Q.
A company has a defined benefit pension plan for its employees. Which of the following changes in assumptions would most likely decrease a company’s reported pension expense? An increase in the expected:
A. retirement age. B. return on plan assets. C. growth rate of salaries.
“Correct” answer is B.
…but, isn’t an increase in retirement age, a change that could result in actuarial gains?
I initially clicked that answer too. But the formula for Net Pension Expense is basically service cost - expected return on assets. So if your ROI is more than you expected, your net pension expense goes down.
I am not arguing whether the answer is correct. My problem is that A is correct too…
increase in expected retirement age would lead to an increase in pension expense.
Why? As far as I understand, unless mortality assumptions go down by more than the increase in retirement age, the pension will have less years to cover in annuities.
and is B simple it is yaar
retirement age increases - you have to provide the defined benefits for longer - so the amount that you have to put aside in the liabilities increases. There is nothing to do with mortality assumptions out here.
it is an annuity payment that is due for a longer time …
I think I am missing something really big here 
Aren’t the defined benefits a pension? Do they provide anything before retirement?
you will learn about it more in Level II - not sure if things are discussed in this detail in the Level I curriculum (at least it was not when I did it way back). But the answer above is right. The benefits are for employees who are part of the plan - for in retirement - but there is an obligation (liab) that gets built up. When the retirement age increases - the pension obligation is required for longer - and as a result the periodic pension expense increases.
When the ERPA (Expected Return on Plan Assets) increases - the Pension Expense is reduced.
This is a very interesting topic!
Sure, the liabilities would be recorded for longer under a later retirement age but wouldn’t you have to adjust the pension obligation accordingly? E.g. decrease the periodic contributions to account for the increase in periods?
And what about the fact that the employee is going to need the annuities for less time - since the upper limit for age does not change?
EDIT : I assume the obligation goes up with years in service but how are we to know if the amount goes up by more than the reduction in amount by the fact that the benefit is going to be needed for less time?
EDIT No2 : I looked it up and came up with a model actuary calculator : http://www.aliactuary.com/retplan.asp
If you tinker with the retirement age, holding everything else constant, the amount that you need to save per year goes down…
Your calculator…doesn’t make any sense since it is a regular retirement calculator. A defined benefit plan defines the monthly benefits by averaging the last salaried periods. There are guaranteed benefits that are paid to surviving dependents and minimum pay out amounts. The higher expected average income might from an actuarial point of view even outweight the lower expected pay out period. I wouldn’t go there. You might be right, but Answer B is simply "more " correct.
Post shows up thrice…sorry for this!
Edit: hm…
Dear I-will-pass
Thank you for the response. It is true that B is simply “more” correct hence for test purposes there is no real issue. However this still looks like a cheap trap. Most evidence I found online is consistent with the “greater retirement age usually means less current obligation” and to a complete novice like me it really does seem far more intuitive.
Regarding the calculator, I can obviously not vouch for its credibility but you will get the same result under the defined benefit one here : http://www.aliactuary.com/maxcalc.asp
I believe (but I am not an actuary), that the defined benefit obligation disregards mortality tables/premature death/longevity risks. This explains why the calculator doesn’t need any information regarding health status, gender etc.
This would explain why there wouldn’t be any actuarial gains in this case.
However, the concept is far more complex and since I am not an expert and you aren’t either I would propose that you go with your own suggestion, as shown here: http://www.analystforum.com/forums/cfa-forums/cfa-level-i-forum/91316517
It cannot be C; A: the question never mentioned any mortality linked changes. Furthermore you cant be expected to have indepth actuarial expertise.
Cross fingers, chose B.
Don’t lose to much sleep over this 
never expected that ;practical question in mock exam on pension expense 
I got really angry at this question too, to me it’s clear that both answers are correct.
Defined benefit gives a percentage of your salary in pension, that is what the company has to put aside for you. If the retirment age is pushed to later time, the have to set aside money for a shorter time when the pension actually needs to be paid out.
If it was a defined contribution pension the contribution would be for a longer time period, but since it is a benefit pension, it means providing a percentage of your salary in pension, and later pension means less money needed to pay out. Look at Greece where they up the retirment age to save money, it’s not a complicated relationship to understand… …shitty Kaplan material.