With the shift to defined-contribution pension plans from defined-benefit plans, it is increasingly individuals — rather than institutions — who are required to make complex decisions about how much to save and how to make those savings last.

In order to assist both prospective retirees and their financial advisors, Shlomo Benartzi, a professor at the University of California at Los Angeles, recently sought out several fellow academics in the behavioural science field, asking each of them for a key insight into creating an effective framework for retirement planning.

Columbia University professor Eric Johnson, for instance, has found that a great many retirees suffer from what he calls “hyper loss aversion.”

It is no surprise that retirees are more concerned now about financial losses than they were when they were younger. But even more surprising was the extent of their concern. For instance, in Johnson’s study, retirees wouldn’t accept a “coin flip” equivalent bet unless they could expect to win $100 on heads for every $10 lost on tails, which suggests participants weighted losses about 10 times more heavily than gains. (Earlier studies have shown that the average investor needs to expect to win much less than that in order to feel, on an emotional level, that he or she will break even.)

Johnson’s study also found that many retirees are leery about financial products that offer substantial protection and guarantees. Johnson has concluded that many retirees view insured investments, which come with a guaranteed minimum withdrawal rider, as a form of sacrificing control of their money — and perceive this as a loss.

Declining purchasing power is another major retirement issue. Most people think in nominal dollars and overlook the corrosive effect of inflation. This is what Princeton University’s Eldar Shafir refers to as the “money illusion.”

In Shafir’s study, when the future value of savings was outlined in nominal dollars — in other words, unadjusted for inflation — retirees focused on the amount of money on hand. But when that future value was stated in real dollars, reflecting the erosion of purchasing power, participants tended to favour an inflation-protected solution. What’s equally interesting, however, is that when a retirement solution was presented in a more neutral way, preferences were similar to those when nominal dollars were the focus.

Numerous behavioural studies have established that “framing” decisions can have a huge impact on investment decisions. There’s considerable difference, for instance, between advising people to spend about 70% of their current income in retirement and telling them they have to eliminate 30% of their existing expenses to manage their affairs. Even though these two proposals are mathematically equivalent, most people find the 30% option to be more of a challenge.

Professor Jeffrey Brown of the University of Illinois applied this logic to creating retirement income, asking prospective retirees to opt for either a life annuity that pays $650 a month or a savings account of $100,000 bearing the equivalent interest. The two choices were designed to have the same actuarial value in order to ensure a balanced comparison.

Half of the group in Brown’s study was presented the annuity choice in a “consumption” frame — a monthly income of $650 for life — and half was presented the choice in an “investment” frame — a monthly return of $650 for life. Seventy per cent chose the annuity when it was presented in terms of monthly income; only 21% chose an annuity when it was expressed in the form of a monthly return.

Benartzi compiled his overall results into a checklist for those charged with putting together retirement plans. Although the checklist is designed for plan sponsors and policy-makers, it can provide advisors with a useful guidepost when mapping out clients’ futures.

His report, entitled Behavioural Finance and the Post-Retirement Crisis, along with that checklist, can be found at www.allianz-investors.com/PublicRelations/External%20Documents/Allianz_DOL_RFI_Response.pdf. IE