Couple of questions on acounting for pension cost. Feel free to answer only one if that's all time/knowledge/inclination allows. All opinions, whether absolute fact or uncertain guess (but do qualify your response accordingly), will be greatly appreciated.
1. In the examples (in chapter 19), all the elements of the pension cost (and/or reconcilliation of balance sheet entries) are estimates fixed at the start of the year.
Do these ever change when preparing the accounts at the end of the period to take into account what actually happened, eg if we assume benefits are 1 million and they are actually 1.4 million? I know the expected return on assets is changed to actual return less gain on assets, but are any other items changed?
Does this change depending on whether it is FAS 87, FRS 17 or IAS 19?
I am answering these questions without having ch 19 in front of me. The pension cost provided at the end of the year incorporates the expected ROA. The G/L incorporating the under/over performance on the asset retrun is swept up in the adjustments to the balance sheet item. Actual benefits and contributions will also be swept up in this but the extent to which differences to the expected amounts would affect the pension expense would normally be epxected to be immaterial (as they are multiplied by a discount rate). However - if it looks like they might be material - a conversation with the client/auditor would probably be the senssble thing to do -as you should remember that they are the client's disclosures - not yours.
2. FAS 87 says that the discount rate "should reflect the yields available to purchase annuities and have regard to returns on high quality fixed-interest (or index linked) corporate bonds of a suitable term."
On initial reading, I assumed this meant taking the yields on bonds less allowances for profit/expenses/contingency loadings etc since this is the return you can expect if you buy annuities as an investment.
However, it seems that discount rate is simply taken to be the yield only.
Which is it?
I would expect the discussion mainly to be based around a suitable bond yield rather than implied annuity rates - as they are more readily available. Yields can be adjusted for any number of features depending on a credible justification e.g. if you choose an index with a 15 year term but your liabilities have a term of 20 years there may be scope to make an adjustment to the yield based on this term differential. Maybe the curve describing the yields making up the index is upwards sloping with term so your client could argue to their auditor that the discount rate should be higher to reflect an extrapolated higher yield. Again - it often boils down to judgement and negotiaitions with the auditor in practice, based around the framework as laid down by the standards board.
3. My experience with accounts covers IAS 19 only which takes liabilities and expenses as positive and assets and income as negative. (Personally I like this approach as it is unambiguous.)
As far as I can tell, FAS 87 doesn't do this.
Eg, it seems to show both assets and liabilities as positive.
It also seems to be inconsistent (unless there is a typo) as on page 10 the return on assets is positive, and on page 40 the return on assets is shown as negative where the return is an income in both cases.
What is the sign convention, or is there no sign convention?
Is there a sign convention for FRS 17?
I don't know whether there is a convention - i usually just try to prepar the figures mechanically in line with the previous set! Be vigilant is all i can advise 
4. If IAS 19 is required for all companies trading in UK/EU, does anyone use FRS17? I'm not sure if "trading" here means operating (eg selling goods/services) or the trading of listed stock on an equity market.
If it's possible, are there any advantages to choosing FRS17 over IAS 19?
I think that IAS19 is only required for listed companies (for ease of comparison in published accounts. Often times a US listed parent company will commission FAS figures for their UK concern in their group accounts, but the UK company will still need FRS17 figures for their statutory accounts filed in the UK. I don't know why they wouldn't have IAS19 figures for this purpose but I suspect it is becuase IAS19 may be more onerous? Or it may be out of habit?
5. And finally this is specific to tutors. I think all references in chapter 19 to "worked examples in section 5" should be changed to ".... section 6".
Also, in my personal opinion, I thought the chapter would have flowed better starting with IAS 19 and then comparing the others to this, rather than the other way around starting with FAS 87, since I prefer to start with what I know. Of course, I understand that my views may not represent the average student (especially since I'm not from the UK although I expect UK students to be more familiar with IAS19/FRS17 than FAS87) but this may be something to consider.
Click to expand...
