Editor’s Note: This commentary was written before the April 6, 2020 announcement by the Actuarial Standards Board of the decision to delay the implementation for the new commuted value standards to ‘not earlier than December 1, 2020’. However, early implementation of the standard for target pension arrangements is still permitted.
“I am so smart! S-M-R-T… I mean S-M-A-R-T!”
For years I have argued that the actuarial profession should not be in the business of setting standards for commuted values (I have also argued that we shouldn’t be in the business of setting standards for ‘annuity proxies’ for similar reason, but that is a rant for another day).
Back up – what is a commuted value?
Commuted values are the hypothetical lump-sum value of a series of monthly/annual pension payments payable for a lifetime (or a joint lifetime). In the good old days, if you were promised a pension then the pension plan delivered those payments when you finally retired. But in the 1980s, as pension reform took hold in Canada, legislators decided that a member of a pension plan should be able to terminate their pension contract with the pension fund at the time their employment contract was terminated with their employer which was often the pension plan sponsor. This sounds fantastic in theory – someone no longer connected to the employer shouldn’t have to remain connected to the pension plan sponsored by that employer.
However, as legislation was being drafted, governments came to realize that they needed to tell sponsors how to calculate commuted values ‘in the prescribed manner’. Somehow in Canada, the Canadian Institute of Actuaries was drawn into the project. After all, who other than actuaries are best suited to figure out the answer to such a complicated question?
Setting the Standard
What could go wrong in letting someone transfer the fair value of their future monthly pension to an RRSP or another employer’s pension plan? Well the problem starts from the beginning with the word fair. The value of a series of monthly payments payable for someone’s lifetime, to state the obvious, varies significantly on how long that lifetime will be, which is correlated (but not precisely) to personal health combined with family medical history. Commuted values can also vary significantly based upon marital status and the date of early retirement where plans offer ‘subsidies’ of one sort or another. And the mother of all political hot potatoes – gender is a big factor too. Finally, you need to have some idea of investment returns for the next 30, 50, or 70 years.
So, when the first commuted value standard was created in 1986 the actuaries did their best – an interest rate taken from government bonds and an assumption that after 15 years the most reasonable estimate would be an investment return of 6% per annum. We should spend some time laughing at the idea that in the absence of better data we were happy to assume that sooner or later bond yields would return to some sort of average – mean reversion at its finest. Governments bit the bullet and decided that gender would not be considered and as a result for almost all pension plans females get something less than fair and males get the reward from woke public policy. Except in Quebec where some smart folks figured out that women would need that extra cash because they will probably live longer and came to a different conclusion on fair.
I have lost track of how many updates the actuarial profession has made to the standard. Each time around there are many voices at the table. Many regulators like systems that produce ‘high’ values since their goal is to protect the members while plan sponsors are seen to have endlessly deep pockets to pay whatever it takes to be sure we aren’t being ‘unfair’. Actuaries that work for plan sponsors, investment firms, or sell annuities all have views – not always the same view. Already there are complaints that the new standards are unnecessarily complex and, in some situations, don’t make sense. The Actuarial Standards Board, already fielding complaints about system programming changes and an absence of education material with examples, is considering delay the effective date of the standard from August 1st to later in 2020 – but as far as I can tell, 2020 for sure.
At this point in my commentary you are probably getting a sense of why I think it is a fool’s errand to try to solve the question of what a fair commuted value looks like.
In the latest round there have again been changes to try to be more accurate in calculating the fair value that is unknown to start with. Dean has spelled out the details last month in his commentary on the standard.
Of interest to the profession, regulators, as well as sponsors and members of target benefit plans (now called ‘target pension arrangements’) are the new rules around plans where the sponsor does not commit to guaranteeing the benefits but rather commits to a rate of contribution. I won’t go into my thoughts on those new standards here, but I will simply acknowledge that this update is a very thoughtful reflection on the difference in value between different types of pension guarantees. The actuarial profession has done something good here. They have correctly put into the hands of plan administrators (i.e. boards of trustees) some of the decisions about what is ‘fair’.
Interestingly, the legislators in Ontario weren’t comfortable with the uncertainty of what the new standards would say for Target Pension Arrangements and have already changed the rules so that the single-employer commuted value rules for ‘guaranteed pensions’ still apply to target benefit plans (at least for a while longer). Gotcha – government is finally admitting that ‘fair’ is a political decision and not an actuarial calculation. It will be interesting to see if other provinces follow Ontario’s lead.
Perhaps it won’t be long now before the actuarial profession gets out the business of setting these standards and just leaves legislators and regulators with the burden to decide what is fair – a burden that they should have owned from the beginning.