Financial Professional
Institutional Investor
Individual & Shareholder


April 2022

Retirement Income as a Psychological Portfolio Problem


We believe a well-formed financial plan can offer investors a solid foundation for success. Even so, any number of current market conditions may lead investors to question their approach. Today, many investors are grappling with the potential effects of rising interest rates and evaluating the role of fixed income in their portfolios.

For investors approaching (or in) retirement today, these concerns may be compounded by uncertainty over whether their portfolios can sustain sufficient income throughout their lifetimes. In this article, Professor Suzanne B. Shu suggests that a more comprehensive approach considers how a total portfolio of investment and insurance strategies can work together. She expands on common psychological barriers we encounter when contemplating insurance-based strategies, using tontine insurance as a case study.


When discussing asset allocation, Investment professionals are fluent in how a well-balanced portfolio can tradeoff between risk and return to help meet a variety of investment goals. But what about longevity risk and especially the question of decumulation?

Retirement income portfolios face several barriers to adoption, be they market-driven (e.g., differences in delivery channels between investment products, insurance products and government benefits) or more psychological (i.e., our tendency to turn to the specific options with which we’re already most comfortable).

In this article, I consider some of the risks and returns of retirement that investors worry about (both financial and psychological) and suggest a portfolio approach of leveraging multiple products to meet those needs. To illustrate the psychological tradeoffs, I’ll start with the story of the U.S. tontine insurance market.

A Case Study in Tontine Insurance

Tontine insurance was a widely successful financial product in the U.S. from the late 1800s into the early 1900s. For example, one large mutual insurance company that began offering tontine insurance in 1881 doubled its business within two years. Estimates suggest that by 1905, tontine insurance may have been held by 50% of American households and totaled $6 billion in value.1

The product was built upon the old European idea of tontines (pooled investment funds that paid out to survivors) with some important changes. Participants in tontine insurance plans paid an annual premium in which one portion of the premium purchased life insurance, while the other portion went into an investment fund managed by the insurance company for a specific period (usually 20 years).

If the individual died before the time frame expired, the beneficiaries received a life insurance payout. If the individual was still alive at the end of the time frame, he or she shared in a division of the entire accumulated amount of original funds and dividends with other fund survivors, whereas non-survivors were defined as both individuals who had died during the time frame and individuals who had lapsed on their annual premiums.

Issues of moral hazard were avoided since participants were part of blind national pools where other members of the same tontine fund could not be identified. Ransom and Sutch, in their review of the history of tontine insurance, call it “life insurance where survivors won.”

The overwhelming success of the funds led to enormous amounts of money under discretionary management of the insurance companies and which government investigators claimed was being improperly used to enhance the status and influence of the managers. As a result of this misuse, tontine insurance was shut down in 1905 following the Armstrong Investigation.

Besides the benefits to the client of combined life insurance and longevity insurance, tontine insurance benefited the financial provider relative to traditional annuities by shifting longevity risk onto the customers themselves; the provider simply collected a fee to administer the tontine product.

Ransom and Sutch sum up the product by writing, “Considered as a financial innovation, it was very successful. Considered as insurance, it was actuarially sound. Considered as a gamble, it was a fair bet in as much as there was no percentage for the house beyond a charge to cover administrative costs. Considered as a life-cycle asset, it proved to be an excellent investment, earning a rate of return substantially in excess of that generally available on other assets.”

While tontine insurance itself is now illegal in the U.S., the success of the product can provide useful insight into new ways to structure current decumulation products. What made tontine insurance unique was that it offered the largest benefits to the client when one of two aversive extreme outcomes occurred: The individual died early or lived beyond the fund’s time frame.

When deciding on savings and decumulation products, consumers are often searching for protection from these two extreme events of early death and long life. Products already exist for each of these outcomes individually—life insurance for early death and products like annuities or longevity insurance (aka deferred annuities or QLACs) for long life.

Both life insurance and longevity insurance are often purchased through relatively small monthly payments during an individual’s working years, in contrast to a traditional life annuity in which a single large payment of funds is exchanged after the individual is already retired. A portfolio of risk products, rather than individual stand-alone products, should have a significantly stronger appeal for the consumer since both types of extreme outcomes are covered. But it is only by combining the possible outcomes of multiple products that the protections from the extremes can be achieved.

Furthermore, the psychological benefits provided by tontine insurance may also prove helpful in better understanding the perceptions and biases that prevent investors from embracing insurance-based risk products. The three main psychological benefits that were offered by tontine insurance include:

  • The aforementioned protection from two negatively correlated extreme outcomes (i.e., it’s difficult to predict early death or living too long).
  • Perceived fairness in fees.
  • Intertemporal comparative optimism (i.e., how each individual expected to fare in the long run relative to other participants in the fund).

Each of these psychological aspects of tontine insurance, and their applicability to the decumulation market of today, is considered below.

Uncertain Futures and Extreme Outcomes

For individuals, decisions around investment decumulation require consideration of several uncertain outcomes. The most important of these uncertainties is life expectancy, an inherently difficult prediction due to the wide range of possible outcomes and unknown health factors.

Other important sources of investment uncertainty are future interest rates and company default, which have become even more salient in recent years. Finally, another major uncertainty concern for many individuals is the desire and ability to manage their own capital and expenses in retirement.

Just as individuals overestimate their abilities in other domains, people are generally overconfident in predicting these skills, especially as they age. The impact of each of these uncertainties may be manageable in small doses but become overwhelming when they lead to extreme swings in outcomes.

Research has found that individuals can be comfortable with small differences within distributions of outcomes but pay particular attention to certain parts of the distribution that are important to them, such as extreme upside or downside risks.2 Consideration of individuals’ concerns about extreme outcomes may help to suggest prescriptive solutions for increasing demand for decumulation products.

Perceptions of Fairness

In research with my co-authors John Payne and Robert Zeithammer, we have found that concerns about fairness are the strongest indicator of who is unwilling to consider life annuity products, regardless of how attractive of a financial deal the annuity offers.3,4

In our work, we asked questions such as whether it is “fair that the insurance company is allowed to keep the excess funds” after the annuitant dies—individuals who perceive this statement as unfair were highly averse to life annuities, even though such outcomes are fair from an actuarial standpoint of the insurer balancing longevity risk within the larger population.

We believe that tontine insurance avoided this perception of unfairness and in fact benefited by appealing to consumers’ desire for outcome fairness by segregating the insurer’s profits from the product’s pooled funds.5  With tontine insurance, an early death still receives a payout from life insurance.

Furthermore, it is highly salient to the tontine insurance owner that, in the case of early death, her portion of the assets in the fund would go to other participants through the tontine payouts rather than the insurance company. The providing company would receive only a flat fee for managing the fund, which may be perceived as significantly fairer. More generally, understanding individuals’ perceptions of fairness for retirement products, and how they impact market participation, can assist us in designing better decumulation options.

Intertemporal Optimism

When considering outcomes that unfold over long periods, as is true for any decumulation solution, individuals must forecast their own uncertain futures. Beyond forecasting their own mortality, the other intertemporal forecast that individuals may consider in these decisions has to do with uncertainty relative to the future outcomes of other people.

Especially for tontine insurance, in which one’s future payments depend not only on one’s own survival but on the survival of the other people in the pool, any bias in judging one’s own outcomes versus others’ outcomes can lead to different willingness to purchase the product. Substantial research on social comparative judgments has shown that individuals make biased predictions for their own outcomes in the direction of expecting more positive outcomes for themselves relative to similar others.

These effects include perceptions of being more athletic, better drivers, more polite and better organized, but also include being less susceptible to heart attacks and less likely to have undesirable health problems.6 Such biases may affect the perceived benefits of a pooled decumulation plan (such as a tontine) if the individual believes she has a better chance of being one of the survivors and reaping the benefits that are associated with that outcome—in other words, a biased assessment of the probability of surviving will lead to a higher expected value for purchasing the product.

Framing the life expectation question in terms of the chance to “live to” an older age rather than “die by” that age can further support these optimistic projections.7 Helping clients think about the positive chances of their own longevity can increase interest in decumulation options that provide guaranteed income for a long future.

Implications for Retirement Planning

When attempting to craft a retirement income solution, it may seem most straightforward to consider each financial product as an independent decision. Trying to jointly optimize over Social Security benefits, pension income, annuity options, housing wealth and retirement savings can be overwhelming for individuals to contemplate.

However, it is only when the full set of options are jointly considered that the downsides of one option can be paired with the upsides of another, and a more balanced but also more optimal solution can be devised. Tontine insurance was successful because it paired two products that addressed opposite risks—living too long and dying too early—and did so in a way that individuals perceived as fair while appealing to their individual optimism.

Just as an optimal investment portfolio balances risk through negatively correlated outcomes, an optimal decumulation portfolio can use combinations of products to help consumers navigate the risks of retirement.



1Roger L. Ransom and Richard Sutch, “Tontine Insurance and the Armstrong Investigation: A Case of Stifled Innovation, 1868-1905,” Journal of Economic History 47, no 2 (1987): 379-390.

2Daniel G. Goldstein, Eric J. Johnson, and William F. Sharpe, “Choosing Outcomes Versus Choosing Products: Consumer-Focused Retirement Investment Advice,” Journal of Consumer Research 35, no. 3 (October 2008): 440-456.

3Suzanne B. Shu, Robert Zeithammer, and John W. Payne, “The Pivotal Role of Fairness: Which Consumers Like Annuities?” Financial Planning Review 1, nos. 3-4 (2018): 1019-1022.

4Suzanne B. Shu, Robert Zeithammer, and John W. Payne, “Consumer Preferences for Annuity Attributes: Beyond Net Present Value,” Journal of Marketing Research 53, no. 2 (2016) 240-262.

5Daniel Kahneman, Jack Knetsch, and Richard H. Thaler, “Fairness as a Constraint on Profit-Seeking: Entitlements in the Market,” American Economic Review 76, no. 4 (1986): 728-741.

6Jonathon D. Brown, “Evaluations of Self and Others: Self-Enhancement Biases in Social Judgments,” Social Cognition 4, no. 4 (1986): 353-376.

7John W. Payne, Namika Sagara, Suzanne B. Shu, Kirstin C. Appelt, and Eric J. Johnson, “Life Expectation: A Constructed Belief? Evidence of a Live To or Die By Framing Effect,” Journal of Risk and Uncertainty 46 (2013): 27-50.


This material has been prepared for educational purposes only. It is not intended to provide, and should not be relied upon for, investment, accounting, legal or tax advice.

The contents of this Avantis® Investors presentation are protected by applicable copyright and trade laws. No permission is granted to copy, redistribute, modify, post or frame any text, graphics, images, trademarks, designs or logos.