As an avid reader of FT Alphaville you’ll recall our previous explainer on UK local authority pensions. You’ll know that the Local Government Pensions Scheme is the largest funded scheme in the country, the fifth largest in the world, and that UK local authority pensions account for around £458bn of assets. You’ll remember that rather than being run as one large fund, the assets are run by 87 different administering authorities (aka local government bodies) across England and Wales, and a further 11 schemes in Scotland.
But what do these schemes actually invest in, what kind of returns have they delivered, and why might any of this matter to us? We’re here to answer these hitherto unnoticeably burning questions.
Buying what?
Each administering authority has a pensions committee responsible for the big asset allocation calls. Elected councillors are the ultimate arbiter of each fund’s strategic asset allocation, and it’s these decisions that almost entirely determine returns.
Strategic allocations differ, and not just at the margin. For example, Kensington & Chelsea’s approach, outlined in MainFT, is to have a lot of equities. Camden is big into commercial property. East Sussex leans heavily on clever Mayfair-types running hedge funds and private equity. Clywd is long infrastructure. And Barnet loves bonds.
To state the obvious, they have made ridiculously different asset allocations choices in pursuit of a broadly similar investment objective. Are any of these asset allocations optimal? Maybe one of them is, but without a crystal ball, all we can do is look at the performance.
How has this all worked out?
Here’s a chart showing the one-year returns for the year ending March 2023. We’ve grouped the administering authorities by the pooling company of which they’re a member for a mixture of (a) convenience (if you’re a LGPS nerd searching for a fund) and (b) prettiness (if you’re not). As you’d have guessed given the variety of asset allocations, returns are pretty disparate:
One immediate takeaway is that the numbers look terrible. But remember, these were for the year in which global stocks fell, property bombed, and the less said about bonds the better. We’ve thrown some index returns in at the bottom for context. Hover your mouse over the dots for more info.
Don’t use the chart to judge how good the pools are at producing good returns. It’s the administrative authorities — and not the pools — that decide each fund’s overall asset allocation, and this decision is the major driver of overall returns. To take an extreme example, the dot showing the worst absolute return represents the ‘Environmental Agency – Closed’ fund, which dropped 18.3 per cent in the year ending 31 March 2023. It was the Agency’s pension committee that chose the strategic asset allocation (long-dated inflation-linked gilts), not Brunel. In fact, while it’s a member of Brunel, the value of assets pooled with them is precisely zero.[1]
Do the numbers even out over the long-term? We scraped manually copied from a PDF some data we found in a PIRC LAPPA report on Hounslow’s website — examining performance over the past decade for a large subset of the full data set.
According to PIRC, Kensington & Chelsea are streets ahead of the rest, with an annualised return of 10.8 per cent over ten years — just behind global stock indices in sterling terms. The laggards are Waltham Forest and Cornwall, with 4.8 per cent and 5.3 per cent annualised returns respectively. To state the obvious, the differences between the highest and lowest numbers are large.
Cui bono?
Beyond bragging rights, for whom does this actually matter? It has got to matter for some mixture of the LGPS’s six million current / future pensioner members and its twenty-one thousand employers. And, by extension, all local taxpayers.
Staff members are promised a defined benefit, and this does not vary with fund performance. So stonking returns from stocks (or lack thereof) shouldn’t really touch them.
Sure, according to a legal opinion published on the LGPS Board’s website, local government pensions are not actually guaranteed by central government. So maybe there’s a case for saying that the strong returns will benefit members by enhancing the security backing those pension promises made by the council. And it’s true that councils have been going bust issuing 114 notices with alarming frequency.
But it’s hard to imagine a political scenario in which pensions are allowed to go unpaid. In fact, the KC paid for his published opinion that LGPS pensions are not guaranteed felt compelled to down tools mid-opinion and write:
I must say that, from a political as opposed to a legal perspective (and for what my views from such a perspective may be worth), I find it well-nigh inconceivable that central government would allow matters to reach the stage of a complete collapse in local authority finances, and default upon authorities’ legal obligations, without undertaking some form of intervention.
If current and former local government employees benefit only at the margin, who else takes a direct hit in a downturn or benefits from a market upturn? Employers. These are — largely, but not exclusively — public sector bodies, backed by taxes.
We say largely, because maybe 1,750 of the approximately twenty-one thousand employers included in the LGPS aren’t actually public bodies. They’re so-called Tier 3 employers. (Tier 1 employers being public bodies with direct local tax-payer backing, Tier 2 being the thousands of academy trusts.) If you’re working for a charity to whom the council has outsourced some contracts, are a non-teaching member of staff for a post-92 university, or someone working in a housing association, you’ll likely have a LGPS pension via a Tier 3 employer.
OK, but why should *I* care?
Employers mostly feel the impact of varying investment returns through changes to the contribution rate, which they stump up each time they turn the payroll handle. Payroll that (mostly) gets paid for by your central and local government taxes, by closing down libraries, or chiselling away at the cost of providing adult social care.
Contribution rates are estimated only once every three years by actuaries, and are the sum of what are known as the primary and secondary contribution rate. The primary contribution rate is to pay for the future cost of brand new pension promises. Variation between administering authorities is surprisingly large given that this is the effective price they are charging themselves to provide the same LGPS pension to their employees.
The highest primary (employer) contribution rates are for the open portion of the Environment Agency and Lincolnshire funds, coming in at just over 24 per cent of salary. The lowest are at the London Pensions Fund Authority and Kensington & Chelsea, at only 15 per cent of salary.
These numbers are arrived at using a whole lot of whizzy maths, locally-specific mortality data, member characteristics, etc. We’re not in a position really to unpick the differences here. But there are two massive hand-wavey numbers that go into the calculation of the primary contribution rate that are pretty important. These are the pension committee’s expectation for inflation, and the discount rate they select.
To get a sense as to how important these are, we examined — for no particular reason — the funding sensitivity analyses for Essex, Merton, Newham and Nottinghamshire as set out in their 2022 triennial valuation reports. In each case, if either (a) expected inflation is nudged up by 0.1 percentage points, or (b) the discount rate is nudged down by 0.1 percentage points, the primary contribution rate increases by 0.7-0.8 percentage points, making it more expensive to employ local government people in cash terms.
We thought we could make a cool dataviz, which would show the funds with the highest discount rates and lowest inflation assumptions had the lowest primary contribution rates. But this is not really how things work out. Had that been the case the darkest dots (funds with the highest primary contribution rates) would be up in the top left quadrant, and the lightest dots (with the lowest rates) would’ve been in the bottom right quadrant. So there are obviously a lot of other things going on.
Once money is contributed to the pension fund it gets invested in stocks, bonds, infrastructure, etc. If the assets outperform all the assumptions set out in the triennial valuation, this produces a ‘funding surplus’ the next time the actuaries run the numbers. If they underperform the assumptions, the fund will look ‘underfunded’.
So how did they look the last time actuaries turned the triennial valuation handle?
Happy days! Unless you’re living in E17 or Havering, your local authority’s pension fund was either in surplus or only a modest deficit. Which is presumably good, because it means that your Council Tax has little prospect of spiking to bail out your councillors’ unwise (or unlucky) asset allocation choices.
But hang on, these triennial valuations are also based on the same bunch of hand-wavey assumptions used to calculate the primary contribution rate. And as the Government Actuary’s Department wrote in a terrific report on the variation in assumptions used:
Our analysis showed that there was no clear influence due to the asset mix, prudence, funding level, type of employer or maturity in isolation on the discount rate adopted.
Gulp! This is how an actuary says ‘you’ve just pulled these out the air’. And they have the charts to prove it (in Appendix B of this document).
Luckily, there’s a way to compare all the funds on a like-for-like basis. This is the so-called funding level on a SAB basis (using some standardised assumptions found in table G1 of this document). And funds are obliged to report this stat in the Section 13 dashboard at the back of their triennial valuation.
So, how do things look on an SAB consistent basis? Actually, marginally better:
Again, Kensington & Chelsea comes out on top — but not by much — with a SAB funding level of 164 per cent. In fact, while calculations using SAB standardised assumptions come out with an answer that is on average around ten per cent higher, this is far from uniform. While West Sussex reported a 125 per cent funding level in its triennial valuation, this funding level jumps to 159 per cent when common assumptions are used — we think indicating that the assumptions they’ve chosen are relatively conservative compared to their peers. And using the combination of scheme assumptions and standardised assumptions we can see that it was really Waltham Forest, Brent, Havering and Berkshire that looked most underfunded in 2022.
So what could Waltham Forest, Brent, Havering and Berkshire do with this information? Do they just pray for booming investment markets?
No. As well as a primary contribution rate (reflecting an estimate as to how much needs to be put aside to pay for the pension being promised), LGPS funds can also charge employers a secondary contribution rate. This is a rate added to payroll to get the fund to wherever it wants to be (eg, fully-funded, 105 per cent funded, 110 per cent funded), over a horizon decided by the fund (maybe 15 or 20 years). And this rate can be positive or negative.
So you’d expect all the funds that were in surplus to have negative secondary contribution rates, and all the funds in deficit to have positive contribution rates. That would at least be logical?
Ahem.
This chart of secondary contribution rates and funding levels shows that this is not the case. Sure, underfunded Waltham Forest, Brent and Berkshire led the pack in secondary contribution rates with nine per cent, twelve per cent and eight per cent respectively. And Kensington & Chelsea has been reaping the rewards of its over-funding by knocking seven per cent off payroll costs. But Staffordshire? Staffordshire’s 2022 valuation put it twenty per cent over-funded. And yet rather than cashing in through payroll discounts, it looks like it is chucking an *extra eight per cent of payroll* into the pension fund, on top of money it reckons will be needed to make good on its promises.
We honestly couldn’t understand why this was the case, so we got in touch with Steve Simkins, partner at the investment consultancy Isio. He pointed us to the existence of LGPS funding strategy statements, and the idiosyncrasies of the methodologies pursued by different actuarial firms. Each fund publishes one of these and it lays out why and how it’s a bit different from the others. We lasered in on the statement published by Staffordshire Pension Fund.
The reason why the Staffordshire fund levies upon its employer members a hefty secondary contribution rate despite its chunky surplus appears to be because it has in place a ‘stabilisation arrangement’. This means that the total (the sum of the primary and secondary) contribution rate employers pay can change only be +/-1 per cent of salary each year. At least, this is the case for local authority employers, as well as opted-in academies.
We can see where they’re coming from. If the global economy takes an absolute dive — dragging asset market values down with it — local taxpayers might not take especially kindly to huge increases in Council Tax to cover for the resultant hole in LGPS pensions. And back in 2016, Staffordshire’s funding level was only 78 per cent — requiring chunky secondary contribution rates as part of the deficit repair process. On the other hand, it seems a bit weird to us that local taxpayers are chucking extra money into the fund because it was, once upon a time, in deficit. Especially with local authority finances strained.
Scandal?
As a terrific piece from the Local Government Chronicle at the back end of 2023 explains, those assumptions summoned out of the air macro forecasts produced by actuaries — and adopted by funds — that are used to calculate funding levels are not only subjective, but can be bent to the will of the administering authorities. They reported:
[a] senior officer told LGC “a lot of funds” try to tweak actuarial assumptions to “hide” large surpluses that might lead councillors to demand a cut to their council’s contribution rate to free up money for other services.
If tweaking assumptions to make the pension fund look like it was in greater need of hefty contributions sounds pretty scandalous to you, you’re not alone. In fact, John Clancy, the former leader of Birmingham city council, has started making waves by pretty much calling it out as such. Or, at least, demanding an urgent review into the running of West Midlands Pension Scheme, saying that its demands for employer contributions were over half a billion pounds more than was needed. And this, he alleges, pushed Birmingham Council over the edge into issuing a section 114 notice (aka local authority bankruptcy) in 2023.
We haven’t checked his workings but can see that there are going to be tensions between the pension funds looking to shore up funding, and other spending priorities like the administration of services that they are legally obliged to provide, such as adult and children’s social care.
So what next?
If you’re reading FTAV it should not be news to you that, compared to March 2022 levels, stock markets are up around forty per cent and long-dated bond yields are around three percentage points higher. While we’ll need to wait for the actuaries to do their thing, we’ve no doubt that they will ultimately find that funding levels are crazily, almost unbelievably, stronger.
To get a sense as to quite how much stronger the funding positions will be, we can turn again to Isio, which has for a while been running a ‘low funding risk’ index of LGPS funds. They attempt to answer the question as to how over- or underfunded each fund is by running two steps. First, they guesstimate the value of each fund’s assets based on their strategic asset allocation and market movements. Second, they guesstimate the present value of each scheme’s liabilities — not discounted with the funky set of assumptions chosen by councillors but instead discounted using gilt yields.
This is a pretty penal way of discounting liabilities, and one that will inflate them to levels much greater than seen anywhere in the private sector — even bulk annuity insurance buyout liability valuation (making the funds look far less funded than triennial or SAB valuations). But at least it’s consistent. And it’s also live.
The chart below shows Isio’s estimates as to LGPS funds’ low-risk funding ratio back in March 2022 (when the last triennial valuation took place), and also at year-end 2024. You can toggle the filter to flip from one to the other.
While these numbers are just estimates, if we can suspend our disbelief to imagine them as credible proxies of the results that might come out of the 2025 triennial valuation round, we can get some staggering impacts.
We know that Kensington & Chelsea is pretty much the best-funded scheme, however anyone wants to cut things. The Isio low-risk funding index had them at 94 per cent in March 2022, while the triennial valuation puts them at 154 per cent and the SAB standardised assumptions puts them at 164 per cent. Whatever the level of funding, this led to them adopting a secondary contribution rate of *minus* seven per cent.
Fast forward to December 2024, and every single LGPS fund that is open to new members (bar Waltham Forest) has an Isio low risk funding index better than Kensington & Chelsea’s back in March 2022. There’s an important caveat that councils’ guesstimated liability discount rates are — according to actuaries we’ve spoken to — are unlikely to move up lock-step with gilt yields. If guesstimated discount rates move up by less than the gilt yields used in Isio’s analysis, funding improvements will look less dramatic.
But still, with the average secondary contribution rate across schemes being around two per cent of salary, a move to minus seven per cent — or more — would have a large impact on local authority finances. MainFT quotes Tim Gilbert, a partner at Lane Clark & Peacock, the actuarial consultants, as saying that the likely funding improvement would support “a significant reduction in the contribution rates . . . a reduction of 50 per cent or so is justifiable given the change of the conditions”.
Although we know from looking at Staffordshire’s funding strategy statement that some local authorities would need to tweak their arrangements to allow such a huge shift to take place in one leap, rather than just allow the contribution rate to decline by one per cent per year. And there seems to be some chat in LGPS-world about whether the funds should lock in their massive surpluses by de-risking (aka, buying gilts), though LPPI reckon in a new report that doing so would make future benefits extremely costly.
Scotland the brave?
Other than in the intro, we’ve completely failed to discuss Scottish LGPS funds. This is partly due to laziness (their valuations, fund accounts, etc are not conveniently gathered on the LGPS Board website). But it’s also because they’re on a slightly different valuation cycle and have already taken a slightly different path.
As the Local Government Chronicle reported back in 2023, Scottish funds’ aggregate funding looked likely to jump to an average 147 per cent from an average 106 per cent in 2020. The boost was pretty evenly split between contributions from strong asset markets and a shrinking in liabilities derived from changes to assumptions (that we’ll assume means higher bond yields).
What did Strathclyde Pension Fund — the largest of the Scottish funds — do? It slashed the contribution rate for its main (mostly council) employers from 19.3 per cent of salary to just 6.5 per cent of salary. But it did so with a caveat. They told employers that the cut was temporary and would jump back to 17.5 per cent in the year 2026-27.
We can see the merits of this approach, and why this might be something that English and Welsh funds might want to pursue. Indeed, close analysis of Kensington & Chelsea’s 2022 valuation shows that their minus seven per cent secondary contribution rate was scheduled to fall away in the third year of the forecast, suggesting that they’d gone down the same road. Local authorities get some relief and a bit more money in their pocket today, but also the knowledge that this isn’t part of their permanent budget.
So maybe they can commit to some shorter-term projects, or do a bit of capital spend (if such budgets are even fungible?). If things turn out even better than expected, funds can then extend the contribution holidays.
In fact, MainFT reported last week that Kensington & Chelsea has decided that rather than hike secondary contribution rates, it would use its windfall investment gains to cut employer contributions all the way down to zero in this third year. This is despite their actuary warning that to do so was “inappropriate” and would foster an expectation “that this is a sustainable rate in the long term”. The actuary agreed:
that a zero rate would only have a marginal impact on future outcomes and should not have a detrimental effect on the ability of the Fund to pay future pension benefits.
But it’s easier to cut a contribution rate than hike it.
Getting to the point
Given the embarrassing riches on the pension side we can even see the opportunity for something radical. Maybe we’ll see authorities in aggregate halving employer contribution rates, and some even scrapping for the next three years. We hate seeing local services shuttered or squeezed. And cutting pension contributions will surely provide some relief.
Moreover, it would be nonsensical to maintain over-contributions to funds that are in significant surplus. But it’s not really pension contributions that have been the problem for local authority finances.
As MainFT has reported extensively, the big driver of deteriorating local authority finances has been the ever-rising cost of adult and child social care. 2025 looked like a bit of a crunch year for some heavily indebted councils, and an accelerating number are forecast to issue section 114 notifications, aka go bust. We’re not local government finance experts, but can see that a cut in payroll costs might save some councils from going over the precipice. And don’t get us wrong, that would be a very good thing. But a bad side-effect of this good thing might be that it helps government kick the political football of sustainable social care funding further down the road.
So sure, let’s not have local authorities pile unnecessarily large amounts of cash into pension schemes, but also please let’s not spend the respite afforded by any cut in contributions on evermore policy procrastination.
Bonus content (!)
We’ve just enough pixels left to treat you to two final dataviz. It’s the one that council taxpayers are probably most interested in, that shows the overall contribution rate (the money that actually gets charged as part of payroll), and the degree to which this is informed by secondary contribution rates (the premium or discount to ongoing estimated cost of providing future pensions). We think it’s a banger. And if you’ve stuck with this post long enough to reach this point, chances are that you will too.
This is where the rubber really hits the road.
On the far right hand side is the London borough of Brent. With an average secondary contribution rate close to 12 per cent of salary, the total employer cost of providing an LGPS pension is around 33 per cent of salary.
At the other extreme is Kensington & Chelsea, which has an average secondary contribution rate of almost minus seven per cent. Given that they have a weirdly low primary contribution rate to begin with, the total employer cost of providing an LGPS pension is a mere 8.1 per cent of salary.
Now, before you get too cocky about your understanding of LGPS, do remember that each of the bubbles on the chart is itself a multiemployer fund. And each of the (often hundreds) of employers included in each fund will have their own individually calculated primary and secondary contribution rates. To give an idea of the multitudes any given dot contains, we pulled out the numbers for the West Midlands Pension Scheme.
The average secondary contribution rate for West Midlands is 1.9 per cent of salary. But as our searchable dataviz showing the data for all 787 employers in the scheme illustrates, employers could be paying anything from zero to half of salary, largely due to the variation in secondary contribution rates. And these will vary within a fund for a bunch of reasons.
Our guess is that much of the variation is probably because some employers will have joined the scheme when assets were flying high (and their performance subsequently underperformed modelled assumptions) or when financial markets were in the doldrums (and the fund’s subsequent outperformance of modelled assumptions mean they have more than enough to pay future benefits).
Anyway, congratulations. You’ve made it to the end, and can now officially call yourself an LGPS nerd.
[1] It’s an odd fund to be honest. It’s the smallest of the funds – at less than a tenth of the median. It’s the only fund closed to new entrants and in run-off (the Environmental Agency has an open fund too). It holds only cash, inflation-linked gilts, and some unquoted stocks that have been written down to nil. It’s got a formal guarantee from central government – something lacking elsewhere. And it looks like it’s run on a non-discretionary basis, with the pension committee giving the fund manager specific directions as to which inflation-linked gilts to hold. None of this is normal. This also happens to be the fund that recorded the largest outperformance of its benchmark (which dropped over 39 per cent for the year).
Further reading:
— Sorry, but it’s time to start caring about Local Government Pension Schemes
https://www.ft.com/content/87c321ab-e5ac-4a1d-a637-c1f7befcc1cb
