February 17, 2020

Rethinking Retirement......Davos Style

“To maximize the advantages of a longer life, both education and traditional career paths need to be reconsidered. Retirement systems need to be sustainable and respond to the needs of a changing society. To disarm the demographic ‘timebomb’ and unlock the longevity dividend, we need to rethink retirement.”

Urs RohnerBoard ChairCredit Suisse Group


Almost a Davos Alumnus

A surprise invitation from Credit Suisse (CS) appeared in the KPA Inbox in last October. It was to participate in a discussion on ‘rethinking retirement’ at the January 2020 World Economic Forum gathering of global luminaries in Davos Switzerland. The discussion would be based on a specially-commissioned CS study titled “Rethinking Retirement”, to which I was invited to make a contribution. Shortly after submitting my contribution to the Study in December, there was another CS email in the KPA Inbox: while the Study would be released at a Davos press conference, the ‘live’ discussion of its contents was cancelled. As things turned out, the Study release date was Tuesday January 21, the day that headliners Donald Trump and Greta Thunberg both made their Davos appearances. That would have been tough competition for a ‘live’ session on rethinking retirement.

And of course, Trump and Greta were not the only Davos attention-grabbers that week. There was serious competition from topics ranging from stakeholder capitalism, tree planting, common ESG metrics, SDG implementation, the future of health care, the future of the media, clean energy provision, disability inclusion, global supply chains, and fighting modern slavery. In short, “Rethinking Retirement” was in a tough match for the attention of the attending global luminaries and media. Which gets us to the goal of this Letter: to highlight the key insights of the very good CS “Rethinking Retirement” Study, and to link you to the full 61-page version should you want to get beyond the highlights.   

While many of the ‘key insights’ of the Study are universal, country-specific application often requires adaptation to the specific culture and rules of that country. The Letter points to some of these specific insights for Australia, Canada, France, Japan, Netherlands, UK, and USA as well.    

A Brief History of ‘Retirement’

The Study reminds us that ‘retirement’ is a relatively new concept. For example, Reichs Chancellor Otto von Bismarck instituted the concept of a universal age-based pension in 1881. Why? Because the industrial revolution was forcing a rethink of the old 2-stage personal life cycle (prework and work) and replacing it with a 3-stage version (prework, work, and post-work). Given the rise of radical socialism at the time, the crafty Chancellor thought it prudent to ward off the possibility of the rise of an aging proletariat that would one day be too old to work in the factories, but without the financial means to support themselves.

And Bismarck was not alone in recognizing this….employers themselves began to see value in promising their employees a pension in retirement. An important factor in these pension provision decisions was the life expectancy of retired workers in the late 19th Century. With pension commencement typically set at age 65, life expectancy was measured in just a few years rather than decades, making pension provision an affordable proposition at that time. However, as decades passed, improved diets and living conditions, as well as major medical advances led to a steady rise in life expectancies, leading to a steady rise in the cost of pension provision.

There were other ‘age’ factors at work with the passage of time. One was a growing gap between ‘chronological age’ and ‘biological age’. As a specific example, a Swiss study showed that the average male aged 65 today has the same mental and physical capabilities as the average 51-year old had in 1950. So if age 51 was not a logical retirement age in 1950, age 65 should not be a logical retirement age today. Further, longevity and biology are not the only two ‘age’ factors impacting the changing retirement landscape. At the other end of the life-cycle, falling fertility rates have also become an        important contributor to reshaping population age distributions in the developed world towards relatively more oldsters and fewer youngsters.

Finally, these ‘age’ issues are not the only drivers leading to the need to rethink retirement. The old work model of a 40-year career with one employer is fading fast, as is the nature of work itself. Young workers are entering the workforce later, are forming families later, and are building work histories involving multiple employers in multiple sectors of economies subject to radical uncertainty from multiple sources. Last but not least, historically-low interest rates suggest future returns on retirement savings will not match historical post-WWII realizations.    

The Global State of Retirement in Numbers

So where do things actually stand today? The Study provides some interesting numbers:

  • Effective OECD retirement ages since 1970: the average worker exit ages were 68 in 1970, bottomed out at 62 in 2000, and rose back up to 65 in 2018.
  • Effective retirement ages for Japan and France in 2018: age 70 for Japan, age 61 for France, compared to the OECD average of 65. Culture is a factor too!
  • Labour Participation Rates for age 65+ workers from 2000 to 2018 for selected countries: Japan 22% -> 23%, USA 12% -> 19%, Canada 5% -> 15%, Australia 5% -> 13%, UK 5% -> 10%, NL 2% -> 8%, France 1% -> 4%.
  • Post age 65 OECD life expectancy today: 18 years…..expected to increase to 23 years by 2060.
  • Average OECD ratio of people over age 65 to working population today: 31%.....expected to increase to 58% by 2060.
  • Poverty rates for age 65+ for selected countries: Australia 21%, USA 21%, Japan 20%, UK 14%, Canada 11%, France 4%, NL 3%.
  • Melbourne-Mercer Global Pension Index (MMGPI): overall country quality grade based on weighted ratings for retirement income system (RIS) adequacy, sustainability, and integrity in each country: A: Netherlands, B+: Australia, B: Canada, C+: France, UK, USA, D: Japan.

The numbers support the ‘rethinking retirement’ narrative. A key element of that narrative is that, regardless of a country’s official retirement age, effective retirement ages have been rising since 2000 and will likely continue to do so. Appropriately, the Study devotes a section to measures that will facilitate the needed move to longer working lives. Success drivers for this move include continuing education, new skills development, flexibility in work and compensation structures, and financial incentives in countries’ Pillar 1 pension plans to delay retirement.

Writing the Study section on rethinking the design of retirement income systems (RIS) was our contribution to the CS project. Here is a summary of that section.

Rethinking RIS Design – Micro Perspectives

The evolution of RIS design at the micro level is best understood through the work of four Nobel Laureates:

Prof. Harry Markowitz got us to rethink how to construct investment portfolios using Modern Portfolio Theory. If we can estimate the prospective return characteristics (i.e, expectations, variances, and co-variances) of individual securities, then we can construct ‘efficient’ portfolios offering a ‘frontier’ of the best reward/risk opportunities. If we can also specify our tolerance for risk-taking, then we can identify the ‘efficient’ portfolio that best meets our needs.

Prof. Robert Merton argued that this ‘optimization’ philosophy should be stretched through people’s  lifetimes. If people wanted to maintain their standard of living after they stopped working, how much would they need to save during their working lives? That depends on: the investment return their savings earn, how long they work, how long they live after retiring, and on how their tolerance for risk bearing and their consumption patterns change as they age. Depending on how these four questions are answered, the required working life savings rate can be anywhere between 5% and 40% of pay.

Prof. George Akerlof pointed out that micro economic theory assumes that buyers and sellers in any market had ‘symmetric information’ about the product or service they were buying and selling. If sellers know more about what they are selling than the buyers know about what they are buying, then buyers will pay too much for too little value. The market for investment management services is the largest ‘asymmetric information’ market in the world. Thus retirement savers will generally pay too high fees for the value they receive from investment managers, unless their savings are managed through fiduciary structures that level the informational playing field.

Prof. Daniel Kahneman and his collaborators showed that humans also exhibit logic and behavioral shortcomings leading to inconsistent investment decisions based on the prospects of relative and absolute financial gains and losses. This has led RIS designers to construct nudges and decision defaults to help guide workers and retirees towards decisions that are ‘right’ for them.

Rethinking RIS Design - Macro Perspectives     

Switching now from a micro to a macro perspective on RIS design, a big breakthrough came in 1994 with the World Bank’s report “Averting the Old-Age Crisis”. Its key message was that there is no single best ‘micro’ solution to RIS design. Instead, we should be thinking of the respective roles of government, employers, and individuals (or families) in creating sustainable lifetime post-work income streams. This will require integrative 3-pillar ‘systems’ thinking:

  • Pillar 1: A universal base pension provided by the state funded on a pay-go basis.
  • Pillar 2: A prefunded employment-based pension sponsored by the employer.
  • Pillar 3: Individual retirement savings programs provided by the financial services industry.

With the passage of 25 years since the Report was published, this is a good time to ask how the 3-Pillar macro RIS model is doing:

  • Pillar 1: The main issue here is sustainability, with demographics being destiny. Countries that have made adjustments as their populations continue to grow older (e.g., through raising the retirement age, changing contribution rates and/or benefits) continue to have sustainable Pillar 1 systems. Countries that have not made those adjustments face increasing difficulties.
  • Pillar 2: Traditional DB plans are becoming an extinct species (i.e., too expensive and risky for most employers). New, more sustainable versions are replacing them (e.g., shared-risk, target benefit, collective DC). Participation is expanding through auto-enrolment regimes and innovative distribution strategies. There is also growing innovation in the design and management of pension delivery organizations.
  • Pillar 3: The Akerlof and Kahneman predictions are playing out in practice, with many individuals implementing poor retirement finance and investment decisions poorly. These problems are best solved by moving these people into Pillar 2 arrangements with strong fiduciary umbrellas, effective lifecycle designs, and strong implementation infrastructures.

In short, much of the world needs material workforce coverage expansion into effective Pillar 2 retirement finance and investment arrangements.

Expanding Pillar 2 Coverage and Designing Decumulation ‘Back-Ends’

The simplest way to achieve this expansion goal is for governments to require employers to offer a qualifying Pillar 2 plan to their employees. Unsurprisingly, countries that achieve top scores in the Melbourne-Mercer Global Pension Index have had this requirement for decades (e.g., Netherlands, Denmark, Finland, Australia). The UK, Canada, and the USA have begun to use auto-enrolment to join this elite group. 

Designing effective decumulation ‘back-ends’ for DC arrangements is also becoming a priority, especially in Australia with the maturing of its almost 30yr-old ‘super system’. The elephant in the decumulation room is longevity risk: the risk of outliving one’s retirement savings. Conceptually, the solution is simple: the provision of longevity insurance through longevity risk pooling. Practically however, the devil has been in the details of making this work in practice. Workable solutions are slowly emerging. 

In Conclusion

The CS Study offers a useful historical context for rethinking retirement as we enter the third decade of the 21st Century. Demographics are indeed destiny, and longer working lives will be an essential part of our lives in the coming decades. At the same time, academia has provided us with a solid conceptual framework for redesigning retirement income systems so that they become both more sustainable, and more effective in providing the means needed to finance life after work. The challenge now is to transform that conceptual framework into operational structures that actually do this.

Keith Ambachtsheer


The full “Rethinking Retirement” Study by Credit Suisse can be accessed HERE. In a related development, Canada’s National Institute on Ageing (NIA) has just released a study titled “Improving Canada’s Retirement Income System: Setting Priorities”. I was a co-author of this study. It can be accessed HERE.


KPA Advisory Services is pleased to share this edition of The Ambachtsheer Letter with all readers; if you wish to become a KPA Advisory Client/gain access to ALL Letters, please see the Services page on our website.

The information herein has been obtained from sources which we believe to be reliable, but do not guarantee its accuracy or completeness. 

Advisory Service clients have access to full issues of the Ambachtsheer Letter.

Become an Advisory Service Client
or Login
Back to Top