Pension shortfalls will be exacerbated by underestimating life expectancies, a report released Thursday by Swiss Re. An essential element to preparing for longevity risk, the reinsurer found, is to develop better predictive methods for determining mortality.
“The failure to consider future drivers of mortality in historical predictions contributed to employer pension funds under-reserving for longevity risk and other bodies, including governments, not budgeting effectively for funding an ageing population,” Daniel Ryan, head of life and health research and development at Swiss Re, said in a statement.
Effective longevity models would have to take social factors, medical treatments and preventative approaches into account. The report recommends that pension plans assess their exposure to longevity risk and decide whether to pass it on to a third party that is better equipped to take on the risk.
Traditionally, the most common way to predict life expectancy is to combine current rates of mortality improvement with a long-term future assumption, the paper found. Computer-based models that provide forecasts based on historical mortality experiences are also used.
The paper notes, though, that the “complex interaction” of various risk factors, including disease and improved treatments and technology, make it difficult to predict longevity based on historical trends alone. As such, a “life course” approach would take more than history into account when predicting mortality, including future developments in societal changes, medical practice and technological advances.
In order to better understand mortality, the paper suggests building a multi-discipline approach that involves experts from various professions:
- Actuaries to assess the financial impact of future events
- Medical experts to provide insight into future developments in diagnoses and disease
- Epidemiologists to analyze risk factors that lead to disease
- Pharmacologists for their expertise on new drugs and their impact
- Demographers to help aggregate influence on population size
- Gerontologists for their insight into the physical, mental and social effects of aging
- Governments have a “vested interest” in understanding their citizens’ longevity
Swiss Re provides an overview of some of the considerations for building a forward-looking model. Below are some of the factors that influence multiple and individual diseases.
General drivers to diagnosis and survival
- Individual risk factors including age, gender, diet
- Health care funding: Public versus private, influence of patient advocacy groups, cure versus prevention
- Patient-doctor interaction: Awareness, increase in available treatment options, growing emphasis on shared decision making
- Research and development: Public versus commercial sponsors, regulators’ attitudes toward developments, pharmaceutical versus biotechnological industries
Developing a disease-centered model for mortality rates, according to Swiss Re, will involve committing resources to scenario development and the importance of a disease to the overall model. For example, drugs may improve life expectancy for breast cancer patients, but will have a limited effect on the life expectancy of the entire population.
“Overall, it is necessary to develop and quantify suitable scenarios for future diagnosis rates and survival rates after diagnosis for use in a life course model,” according to the paper. “This is relevant for each disease—or combination of diseases—that is tracked in the model.”
Pension plan managers should understand the risk of underestimating the costs of longevity, the paper recommends, while acknowledging that however much analysis is undertaken, there will always be uncertainty.
Swiss Re recommends that insurers work together and with reinsurers to manage longevity risk.
Regulators have already recognized and responded to the need to increase retirement ages. “It is logical that retirement ages should relate to the expected longevity of citizens and more forward-looking approaches can help create a fairer system, along with demonstrating the difference in life expectancy between various groups with society.”