LGPS Annual Data Dump – The Messages Behind The Headlines

MHCLG recently released detailed March 2025 data (i.e., SF3 data) for the English and Welsh Local Government Pension Scheme.  As well as data on contributions, expenses, and the member profile for each fund, it is also possible to calculate investment performance from the data both in aggregate and by fund.

Pooling is not (yet) saving funds money

At a Scheme level, it is not surprising that overall expenditure has risen by about 9% over the past twelve months.  The bulk of expenditure is investment costs, which is in the main charged on an ad valorem basis and tend to grow broadly in line with the growth in assets.  Over the last 12 months investment expenses in fact grew by 8.0%, more than the 2.7% increase in fund assets.  Governance and administration costs rose even faster, by 11.6% and 13.3% respectively.

There are many reasons why investment costs might have grown faster than assets.  Reporting standards may be improving;  allocations to higher fee private investments are on the rise (though so are those to ‘cheap’ gilts);  the cost of building out the surviving pools’ investment expertise must have filtered through to fund costs, and some of that may well have been allocated to investment expenses.

There are also still caveats about the quality of the data:  ten funds reported costs of below 20bps of assets, while four reported over 100bps.  That degree of variance is unlikely, but any error will be on the side of under- rather than over-reporting costs.  The blunt message for the Government is that the aggregate figures do not yet support any claim that pooling is saving money.  It may be that the costs being incurred by funds today will save money in the future.  But equally they may not.

Aggregate net cashflow continues to deteriorate

Aggregate cashflow before investment income continues its negative trend.  This has been a theme in my commentary on the SF3 data over the past ten years and should not be a surprise to anyone.  Benefits payments rose by 12% in the past year and have risen by 7.1% annualised over the last five years.  Benefits inflation is an obvious factor, but payments are also higher because of increasing lump sum payments and greater numbers taking pensions early for ill health or similar reasons. 

This year, total contributions (employer and employee) fell short of benefit payments and expenses by £4.1bn or 1% of assets.  Over the last decade or so, primary contributions have often been pre-paid in the first year and therefore dipped in the third year of the valuation cycle.  This time round, there were fewer prepayments in 2023, presumably because the higher returns available on cash gave employers less incentive to pre-pay.  Correspondingly 2025 primary contributions held up better than in either of the previous two cycles, and were only about £300m lower than 2024.

Secondary contributions (i.e., those to finance shortfalls in past service funding) unsurprisingly continued to fall, as employers increasingly find themselves in a fully funded position.  Some funds have also chosen to return money to employers by levying negative ones.  In 2020 secondary contributions formed 21% of the total, whereas in 2025 the number was only 8%.

Can investment income continue to plug the gap?

Investment income, which constitutes about a third of total income, made up the shortfall between rising benefit payments and falling contributions, and grew by 14% in 2025.  Behind this were higher interest rates and bond yields, and perhaps funds switching from accumulation to income units.  Net cashflow excluding transfers therefore ended up at a positive 0.7% of assets, broadly in line with previous experience.  

But that bald figure hides the fact that the increase in investment income is unlikely to be repeated every year.  The combination of lower contributions, a 3.8% rise in pension payments, and a much lower increase in investment income means that over the next valuation cycle net cashflow is likely to reduce closer to zero.

Investment performance over 3 years matched the average discount rate

At Scheme level, investment performance, calculated somewhat crudely as the growth in NAV net of i) benefit payments less contributions and ii) net transfers), over 12 months was 4.1%, and over 3 years on an annualised basis was 4.4%.  The real picture today is almost certainly better, because 31^st March 2025 was close to the low point in markets ahead of Trump’s Liberation Day, and investment returns since then will certainly have been higher.   

These data show that the LGPs overall has done its job in growing assets by a similar rate to the discount rate (average 4.4%).  In comparison, an even cruder reference index consisting of 60% global equities (unhedged)and 40% gilts would have returned 0.2% annualised over the same period.  While the results look reasonable, they give no information whether the level of risk taken was appropriate – or put another way, whether the return justified the risk taken.  

Maybe scale does not matter?

Among the 86 funds in England and Wales, as one would expect, there is a wider range of performance.  Over three years the best individual fund performance was 9.4% annualised, and the worst was -0.6%. 

There is little or no correlation between individual funds’ investment costs and performance over three years, but perhaps that should be expected.  I do find it interesting that four of the best fifteen performing funds over three years were from Wales.  It may not be the case that, as the Government’s asserts, size and scale are needed to deliver effective performance, since the Wales pool AUM is well below the pooling minimum of £50bn.  As it happens, all four funds came in with expenses at or below the LGPS average figure.

Messages for the future?

Looking forward, what can we expect?  If pooling is to succeed, it must demonstrate real cost savings for funds.  Secondly, I suspect that funds have taken more risk than is necessary to deliver a 4% annualised return over the last three years, and I would hope to see higher returns, or alternatively some de-risking.

But maintaining positive cashflow so that pensions are paid is far more important than either of these.  Here the 2025 data shows 20 funds with negative cashflow even after investment income is included.  Anecdotally more and more funds are finding themselves in the same position.

As funds mature, there is nothing wrong with having negative cashflow, but it does have to be managed.  And in these uncertain times, I suggest that looking at a wide range of scenarios is important.  Could a future Reform UK Government close the scheme to new entrants?  Will central banks target 4% rather than 2% inflation?  Could a financial crisis make it difficult to remit foreign currency?  Will investment income be imperilled? 

In my view the overall message has not changed, that every fund should create a robust plan to manage cashflow.  Perhaps the increased volatility in markets and politics means that considering lower probability negative scenarios has become more important.