USS - General discussion

Barralad77

My Annual Allowance figure looks to be correct; how are you calculating it?

My worry is I have other pensions and these will be taken after I’ve taken the USS pension. I’m assuming they check with (in this case) USS to see how much of my LSA I’ve used up. The last thing I want is for USS to tell them that I’ve already used it all up. The LSA figure on the statements is clearly wrong, and I’ve no idea what else they might get wrong in the future.

kermchem

For a defined benefit pension scheme, which the main part of USS is, the calculation of the annual allowance is not as simple as counting how much has been added by the employee and employer. See https://www.gov.uk/hmrc-internal-manuals/pensions-tax-manual/ptm053301

Barralad77

Good Grief! Thanks for the link but I’m at a loss how their calculations work. For example, ‘Example 3’ has someone earning an average of around £60k across the year, and where a year’s contribution gets them 1/80 of that salary as annuity (plus a 3 x salary LS). By my calculations 1/80 of 59,600 is £745 and the corresponding LS would therefore be 2,235. But the way it’s been calculated on the Govt site their pension has risen in value from £13,500 to £18,625, an increase of £5,125 (which as a % of the annual salary is around 8.5%, not the 1.25% I assume they get if they get 1/80 of their salary). So now I’m completely flummoxed. I calculate my AA as (salary/75 x 19) plus whatever has gone into the IB. That has always pretty much matched with the AA quoted on my statement. 

Edit: I’ve just used the AA calculator on My USS and the resulting figure tallies with my manual calculation method.

d6fs1l

Barralad77 said:

My Annual Allowance figure looks to be correct; how are you calculating it?

In general, to calculate the pension input of any defined benefit pension for the annual allowance, you measure (the difference in benefits accrued over the input period)*16 -- but NB the opening value of benefits is adjusted for CPI. 

For USS the multiplier used is actually 19 (16 for pension accrual plus the additional lump sum accrued). Again, though, you cannot just take the annual accrual and multiply it by 19 -- you have to adjust the opening value of accrued benefits for CPI before measuring the difference between that opening value and the closing value of accrued benefits at the year-end. The figure used for CPI is the familiar September figure used for official pension increases. For 24-25 this is 6.7% (I believe).

See for example the description of the calculation in step 4 of “How to find the opening value”: https://www.gov.uk/hmrc-internal-manuals/pensions-tax-manual/ptm053301

There is a clear M&G calculator available online (for example) which you can use for this purpose (option 2): https://www.mandg.com/pru/tools-calculators/defined-benefit-pension-input-tool/index.html

Obviously, to calculate your total pension input for USS if your salary is above the salary threshold, you also need to add the amount saved into the Investment Builder too.

For some reason, the USS description of the process (https://www.uss.co.uk/for-members/pension-tax/annual-allowance) and the worksheet it provides, linked on that page, does not include the CPI adjustment. I do not know why not. Past versions of the worksheet did include it -- https://www.imperial.ac.uk/media/imperial-college/administration-and-support-services/hr/public/pensions/Annual-Allowance-USS.pdf  (which I found by googling -- I could remember having seen the older version in the past, which is why I searched for it).

Maybe I am wrong, but if so, I'd like to understand why. I asked USS and got an answer which I felt was incorrect.

The reason it matters is because the large CPI figures for 23-24 and 24-25 would have a very significant impact on the DB pension inputs for those years.

But like I said -- maybe I am wrong.

Barralad77

Thanks @d6fs1l. I’m clearly out of my depth at this point. From what I see on the Govt site (and elsewhere) calculating the AA looks to be something designed to bring about maximum distress; it’s as if it’s been designed precisely so no-one (unless an expert in such matters) could possibly grasp how it works (at least, that’s the impression I’m left with).

Can I ask, when you do the more complex calculations (i.e., those that factor in CPI) does the AA figure come out as being higher or lower than the figure using the USS (just multiply by 19) figure?

Is it possible that USS don’t factor CPI in as the pension increases aren’t 1-1 linked to CPI? For example, if there was a year where CPI was 5% but the USS pension uplift was 2.5% (due to it being capped) then the 5% figure would be incorrect/misleading (possibly?).

And one final thing; if it turns out that people have exceeded their AA even though USS has told them they haven’t, who will be liable for any consequences? I suspect it falls on the individual, but feel that USS would bear moral responsibility to a large degree.

Barralad77

One last shot at understanding how this works.

I’ve just calculated my AA for 2024-25 two ways:

[1] Increasing my ‘opening value’ by CPI of 6.7% (following the instructions on the old form you linked to on the IC site).
[2] Increasing my ‘opening value’ by 1.4% (which I think is what we got as the uplift that year?).

Using [1] the AA figure comes out at many £000’s lower than the figure quoted on my statement (it’s around £18k lower).
Using [2] the AA figure is within £4 of the figure quoted on my statement.

In sum, the difference between the two figures calculated and the figure quoted on my statement are [1] £18,000 and [2] £4.

It does look to me as though multiplying the annual pension bought (1/75) of salary by 19 and adding in any IB contributions is correct. I’m not a maths/finance expert but I’m assuming the increase of 1.4% (to the opening value) is accounted for in the higher value of the 1/75th, so no further consideration is necessary?

Happy to concede I could be way off the mark, though.

d6fs1l

Barralad77 said:

Can I ask, when you do the more complex calculations (i.e., those that factor in CPI) does the AA figure come out as being higher or lower than the figure using the USS (just multiply by 19) figure?

You would expect it to be lower because the pension accrual will be smaller if the opening value is higher. This is especially significant if you are adjusting the opening value by 10.1% (as in 23-24) because the opening value would be larger than the closing value, and thus the input would be zero!

It looks to me that the USS figure is simply the (new retirement builder accrual * 19) + investment builder contributions. That does appear to disregard the CPI adjustment.

You might think -- well, better to be wrong that way around (i.e. better to overestimate the input rather than underestimate it) and there is some truth in that. But that may have undesirable consequences for people who want to maximise the use of their annual allowance.

Something else has just occurred to me. When I plug my own pension figures into (say) the M&G calculator I get wildly different pension inputs from the ones given by USS. The absence of a CPI adjustment does not explain all of the differences.

What I've just realised is that if the calculation in the annual statement is based on the new RB accrual * 19, the value of the final salary component (for those who have it) is being disregarded. I now can see that this (alongside the absence of the CPI adjustment) is what is causing me to generate such different figures from USS -- because when I've looked at the opening and closing values, I have used the total pension from both parts (and the majority of my DB accrued pension is from the final salary section).

Maybe that's what is confusing me. Obviously, there is no new final salary accrual, but there is a revaluation. So the FS component does increase. I thought that you factor the value of that FS pension into the input calculation -- but maybe not. It now occurs to me that if the FS section is treated as a deferred scheme, then the increase in pension in it would not count towards a pension input. (see e.g. Annual allowance).

So that would (finally) explain to me why I cannot understand the number in the annual statement!

But even if I'm right about the final salary point, it still would not explain why the CPI adjustment does not appear to factor in the USS number.

There may well be further subtleties that I do not appreciate -- I did wonder about the linkage with CPI that you mention, although I can't find any source which suggests that this would affect the calculation of the input.

At this point one wonders how anyone is supposed to make their way around any of this. I think USS would disclaim any reliance on the figures in the statement, but what else could a non-expert person reasonably do?

Barralad77

I think the ‘old’ form you found on the IC site mentioned something about the final salary element being irrelevant to the way it’s now calculated. 

But going back to the CPI, I worked out that my pension at March 2025 was 6.8% higher than it had been at March 2024. Accordingly, if you increase the opening value by 6.7% then you are effectively wiping out the notion of any increase. It resulted in the difference between my opening and closing values being under £50, which even when multiplied by 19 wasn’t a big number. 

The more I’m looking at this the more confident I am that the USS figure for AA can be trusted although I completely agree that it’s a potential minefield to navigate if you refer to other resources (esp. the Govt’s own websites!).

Of course, none of this accounts for why the LSA figure quoted on the statement is completely out of whack. 

hyubh

Barralad77 said:

Is it possible that USS don’t factor CPI in as the pension increases aren’t 1-1 linked to CPI?

No, this is the statutory calculation. Part of the complication comes from standardisation, e.g. that + the fact AA input periods are aligned to tax years not scheme years (as they were originally). Some public service schemes have changed their in-service revaluation date to allow for this (e.g. see this for the LGPS), might be worth double checking the situation with the USS.

d6fs1l

You're right about that old guidance mentioning final salary (well spotted), but it says those benefits are carved out because they increase by CPI. However, I don't think that is true for post-2011 benefits -- the increase is subject to caps (I think). Again, where this matters is in the two recent years of high inflation...

I am not sure I can trust the statement figure for AA until I understand why the CPI adjustment appears not to have been applied. As you have shown, it can make a material difference in times of higher inflation.

(I agree re the LSA -- I don't really understand why they have included that figure, because the amount of LS you take is a matter of choice.)

And we wonder why people find pensions difficult....