Civil Service pension: alpha v. partnership & voluntary topup

hugheskevi

chiefnoodle said:

hugheskevi said:

I'm pretty impressed with the modelling, and the speed at which you have come to accurate conclusions - it usually takes people a lot longer, if they get there at all.

There are lots of things that can be done to refine this (eg using exact State Pension ages rather than 68 throughout) but they won't meaningfully change the conclusions.

Thank you! That is quite a compliment from the MSE forum pension expert. I tried listening to their podcast, but it was one-sided to the point of being an advert, and I quickly realised I wasn't going to get any thoughtful critical discussion there. Then I spent a while going through the CV pension website. It would have been much quicker if they could have included some plots like the one I made on their "Alpha or Partnership" page to give some quick "rule of thumb" intuition about how the two compare. And yes, plenty of scope to improve my model, but as you say I decided to only include factors which would change the outcome.

I always think that unless you give a fair assessment and highlight weak points as well as strengths then anyone who wants to do some critical thinking and come to their own conclusions won't listen to you. I've just listened to that podcast, and there were some questionable things said (I made a list of them below).

I acknowledge that opting out of a DB scheme for a DC scheme is a big decision. I always say that members should choose alpha unless there is a compelling reason to choose Partnership, and also say that alpha is more generous for older employees. If an individual is motivated enough to consider it in detail as you have done, they may well find a compelling reason, if not, alpha will give a decent outcome in all scenarios and they won't go far wrong. The problem is that if someone just does something without understanding it, I've seen some people still in Partnership at ripe old ages because their mates said it was great to switch when they joined 15 years earlier.

• Starting off by describing the Civil Service scheme as a Relief-at-Source scheme when alpha is a Net Pay scheme for tax relief (Partnership is RAS)
• Saying contributions are 'kind of matched' - again, why confuse people when you can give an accurate description
• Describing Partnership members as usually having a lack of understanding. Some of the most knowledgable employees are Partnership members - in the past it may have been due to Annual Allowance and the tapered Annual Allowance, there are very good reasons for classic members to break their final salary link as they get closer to retirement if they think their salary will go up by less than inflation (and that even applies to Premium and Classic Plus members now), younger members may well prefer to put 3% into Partnership (to get full matching) and have higher income when they most need it - which as you have shown may well be a better outcome anyhow, and some members who simply can't afford the standard member contribution can move into Partnership and contribute nothing yet still get a good employer contribution. To simply dismiss Partnership members as generally lacking understanding shows a complete lack of understanding of just how many members will be better off in Partnership in particular circumstances, even if you ignore the potential for younger members to end up better off from a spell in Partnership before moving to alpha as they age.
• A complete lack of knowledge of the employer contribution rate to Partnership - how can you dismiss it if you don't even know the employer contribution rate?
• An obsession with quoting the (then) alpha employer contribution rate of 27% - which is irrelevant. The alpha scheme hasn't changed since it was introduced in 2015. The employer contribution rate then was 21.1. From next week it will be 28.97%. The scheme hasn't changed, so how has it become almost 50% more generous to employees? There is even a perfectly good figure calculated by the Scheme Actuary which shows the cost of newly accrued pension less the average employee contribution - it is 23.6% (changes in the scheme discount rate and notional past service deficit cause the change over time). As shown in this thread, just quoting the average is very misleading given the significant age bias, but at least if you do quote an average figure you should make it the most appropriate figure.
• Saying typical employer contribution rates are 4%. Sure, statutory minimum schemes might be down at that level, but why not compare to a decent scheme, one meeting the PLSA standards, for example? If you own a Ferrari, do you persuade many people how good your car is by comparing it to a Ford Fiesta?
• The portrayal of restrictions on access as a bonus of a pension. Really? Perhaps in niche cases, such as benefit recipients (although I would prefer my assets to be liquid if I am in that much need), or those with control issues, but for most people you prefer liquid and trade liquidity for return (eg instant access vs fixed rate savings accounts). 
• Going into the FSCS guarantee on bank savings - that really felt irrelevant. For a start, just put money in different institutions. They discussed that bank collapses have happened, but didn't mention that depositors have been fully protected beyond FSCS levels. My main objection is that this podcast is clearly pitched at a very low level of understanding, and I wonder how many low-paid Civil Servants really have £85,000+ lying around in a bank account?
• Death benefits being portrayed as a key strength of alpha. The survivor receives 37.5% of the pension payable from Normal Pension age and taxable. That really isn't great. The survivor of a DC holder receives 100% and can be tax free depending on age of member death. They argue that if the DC pot has been mostly run down the survivor will inherit very little, but don't mention that if the DB holder dies shortly after retiring a large proportion of the value will be lost. That section really lacked balance.
• The emphasis on 'guaranteed' benefits, with no reference to changes to State Pension age which affect Normal Pension age, nor the potential for change to the indexation level, as with the RPI/CPI switch, both of which would affect accrued benefits.
• I did like the line "whereas the government has taken the view, well we're not going to do that and we want the best for the people that work for us and we're going to keep this DB scheme alive." - conveniently ignoring that alpha replaced the nuvos scheme, and alpha is worse in every way except in the event of ill-health or death. Arguably Premium, the scheme nuvos replaced, was better than nuvos although it is less clearcut, and nuvos probably is better than Premium for older members. And no mention at all about how difficult in terms of both the deficit and national debt it would be to change from an unfunded scheme to a funded scheme, which provides a powerful incentive for the government to keep unfunded DB pensions.

chiefnoodle

hugheskevi said:

chiefnoodle said:

hugheskevi said:

I'm pretty impressed with the modelling, and the speed at which you have come to accurate conclusions - it usually takes people a lot longer, if they get there at all.

There are lots of things that can be done to refine this (eg using exact State Pension ages rather than 68 throughout) but they won't meaningfully change the conclusions.

Thank you! That is quite a compliment from the MSE forum pension expert. I tried listening to their podcast, but it was one-sided to the point of being an advert, and I quickly realised I wasn't going to get any thoughtful critical discussion there. Then I spent a while going through the CV pension website. It would have been much quicker if they could have included some plots like the one I made on their "Alpha or Partnership" page to give some quick "rule of thumb" intuition about how the two compare. And yes, plenty of scope to improve my model, but as you say I decided to only include factors which would change the outcome.

I always think that unless you give a fair assessment and highlight weak points as well as strengths then anyone who wants to do some critical thinking and come to their own conclusions won't listen to you. I've just listened to that podcast, and there were some questionable things said (I made a list of them below).

100%. I don't have the wealth of knowledge about CS pensions you have to evaluate each claim they made, yet just from the tone of the podcast I picked up suspicions that it wouldn't be fully reliable - which is what motivated me to look at it in more detail. Your careful unpacking of their assertions has confirmed that :-) Thank you!

Andy_L

chiefnoodle said:

If I were you, I'd consider reducing your work hours slightly (by a fifteenth). If you are now 37h/week, consider moving to 34.5h/week. Your salary will go down pro-rata to £55,945, bringing you within the 5.45% band. While your gross salary will drop by £4k, your net salary will drop by less than £3k (if I've understood the calculations correctly). So you effectively get those hours off at a discounted rate.

Are contributions rates for part timers based on the reduced salary or the full time rate?

chiefnoodle

Andy_L said:

Are contributions rates for part timers based on the reduced salary or the full time rate?

I believe based on your actual part-time salary, and assume someone would have corrected me if I'd got that wrong...

Details about what items of pay count towards your "pensionable pay" are here.

r6mile

chiefnoodle said:

Andy_L said:

Are contributions rates for part timers based on the reduced salary or the full time rate?

I believe based on your actual part-time salary, and assume someone would have corrected me if I'd got that wrong...

Details about what items of pay count towards your "pensionable pay" are here.

They are based on actual salary.

amanda1024

chiefnoodle said:

Here is the plot for your band: 

I wish I’d had this ~14 years ago when I joined the civil service! Prevailing wisdom seemed to be that defined benefit was the gold standard for pensions, but from this I’d have been a lot better off choosing the partnership scheme. I’m now in my mid-thirties and in the 7.35% contributions band, so it looks like it would still be worthwhile switching over to partnership for the next 15 years or so?

What’s the reason for the kinks in the curve from age 30 to 45?

chiefnoodle

amanda1024 said:

I wish I’d had this ~14 years ago when I joined the civil service! Prevailing wisdom seemed to be that defined benefit was the gold standard for pensions, but from this I’d have been a lot better off choosing the partnership scheme. I’m now in my mid-thirties and in the 7.35% contributions band, so it looks like it would still be worthwhile switching over to partnership for the next 15 years or so?

What’s the reason for the kinks in the curve from age 30 to 45?

TBH I'm not sure what the CS pension plans were 14 years ago. I understand they were more favourable then, so it is possible that you made the right decision. I'll leave that for forumites who know more history to weigh in on, this plot only compares Alpha with Partnership which are the current plans.

Purely from the perspective of expected returns yes it looks like it might be worth considering the switch. But there are other questions as well such as certainty which have been previously brought up in this thread which might swing things. If I were you, if all your pension until now is in DB, it might be worth paying a few years into DC - that way you will have hedged your bets - and then going back to DB in 10-15 years.

The kinks in the plot are due to fluctuations in the employer contribution to the Partnership scheme. See here under "What will my employer contribute in my partnership pension account?" - the rates jump at ages 31, 36, 41 and 46, and this is reflected in the plot.

hugheskevi

amanda1024 said:

chiefnoodle said:

Here is the plot for your band: 

I wish I’d had this ~14 years ago when I joined the civil service! Prevailing wisdom seemed to be that defined benefit was the gold standard for pensions, but from this I’d have been a lot better off choosing the partnership scheme. I’m now in my mid-thirties and in the 7.35% contributions band, so it looks like it would still be worthwhile switching over to partnership for the next 15 years or so?

What’s the reason for the kinks in the curve from age 30 to 45?

14 years ago a new member would go into the nuvos scheme. Nuvos is extremely similar to alpha, except that it has a fixed normal pension age of 65 whereas alpha is linked to State Pension age (so 68+ for someone now in their  mid 30s). You have now been returned to nuvos for the period up to 1 April 2022 due to the 2015 Remedy following the McCloud judgment, after which you are in alpha. Deferred revaluation of nuvos benefits is the same as in-service revaluation. Deferred revaluation of alpha benefits is marginally more generous than in-service, as in-service revaluation can be negative if there is deflation whereas deferred revaluation has a floor of 0%.

Paying a contribution rate of 7.35% suggests a higher rate taxpayer. It may be attractive to put that higher rate income into a DC AVC or personal pension to avoid 40% income tax, and possibly regain entitlement to full Child Benefit.

Remember if considering switching to Partnership to review the death-in-service, dependent's benefits and ill-health benefits, as they are all very different under Partnership and someone with a decent amount of service in the DB scheme would potentially lose a lot if these things occurred after they moved to Partnership.

nottinghammer

Great thread! I had come to a similar conclusion using a somewhat similar method (see https://forums.moneysavingexpert.com/discussion/6421388/civil-service-pensions-why-i-plan-to-pick-partnership-over-alpha), but the calculations here are certainly more sophisticated. 

Another way to look at this would be, for Partnership, to assume a drawdown of the pension pot from age 68, withdrawing the amount that Alpha would provide. The remainder of the pension pot remains invested, and grows with the assumed annual real yield. For someone today in their mid-40ies with a 60k CS salary and assuming a 2% real yield every year, the pension pot would then last for about 19 years. 

All that said, and having picked Partnership myself, I still think that even for younger civil servants the risk is considerably higher and probably too high for most. If the CS pension is expected to be a large part of retirement income (i.e., no other sizeable pensions, inheritance, large savings) then having most of the retirement pot invested in the global stock market is a significant risk (just think of sequence of return risk). And if the Partnership pot is invested more conservatively (large bond share) then one should not assume a CPI+3% long-term yield. Long-term inflation-linked gilts give 1% real yield at the moment (and that used to be much lower before the interest rate rise in 2022). And even if one had large other savings and followed a medium-risk 60/40 equity/bond allocation, then picking Alpha and having the Alpha pot mentally in the 40% bond share could make more sense. 

N

hugheskevi

nottinghammer said:

All that said, and having picked Partnership myself, I still think that even for younger civil servants the risk is considerably higher and probably too high for most. If the CS pension is expected to be a large part of retirement income (i.e., no other sizeable pensions, inheritance, large savings) then having most of the retirement pot invested in the global stock market is a significant risk (just think of sequence of return risk). And if the Partnership pot is invested more conservatively (large bond share) then one should not assume a CPI+3% long-term yield. Long-term inflation-linked gilts give 1% real yield at the moment (and that used to be much lower before the interest rate rise in 2022). And even if one had large other savings and followed a medium-risk 60/40 equity/bond allocation, then picking Alpha and having the Alpha pot mentally in the 40% bond share could make more sense. 

I think there are two big issues to consider for young people (or indeed anyone) considering Partnership vs alpha that go beyond the comparison of expected risk / return:

(1) Sub-game perfection - or to check that you would actually see through whatever your plan at the outset would be. For example, let's say you model a 60/40 portfolio you plan to hold throughout. Look at the pot size in the year before planned retirement and consider investment volatility. Would you be confident enough to stay heavily invested in equities given the size of your pot? If the pot is a relatively small part of retirement resources that would be fine, if it is a very significant proportion would you want to reduce the risk? If so, the plan is not sub-game perfect.

(2) Life and policy changes - it is all very well planning that, for example, 10 years in Partnership then moving to alpha would be optimal. But in practice you may well move to a private sector employer during those 10 years, and access to a DB pension would vanish. Whereas access to DC pensions are likely to be on offer regardless of employer. The same applies to future policy change - who knows what will happen, but the probability of having access to a DB pension in the future will be lower than having access to a DC pension. 

These things might well bias a decision to err on the side of caution when it comes to planning the alpha / Partnership age optimality. Even so, there is a lot of value in Partnership for young people, but maybe don't be too greedy and plan to switch back to alpha a bit sooner than a straightforward comparison might suggest.