Veritas propter investigationem [Truth through research]
TitleWhat Makes a Better Annuity? (2009)
AuthorsJason Scott
 John Watson
 Wei-Yin Hu
InstitutionPension Research Council
 Longevity products
CountryUnited States
Date Published2009
Date posted on PR16 Sep 2014
Scott, J Watson, J Hu, W, (2009). What Makes a Better Annuity? (2009) Pension Research Council,

PensionReforms’ summary and comments

The so-called ‘annuity puzzle’ has been widely discussed.  Despite the superficial attraction of a life-time-guaranteed income stream, retirees just aren’t buying annuities unless they have to.  That’s a global ‘problem’.  This 2009 report provides a theoretical analysis of what providers might have to do if they want to sell more annuities.


“In a world with costly and limited annuity products, we investigate what types of new annuity products could improve annuity market participation and increase individual welfare.”


The answer seems to be ‘late-life’ products where the annuity’s starting date is some way into retirement, seemingly because they can “…offer a large price discount relative to their financial market analogues.”


What this means is that the annual income from a given purchase price that starts, say, 15 years after retirement is much larger than an immediate annuity.


“Given access to a complete market, we find all individuals only purchase annuity contracts with a significant time gap between purchase and pay-out.  At a minimum, enough time must pass between purchase and pay-out to build up a mortality discount sufficient to overcome the cost of creating the contract.”


The report concludes that “… annuity markets will never offer actuarially fair pricing nor will they likely approach financial markets in terms of flexibility.”  So the offerings must change and “late-life pay-outs” seem to both increase the welfare of purchasers and increase participation rates in annuity contracts (good for providers as well).


“Given complete annuity markets, annuity purchases should only occur when the mortality discount is sufficient to overcome any annuity costs.  With positive costs, an annuity purchase thus requires a gap between the purchase date and the pay-out date.”


The report suggests that not having that time gap is a “fundamental inefficiency”.  If that were remedied, “…innovative new products may ultimately supplant many incumbent annuity products.”


And, the report suggests, there are “ample opportunities for improvement.”


PensionReforms thinks there is perhaps a different explanation for the superficial attraction of “late-life pay-outs” that is grounded in the relative unattractiveness of immediate annuities.  The bargain looks poor for immediate annuities because they are priced to:

-           reflect the mortality risk (or living longer than expected);

-           insure against the investment risk (or earning less than expected);

-           pay for administration costs;

-           pay for the set-up costs, including introductory commissions;

-           deliver a profit to the provider.


Then there is the not insubstantial risk of provider failure.


A ‘late-life pay-out’ essentially disguises these costs by building in a long lead-time before the insurer has to start payments.  The lottery element (of making it to the start date) and the reducing real value of money also help.


PensionReforms thinks there is probably a role for ‘late-life pay-outs’ but suggests that many or the same cost/risk issues are still present and even magnified in ways that may help providers’ bottom lines but, necessarily, at the expense of the purchasers.  There is a need for great transparency and regulatory supervision if the start date of the annuity is after the purchase date, especially if that is some years after that date.  (File size 285 KB; 41 pp) 698