The Canadian Institute of Actuaries says a 65-year-old man now has a life expectancy of 87 years. This is an increase over projections made just a few years ago. Uncertainty about future improvements are forcing actuaries to rethink longevity risk management.
The Institute’s new mortality table assumes a life expectancy of about 1 per cent to 1.5 per cent higher than what was projected in 2014 for both men and women, says a Morneau Shepell bulletin, recently published on this subject. The consulting firm says the impact on the value of an annuity would be approximately 0.5 per cent.
Morneau Shepell says that this is the first mortality table to be based on the work of both life insurance and pension specialists. The bulletin says the new mortality table is innovative because the rate of improvement in long-term mortality was based on the rates used for projections of public plans in Canada, the United States and England. The previous measure was based solely on the projections of the Canada Pension Plan.
Thierry Chamberland is a member of the Canadian Institute of Actuaries’ committee on pension plan financial reporting. He is also a consulting actuary and a partner with Aon Hewitt. He participated in an interview with The Insurance and Investment Journal on behalf of the Institute.
In a presentation on longevity risk delivered at the Annual Employee Benefits Conference in 2017, he underlined that demographic trends exacerbate this risk. The Center for Strategic and International Studies’ 2013 GAP Index says public plans payable to the population aged 60 and over reached 9.3 per cent of Canadian GDP in 2010. This is forecast to reach 11.8 per cent in 2020, 14.2 per cent in 2030, and 15.8 per cent in 2040.
For its part, Mercer Canada has released a report on the rapid changes that are shaking the industry. The report, published in 2017, called Brace for (or Embrace) The Change in the Canadian Retirement Landscape, says 5,000 people retire every week in Canada. The pace is expected to increase to 8,000 per week by 2020.
Faced with this growing risk, actuaries working with defined benefit pension plans welcome the Canadian Institute of Actuaries’ new mortality tables. Mathieu Tessier, Sun Life Financial’s director, client relationships, says that the final tables will apply to all insurance companies for all life products, including annuities. “The Institute would eventually like to prescribe new tables for pension plans,” he adds.
Tessier says the tables help plan sponsors who want to mitigate their longevity risk. These projections are a crucial first step in a continuing process that requires regular updates. “This aspect is important because medical research moves extremely fast,” he explains.
An unknown risk
Michel Fortin, Quebec wealth leader for Mercer Canada, says it is easier to manage mortality rates determined in a uniform table across Canada than to predict the future. “Mortality is a basic assumption, about 87 years for men aged 65 and over in this case. Longevity, meanwhile, is the future improvement of life expectancy. It’s the biggest unknown,” he says.
The Institute recognizes the limits of its work on longevity. There is very little data available to study the improvement of mortality at a finer level than the general population and to determine whether the best estimate assumption for the general population could or should be modified for different segments of the population, it explains in a consultation document.
The document explains that very little data exists to segment populations based on indicators such as socioeconomic level, gender or lifestyle of plan members. It says variables such as future technological and medical advances could have a significant impact on the longevity of Canadian citizens. Uncertainty is even more pronounced for the more advanced ages (90 years and older), an age group for which there is minimal data.
To better understand longevity risk, Mercer has partnered with an international firm specialized in risk modeling, Risk Management Solutions (RMS), says Michel Fortin. “Life expectancy has improved at an accelerated pace. Will it continue? With RMS, we are developing a tool that will allow us to more precisely establish the longevity prognosis of an individual, to better ensure the sustainability of group pension plans.”
Changing lifestyle, technology, medicine, public health initiatives, environment, genetics can indeed change the situation. RMS will include these vectors in the development of its model.
“Before, the industry was limited to a man-woman distinction. We will go further: does the individual begin his or her retirement in good health or disabled? What is his postal code? And the amount of his pension? It is known that pensioners with higher incomes have a better life expectancy. And what about the morbidity rate among young people? Obesity is a good example of the unknown.” says Michel Fortin.
Mercer will launch its mortality model in Canada at the end of 2017 or in early 2018, he says. “Rather than having only one general mortality table, we will have refined longevity assumptions.”
Sun Life also put a longevity team together. It uses data provided by RMS.
“We want to model risk based on global research. It is not only a matter of segmenting the population, but also of understanding the causes of death. We want to understand how there could be improvements in the future, as has been the case in recent decades with cardiovascular disease,” says Mathieu Tessier.
He says longevity risk plays out over the very long term which can lead to surprises.
“Unanticipated mortality improvements over the last four decades have resulted in increases in pension liabilities of 20 per cent to 25 per cent. Nobody knows if we have the right answer today. The main causes of improvement could be medical advances such as precision medicine and immunotherapy, a new approach to old age, which could now be treated as a disease...Or, on the contrary, an external force could decimate populations, such as new unknown and difficult to treat illnesses. “
Increase in plan liabilities
Tessier estimates that while the change to the new mortality table should lead on average to a 0.5 per cent increase in pension liabilities, this may vary according to the age and gender of a plan’s participants, among other factors. A year of unanticipated improvement in longevity can increase plan liability by about three per cent to four per cent.
Models based on socioeconomic factors are increasingly being used to measure expected longevity, says Chamberland. The United Kingdom is a pioneer in this area. “This approach did not exist here five years ago. The environment in which we live greatly influences our life expectancy. For example, we can segment the members of a plan according to postal codes, in order to have a more personalized table,” he explains.
In his presentation, he explained that postal codes are strongly correlated with socio-economic and geographic factors that can influence mortality experience: relative wealth; urban or rural region; proximity to hospitals and other community services; geography; cultural backgrounds; whether or not a zone is industrial.