21 Mar 2024 5 min read

Unlocking surplus: adapting to an evolving endgame

By John Southall

DB schemes are maturing, but the endgame is in a fascinating state of flux. Here's how we're ensuring our quantitative frameworks remain flexible as the regulatory backdrop shifts.


Moltke the Elder, the Prussian field marshal, is known for his belief in developing a series of future strategic options, instead of a single plan.[1]

We’ve been thinking about DB strategy design for many years, and have developed various quantitative frameworks to help trustees manage their strategies. However, some of our modelling assumptions about the endgame have not played out quite as we expected. Our frameworks often included the following assumptions:

a) No extra utility is gained from being able to pay more than 100% of benefits

b) As soon as a scheme can buy out, it does. This can be achieved in short order

Assumption (b) clearly hasn’t stood the test of time! Many schemes are now more than fully funded on buyout, and willing to buy out, but can’t. Reasons include administration hurdles, capacity constraints in the insurance market and illiquid assets holdings. In recognition of this reality, last year we outlined a framework for well-funded schemes that want to buyout but can’t yet.

On top of this, some schemes may not wish to buy out immediately even though they can, challenging assumption (a). The motivation is that persistent surpluses might benefit the sponsor, DB members (in the form of increased benefits) or be used to help DC members.

In this blog we explain how our frameworks have adapted to the evolving landscape. We also note one potential implication: high funding levels may allow more than just the surplus to be invested in growth assets.

A new dawn for extracting surplus

The Department for Work and Pensions (DWP) launched a consultation on Options for Defined Benefit schemes on 23 February 2024[2]. It runs until 19 April and concerns "Plans to ensure that the £1.4 trillion held by pension schemes in the UK delivers maximum benefit for both savers and the broader economy." Of relevance to this blog, the consultation examines the eligibility criteria for surplus extraction:

“Any extraction of surplus will reduce security for members. In establishing a new basis to permit surplus extraction we need to ensure there remains a very high probability that member benefits will be paid in full. This implies that any surplus extraction should still leave the scheme over 100% funded on a prudent basis. However, the level of investment risk and the strength of the sponsoring employer will also have a significant bearing on what level of surplus is ‘safe’ to extract.”

The security of benefits will always be a key goal. However, rather than pursue security of members benefits at all costs, schemes could be permitted to tolerate a small reduction in security. In exchange they have the possibility of much larger surpluses in the future. Two important questions trustees need to consider, which our frameworks can help answer, are:

  • At what point is extracting surplus a ‘safe’ activity?
  • How should a scheme tailor its investment strategy towards growing the surplus, while seeking to protect accrued benefits?

Considering strategic options

As an illustration, we considered two relatively simple optimisations of investment strategy across three building blocks – diversified growth, CDI (cashflow-matching credit) and LDI[3].

In the first optimisation, the old-school approach, we only sought to maximise security. But in the second, adapted approach, we allowed for more aggressive strategies provided a certain level of security was met.

For the purposes of this exercise, we assessed security using the probability of failure (PoF) i.e. the chance of failing to ultimately pay all promised benefits, assuming no further contributions[4]. Trustees might, for example, consider benefits to be secure if there is less than a 1% chance of ultimate failure. Accordingly, the two approaches were:

  • Approach (1): Seek to minimise the PoF
  • Approach (2): Seeking to maximise the expected return subject to the PoF being less than 1%. If this isn’t possible then minimise the PoF

We also considered an intuitive third strategy:

  • Approach (3): Invest up to the level of the liabilities in LDI and CDI and invest the surplus in growth. The idea is that you could use low-risk assets to match promised benefits, but the excess assets to invest in growth to generate surpluses

The charts below show our results. For different initial funding levels (on a gilts basis) we’ve shown the optimised asset allocation expressed as a proportion of the liabilities.


The results under approach (1) are straightforward and unsurprising – the asset allocation de-risks without limit as the funding position improves. But the results from (2) are potentially more interesting – the picture is like (1) at lower funding levels, but at higher funding levels the growth allocation increases.

This reflects that once you have a larger buffer, matching is no longer really needed to achieve an adequate level of security, and you are freer to focus on growth[5]. Comparing (2) with (3), you can see it’s possible to justify putting more in growth than just the surplus.

The chart below shows how the probability of failure evolves with the funding level under approaches (1) and (2). They diverge at around 115% funding on a gilts basis. This could be used as a threshold for surplus extraction since the sponsor could withdraw surplus above this funding level without the probability of failure falling below 1%:


A refreshed perspective

There are, of course, plenty of ways we can enhance the analysis above. These include using metrics that account for the degree of shortfall, which we prefer in general[6]. We can also allow for covenant risk and incorporate private markets (including ‘productive finance’). Finally, we can model buyouts and buy-ins at some stage – even for schemes aiming to grow their surpluses we expect buyouts or buy-ins to eventually become at least part of the solution.

The analysis here is just a taste but shows how the frameworks can support strategic decision making in a dynamic endgame. As Moltke the Elder would agree, planning is important, but adaptability is essential.


Key assumptions for the illustration: 

  • Diversified growth and CDI are expected to earn 3.0% and 1.0% excess geometric returns over cash respectively
  • The pension scheme has a duration of 14 years
  • Liabilities are fully hedged against rates and inflation risks and leverage constraints mean at least one third of the liabilities must be held in LDI


[1]  He is thought to have said: “No plan of operations extends with any certainty beyond the first encounter with the enemy's main strength.”

[2] The UK's new DB funding framework will also apply to scheme valuations from September. This guides schemes towards investing in line with a low dependency asset allocation once it reaches ‘significant maturity’. Considerable uncertainty remains, but there is likely to remain considerable flexibility in terms of how to invest, particularly when it comes to surpluses.

[3] We assume the scheme hedges all the rates and inflation risk in accrued benefits using LDI and that this is subject to sensible leverage constraints.

[4] The calculations involve projecting the whole future life of the scheme under many scenarios and, for each scenario, observing whether the assets were sufficient to meet all the cashflows promised.

[5] You might note that the asset allocations for (1) and (2) have a significant amount more in growth than (3) even when 100% funded. This can be explained by the fact we allow for other risks (including longevity) in the model, which can encourage targeting a higher return.

[6] Replacing the PoF with expected funding shortfall (ES), for example, leads to more cautious strategies. ES is 100% minus the Expected Proportion of Benefits Met, for fans of our previous work!

John Southall

Head of Solutions Research

John works on financial modelling, investment strategy development and thought leadership. He also gets involved in bespoke strategy work. John used to work as a pensions consultant before joining LGIM in 2011. He has a PhD in dynamical systems and is a qualified actuary.

John Southall