High-level strategy and implementation

Written By: Eric Lambert Actuary and Independent Investment Adviser

Eric Lambert, Actuary and Independent Investment Adviser, discusses the fundamental question of the difference between strategy and implementation and shows how the current dispute about whether pools or funds should be the decision-makers could and should be decided.

At LAPFSIF (Local Authority Pension Fund Strategic Investment Forum) held at the Andaz hotel in London in early February I facilitated a roundtable discussion on “Pooling objectives for the LGPS (Local Government Pension Scheme)”. One particular topic I was keen to explore was whether the line between strategy (a sovereign fund activity) and implementation (by pools) is currently blurred. This article will discuss two important topics – what constitutes high-level strategy and the pros and cons of pools advising on strategy.

The catalyst for these questions was the DLUHC (Department for Levelling Up, Housing and Communities) response¹ to their pooling “consultation” (sic). In particular, section 9 states that the government will revise pooling guidance to set out a preferred model of pooling including delegation of manager selection and strategy implementation. The rationale given in section 40 is that this will allow the pool to deliver the benefits of scale.

What constitutes high-level strategy?
To discuss what high-level strategy is and who should provide advice on it we need to understand what investment strategy entails and how it is normally derived.

Strategy can mean different things to different people: indeed, this is the essence of the issue. While there are other possible definitions I believe a reasonable description of investment strategy from a fund’s perspective is the strategic (i.e. long-term) mix of assets deemed to provide an appropriate investment profile. This is usually understood to be a combination of several features such as return, risk, efficiency and investment income. Each fund will focus and prioritise whatever features of investment strategy are most relevant and significant to it, reflecting its fund-specific circumstances. These typically include the investment beliefs of the decision-makers, the fund’s liability profile (especially its cash flow and liquidity needs), its return appetite, its risk tolerance, current funding level, current and future contribution level and so on.

Note that several of the profile features are recognised investment trade-offs, for example return and risk, and so the chosen strategy will indicate for each fund its most suitable compromises. For example, a super-mature fund with structurally negative cash flow (before taking account of investment income) might prioritise investment income over total return.

While not the subject of this article, other fund strategic decisions might be with reference to LDI (Liability-Driven Investing), CDI (Cash-Driven Investing) and overall currency strategy. The exposure to Levelling-Up (LU) assets is also a strategic decision but they are not a separate asset class; the composition of LU assets will likely come from across several different asset class exposures.

The most significant investment decision
Strategy is indisputably the most significant investment decision. In terms of investment profile the strategy is likely to dictate the majority of the return (typically >70%) and an even higher percentage of both the investment risk and the investment income. The beauty – and the curse – of strategy is that it takes a very long time (usually at least 10 years) to assess whether it has been appropriate or not and, except in extremely rare circumstances, it will almost certainly turn out to have been sub-optimal. Hindsight does indeed provide 20:20 vision. The best that a decision-maker, typically a PFC (Pension Fund Committee) for LGPS, can hope for is to make a rational decision with the benefit of reasonable, objective input and advice.

Investment strategy is typically articulated by an SAA (Strategic Asset Allocation) benchmark with some scope for modest variation, either deliberately under some form of discretionary TAA (Tactical Asset Allocation) or, more commonly, to facilitate asset relative market movements without triggering frequent and potentially costly rebalancing. A popular (especially in the US) simple SAA benchmark is “60:40” meaning 60% equities and 40% bonds. Historically this has shown a decent balance between return-seeking assets (equities) and defensive stabilising assets (bonds). I wonder how many of the modern vastly more complicated SAA benchmarks have delivered significantly improved results on this most basic benchmark?

It is clear to me – and I do not think I am alone – that the government believes funds and pools are deliberately blurring the lines between strategy (a fund decision) and implementation (a pool activity). This belief encourages a model of pooling where this separation is more distinct, to more fully realise the benefits of scale. Some consultation respondents suggest that the distinction between strategy by funds and implementation by pools was not clear cut – a great understatement! The government’s reference to high-level strategy being a sovereign fund activity and advising specifically that investment product and/or manager selection is a pool activity is an attempt to confirm and/or redraw the lines. This however begs a potentially awkward question: what is meant by high-level strategy? And how have funds, and pools, been able to fudge the grey area between strategy and implementation?

For example, let’s consider the predominant asset class for LGPS funds: equities. High-level strategy might be interpreted as the overall exposure percentage (like the 60% in the 60:40 default US SAA benchmark) but before actual implementation there are many further layers of decision: geography (global, regional, individual markets), active/passive, internal/external, style of management, (quant/fundamental, growth/value, bottom-up/top-down), publicly-quoted or unlisted (private equity) etc.

Who makes these decisions? Where does strategy stop and implementation begin? It is entirely possible for a fund to specify an equity exposure in sufficient detail and with myriad constraints ensuring that only one asset manager could offer this product, so that fund can effectively do asset manager selection. For example, is a strategy with a prescribed percentage exposure to internal active UK equity not just manager selection?

A potential contradiction
Even the government’s own position contains a potential contradiction as it is encouraging all LGPS funds to target 10% exposure to Private Equity. This suggests that this is to be regarded as a fund decision rather than an implementation (pool) decision. So, at a minimum, high-level strategy in equities should perhaps specify the percentages to publicly-listed equities and unlisted (Private Equity). I personally do not believe funds should be specifying geographic/regional/individual market (typically expressed as a positive bias to UK equities and/or a bias for – or against – Emerging Markets), active vs passive, style of management etc. These all strike me as implementation decisions.

It is important to stress that exampling exposure to an internal UK equity team as “implementation masquerading as strategy” should not be interpreted as anti-internal management: indeed, quite the reverse. Government, in encouraging internal management by the pools, has captured a potentially genuine benefit of scaling. Larger pools of assets – the scale argument – facilitate internal management, which can indeed be a very efficient way to manage assets. Although the evidence is sometimes skewed by a less than comprehensive view of the true costs, internal management is usually cheaper than external third party management. In addition to having a lower headline asset management charge, internal asset management tends to exhibit lower security turnover and therefore also lower implicit costs from transaction fees.

Given asset management cost savings was the original – misguided – primary driver of pooling it is easy to see why government supports this. Ill-timed third-party asset manager changes by funds, a truly destructive feature of any investment industry, are also largely negated by internal management. Historically funds have been poor at manager selection and the timing of their subsequent release, which is probably the reason that government does not want this to be a fund activity. The evidence has not yet accumulated that allows us to pronounce that pools are better at this activity; the jury is still out, although in Scottish law the appropriate verdict might be “not proven”. Strangely I find myself in agreement with both DLUHC and the Labour Party’s Plan for Financial Services “Financing Growth” as it also favours the LGPS adopting cost-effective in-house fund management.

In earlier times fund strategy would have been a simple choice of a very limited range of assets: equities, bonds and (possibly) property, probably umbrellaed by a split between growth (return-seeking) assets such as equities and defensive (risk-mitigation) assets such as bonds. The 60:40 equities:bonds split is a famous example. In more “enlightened” times in the ‘80s and subsequently, strategy was driven by ALM (Asset Liability Models) although these were always more asset-focused than they were liability-aware. This resulted in vastly more detailed SAAs with spuriously accurate weightings, for example the ridiculous 3.75% to Pacific exJapan equities. This meant funds had to select, monitor and (all too frequently) de-select asset managers for each of these detailed exposures within their SAA. How convenient for the consultancies which advised on the strategy and then also provided the asset manager selection(s)!

Today’s situation is somewhere between the simple original and the over-prescribed ALM-driven SAAs. Perhaps a large part of the current confusion as to what constitutes strategy and implementation and finding reasonable separation is a direct result of the legacy linkage between the two activities and funds historically taking decisions on both elements.

Too much choice?
In today’s (England & Wales) pooling world there is a certain amount of tension between what funds want to define as strategy and what pools should be implementing. Pools are obliged to provide the asset classes and sub-pools (government refers to these as sub-funds) that their partner funds desire and this can be very fund-specific: e.g. active or passive, return and risk targets, style of active asset management, internal or external and so on. In my opinion this has led to an excessive choice of offerings from some of the pools. Does offering a wide range of sub-pools help or hinder the setting of strategy? I personally think the latter. A possible analogy is with restaurant menus: the best restaurants tend to offer a very limited choice compared to the massive menus in cheaper ones. In which restaurant is it easier to order an interesting and appropriate meal? Government would appear to have similar concerns as they encourage high-level strategy and not an excessive choice of sub-pools. The requirement for funds to pool (at least) listed assets by 31 March 2025 will force further consolidation and may also trigger acceptance of a clearer distinction and transparency between what is strategy (by funds) and what is implementation (by pools).

If funds continue to set strategy at too detailed a level, either they will have many non-pooled asset portfolios and/or the pools will be obliged to offer excessive sub-pool choice. Neither are seen as optimal by government. Pools should expect funds to invest via their existing sub-funds where possible, possibly a response to some cynical industry commentators who scurrilously suggest that consultants advise strategies which they know the pool cannot accommodate. This could avoid an unfavourable scenario whereby an excessive number of similar sub-funds undermine the purposes and benefits of pooling.

A simple rule
A simple rule for what is meant by high-level strategy might be to go by the essential asset characteristics such as return (or risk) drivers. I always prefer simple to complex and so would argue that high-level strategy is basically a choice of equities, bonds and real assets such as real estate and infrastructure. Many of the more detailed strategies are really a choice of product and/or product provider. Equities are essentially one asset class, although as previously mentioned the government has muddied this somewhat by their encouragement for each fund to have an ambition to have 10% Private Equity. Bonds could be characterised by government and/or credit, and by nominal and/or index-linked. Real assets (Real Estate and Infrastructure primarily) are relatively straightforward.

If a pool knows why a fund has chosen a particular strategy it is in a much stronger position to implement that strategy more effectively. For example, if investment income is a priority and the strategic real assets exposure is the main generator of that income, then this tends to dictate to the pool the type of real estate and infrastructure most suitable. Knowing the reason for the strategic bond exposure will again allow a pool to implement appropriately. If asset modelling has been correctly done at a suitably high level (and often this is not the case) then the asset investment profiles embedded in the model could also be used as a guide to the type of asset deemed suitable for implementation by the pool. These and similar types of communication could give funds the trust and confidence that the pool will implement the strategy appropriately.

The defining line between strategy and implementation
Is it time to reconsider whether lower levels of detail are truly strategy or stray into implementation? It seems to me that in the new pooling world fund decision-makers should not be making decisions on, for example, active or passive, internal or external, geographic exposure, asset manager style etc. Properly resourced pools can – and should – make these decisions. While today we are a long way from this ideal, this has to be the direction of travel.

Let’s say for the sake of argument that a fund has chosen an SAA with 45% listed equities and 10% Private Equity. I believe all the subsequent decisions are implementation issues to be made by the pool where the pool has a management organisation capable, resourced and regulated to do this. I believe this is the government preferred pool model. Government has stated that “Investment strategy should be interpreted to mean a broad instruction regarding asset classes and level of risk. It should not include an excessive number of classes or a choice of specific assets.” Of course, any half-decent lawyer would happily confirm (for a small fortune) that this statement allows any number of valid interpretations of “broad”, “excessive” and “specific assets” to name but three potential loopholes.

Perhaps the concurrent government drive, overwhelmingly (81%) supported by pooling consultation respondents, to get actual LGPS industry asset allocations on a more transparent and intelligible footing via fund annual report and the Scheme Advisory Board may provide some guidance as to the sort of asset classes deemed high-level.

The pros and cons of pools advising on strategy
A fund can have multiple strategy inputs if it wishes, and there is now a wider range then ever of firms (and individuals) who offer this service. A definite and very welcome LGPS industry trend is for funds to employ whichever strategy adviser(s) they believe to be most appropriate and add most value. This may not necessarily be the normal retained investment consultant. This trend nicely echoes the same rationale that consultants themselves used to push funds from traditional balanced (multi-asset) managers to specialist management: “no one asset manager is equally good and/or best at each asset class”. Russell Investments famously used an athletics analogy by noting that no decathlete holds the world record in any of the individual events. Rephrase that as “no one consultancy is equally good and/or best at actuarial, investment consultancy, setting strategy, manager selection etc”. I always encourage funds to choose the best available input.

Government appears particularly keen to further broaden the range of fund strategy advisers, for example by encouraging pools to offer this service. This service can only be offered by the pool where the pool entity has appropriate FCA regulation. Only one pool, LPPI (Local Pension Partnership Investments) currently offers this service to their three partner funds/clients.

In section 10 the government states that it will amend regulations to require funds to set objectives for their investment consultants. This confirms its commitment to the CMA (Competition and Markets Authority) Orders (July 2019) in respect of setting and evaluating consultants (strategic) advice. Note in this respect that consultants include IIAs (Independent Investment Advisers) who advise LGPS funds.

Improving strategic advice
The current industry trend to take advice from a broader range of strategic advice suppliers, along with government encouragement to pools to provide this service, and the CMA Order application, suggest to me that both funds and the government believe strategic advice is an area that could be improved, both in the initial setting of strategy and subsequent evaluation.

Section 45 acknowledges that the majority of consultation respondents were opposed to pools actively advising funds regarding investment decisions, including investment strategies. Industry events I have attended have confirmed that – at least currently – funds have little interest in having their pool provide this service. Reasons cited included conflict of interest, regulatory (FCA) difficulty and advising appropriately. Several respondents argued that a fund’s right to seek its own sources of advice was part of its fiduciary duty.

A counter argument put forward was that pools should know their partner funds and can advise on strategies they can implement. Section 51 states that government does not regard pools offering investment advice as a conflict of interest and the fund does not need to embrace public procurement to appoint its pool to provide this service. If the pool has an external operator it can procure investment advice through a separate contractor to avoid any potential conflict of interest. Despite all this the government continues to encourage itspreferred position.

My wish is that this article raises the profile of the significant high-level strategy vs implementation debate and, as a minor corollary, discusses the question of who provides strategic advice. While there may be no unique right answer, the debate is long overdue. The industry is currently embracing a plethora of potentially inefficient answers which are not preferred by government. Greater choice in strategic advice is welcome. With open and honest discussions the currently massive grey area between high-level strategy (by funds) and implementation (by pools) could be reclassified as either black or white.

1. LGPS (England & Wales): Next steps on investment – government response, November 2023. All references in this article to “section N” are from this document.