deferred DB pension Revaluation. Are the 5%, 2.5% limits compounded? Also partial years.

optoutDB

As part of my considering a Transfer out of my private company DB pension (which is unlikely to happen). I have for the first time looked in detail at my deferred pension. 

My service ended June 2010.
My NRD is 1/6/2033  (age 65).

Compounded Cap?

Given the current rate of inflation my attention was drawn to the 5% and 2.5% maximum revaluation limits for various tranches ( 95% of my pension is subject to the 5% max).  When I read the details of my pension (as described in the Transfer value documentation) I assumed the 5% cap was in every year and the values were recalculated each year.  But now I have seen the word cumulative mentioned in several threads on the subject. 

So, I'm hoping to hear that indeed the caps are applied on a cummulative basis over the whole period. As it will take a lot of  >5% inflation years to bring the average from about 2% to over 5%.


Partial Years

The other question I have is about calculation dates and partial years.
I dug out my first yearly pension statement for the first year after I left and it tells me:
Pension at date of leaving: £11305
Pension increased to 5/4/2011: £11305.
That was a 3.1% inflation year so it looks like I've been done out of 10 months of inflation (about £260 at 2011 prices). 

I'm wondering if that is normal? 
Also, maybe I will recover it at the rear end, eg if I choose July to retire (early) will I get a full years revaluation then [seems unlikely]. 


thanks,

zagfles

The revaluation cap is on compounded years, so eg the odd year of high inflation year will not be capped as long as the overall inflation over the period of deferment doesn't exceed the cap.

Partial years usually are lost. You only get whole years, and which years can make a difference (it changes at the turn of the year so the pension can rise and fall on 1st Jan).

Those are the statutory rules, the scheme could be more generous, but I don't think most are (some will use RPI and so be better than the CPI used in revaluation orders).

See this thread https://forums.moneysavingexpert.com/discussion/5962314/rules-on-using-occupational-pensions-revaluation-orders/p1

optoutDB

zagfles said:

See this thread https://forums.moneysavingexpert.com/discussion/5962314/rules-on-using-occupational-pensions-revaluation-orders/p1

I haven't fully worked through the problematic case there, and I'd rather see the rules themselves than the result when applied to his dates (I'll get to it eventually).  In my case the inflation rates in the year I left work (3.1%) and retire are likely to be similar, so any flipping will be between similar values. 

What matters to me is whether months get truncated at either the front or tail end, or both.  Or whether the calculation is pro-rata'd   ie if I take my pension in Aug 2028 (28.2 years after leaving) will I get 28.2 years of uplift, or 28 years, or  27 years.

It appears the values were flipping because the method doesn't pro rata the partial years at either end, eg the sensible calculation would give him 5/12 of 10.9% for the period 1 Aug to 31 Dec 1990.  Or better still just publish and use monthly uplifts.  The method appears to me to have been devised to allow the use of a look up table rather then doing the proper calculations. 

Marcon

Statutory revaluation is based on whole years, so you haven't been 'done out' of anything to which you are entitled.

Chapter and verse: see https://www.barnett-waddingham.co.uk/comment-insight/blog/revaluation-for-early-leavers/

optoutDB

Marcon said:

Chapter and verse: see ......

That looks promising then, as long as the method used there applies to any chosen retirement date and not just NRD.  I can retire in a July and get my full uplift (because the integer years will be close to actual years since leaving). 

And just because that's the specified method of calculation, it doesn't change my opinion that people are being done out of between 0 and 12 months of uplift. 

zagfles

optoutDB said:

zagfles said:

See this thread https://forums.moneysavingexpert.com/discussion/5962314/rules-on-using-occupational-pensions-revaluation-orders/p1

I haven't fully worked through the problematic case there, and I'd rather see the rules themselves than the result when applied to his dates (I'll get to it eventually).  In my case the inflation rates in the year I left work (3.1%) and retire are likely to be similar, so any flipping will be between similar values. 

What matters to me is whether months get truncated at either the front or tail end, or both.  Or whether the calculation is pro-rata'd   ie if I take my pension in Aug 2028 (28.2 years after leaving) will I get 28.2 years of uplift, or 28 years, or  27 years.

It appears the values were flipping because the method doesn't pro rata the partial years at either end, eg the sensible calculation would give him 5/12 of 10.9% for the period 1 Aug to 31 Dec 1990.  Or better still just publish and use monthly uplifts.  The method appears to me to have been devised to allow the use of a look up table rather then doing the proper calculations. 

That's exactly what it does. I explained the rules in the linked thread. It uses whole years of deferment, back from the previous year. If you start taking the pension later in the calendar year to when you went you left the scheme, you get inflation for the year you left until the year before you start taking the pension, inclusive. If you start taking the pension earlier in the calendar year to when you left it's from the year after you left till the year before you start taking the pension.

If you have a GMP element that works completely differently, it's on the number of tax years since you left and is usually a fixed rate depending when you left, the above links explain.

zagfles

optoutDB said:

Marcon said:

Chapter and verse: see ......

That looks promising then, as long as the method used there applies to any chosen retirement date and not just NRD.  I can retire in a July and get my full uplift (because the integer years will be close to actual years since leaving). 

And just because that's the specified method of calculation, it doesn't change my opinion that people are being done out of between 0 and 12 months of uplift. 

That isn't opinion, it's fact. Particularly as most definitions of "final salary" mean the salary over the last 12 months, and for CARE schemes it's what you earned over the last year. So effectively final salary is pro-rated, but inflation uplift isn't. Plus most schemes will pro-rate the first inflation increase when you start taking it (eg if the first annual increase is 6 months after you start taking it you get half the increase).

optoutDB

zagfles said:

......  Plus most schemes will pro-rate the first inflation increase when you start taking it (eg if the first annual increase is 6 months after you start taking it you get half the increase).

I should hope so !   

I will obviously have to study all the methods used by my scheme before I choose my date to start drawing this pension, so far it looks like a July.

snowlaser

optoutDB said:

Marcon said:

Chapter and verse: see ......

That looks promising then, as long as the method used there applies to any chosen retirement date and not just NRD.  I can retire in a July and get my full uplift (because the integer years will be close to actual years since leaving). 

And just because that's the specified method of calculation, it doesn't change my opinion that people are being done out of between 0 and 12 months of uplift. 

Assuming your scheme is indeed statutory revaluation it's based upon complete years, with the 5% / 2.5% caps compounded over the whole period (not year by year).

If you retire early the scheme should ensure that you get at least the revaluation from date of leaving to your chosen retirement date - though the pension will likely be actuarially reduced for early payment (often somewhere between 3% and 5% for each year earlier than NRD, but your scheme will have their own way of doing it).

I can understand how you feel may lose out on up to a year's revaluation that way - I guess they have to draw the line somewhere and could do it at days, weeks, months, years or something even longer and have gone for complete years for simplicity and, no doubt, to also keep the costs down. 

optoutDB

It's possible that what you miss out in the revaluation during the deferred period, comes back to you in the pension payment phase (notwithstanding the problem of different inflation numbers at the start vs the end) .

Nearer the time I will obtain the details of how the indexing is done when the pension is in payment.  

eg If I choose July to take the pension, to minimise partial year losses at the final revaluation, will I then lose some months of indexing up to some date  ( 1st Jan, 5 April, 8 Jun ).  

artyboy

zagfles said:

The revaluation cap is on compounded years, so eg the odd year of high inflation year will not be capped as long as the overall inflation over the period of deferment doesn't exceed the cap.

Partial years usually are lost. You only get whole years, and which years can make a difference (it changes at the turn of the year so the pension can rise and fall on 1st Jan).

Those are the statutory rules, the scheme could be more generous, but I don't think most are (some will use RPI and so be better than the CPI used in revaluation orders).

See this thread https://forums.moneysavingexpert.com/discussion/5962314/rules-on-using-occupational-pensions-revaluation-orders/p1

Interesting thread and one that I'm just getting my head around. I have a small DB pension with a deferment date of 15 September 2008 (a date that may be familiar to any finance professionals here!). Interestingly the scheme closed some 10 years before but benefits were preserved and it would have been based on my final salary at the 2008 date.

Revaluation is 5% capped, but CPI has been nowhere near that over the last 14 years. My policy does not make it at all clear that revaluation is done at an overall level between deferment and when I take it, which will be at least 20 years later.

However if that is the statutory rule, then a few years of high inflation could actually bump it up considerably! It's almost like having unused allowance from previous years that can be called on...