottest insurance coverages, as the baby boom generation approaches retirement age," Vinsonhaler says.
There are two kinds of long-term care insurance products, he says - standalone coverage and coverage that treats long-term care payments as early withdrawals from life insurance death benefits. Yet in determining a fair price and estimating the net cost of future obligations, there is a lack of credible data to estimate when a disability may occur and how long it will last. It is also difficult to quantify the full range of possible costs.
The research project is trying to solve these problems by building a complex actuarial model. "The model is flexible enough to capture both types of coverages, design variations within those coverages, and determine a pattern of benefit payments," Vinsonhaler says.
Long-term care insurance covers a person against the cost of services for people who have disabilities or chronic - long-lasting - illnesses. Those services can include:
The American Association of Retired Persons estimates that the annual cost of a nursing home stay is about $50,000, and a person who receives three home health visits a week could pay about $12,000 a year. Data from the New England Journal of Medicine suggest that for a man over 65 the odds are one in three that he will need long-term care at some point. For a woman over 65, the odds are one in two, because women live longer.
"The demographics support the research," says Vadiveloo, adding that the government also has a vested interest in it, because care paid for by private insurance can save tax dollars.
"Long-term care is one of the hottest selling products by insurance companies and standalone health companies," he says. "This research will be well received by individuals and by regulatory agencies."
Vadiveloo says it is difficult for companies to properly price long-term care insurance because of inadequate actuarial data for older people, especially as improvements in medical technology will lengthen nursing home stays.
"You want to be able to establish models to see all the possibilities of things that may happen," he says. "Companies may recognize the volatility of the product but not be able to determine the price. There is no literature on the subject."
For example, Vadiveloo says some preliminary results of the research indicate those costs to insurance companies may be as much as four times higher than their own estimates.
The research, which will take about a year to complete, will have academic benefits and practical benefits for the actuarial profession, for the insurance industry and for regulators. It is expected to lead to a marketable product.
"It's great to provide benefits, but regulators want to know if the company is able to pay the benefits," Vadiveloo says. "One of the things to come out of the research may be a way to capture the volatility and estimate the correct amount of reserves companies must establish."
"This research could even benefit Aetna, because long-term care may be added to the payout of annuities," Vadiveloo says. "There's a variety of research in actuarial science at the University, but for a practitioner, much of it isn't so useful. But in the industry, I wish I had the trained resources the university has. This is a good thing for both the University and Aetna."
Vinsonhaler agrees: "We have the math skills, but without Jay, we would have no way of getting into something so practical."