Active Management: False economies?

An influential consultancy tasked with finding savings in the UK’s local government pensions scheme has put forward the idea of pooling its funds into passive investment. Brendan Maton looks at the issues and the sector’s response

UK taxpayers could be saved £660m (€850m) a year if the Local Government Pension Scheme (LGPS) centralised its asset management. This was the headline finding from a report earlier this year by actuarial and investment consultancy Hymans Robertson.

The savings, which would effectively halve the service costs of the LGPS, would come primarily from two proposals: managing the centralised equity and bond assets on a passive basis; and avoiding funds of funds to access private equity, infrastructure and hedge funds. 

Liabilities and strategic asset allocation would continue to be managed locally across England, Scotland, Wales and Northern Ireland. But day-to-day investment management would be overseen by two megafunds, one in listed securities and one in unlisted.   

Huge membership
LGPS appears huge: £180bn assets serving 4.6m members. By membership, it is Europe’s biggest funded retirement scheme. But its management is highly devolved across the UK, with 89 local authorities (plus a few other government bodies) responsible for everything from strategic asset allocation to monthly pension payouts. Fund size varies from £15bn to £400m. 

Despite the devolution, administering authorities frequently end up duplicating investment mandates. Just 10 external managers run half of total assets, according to the trade union Unison, while a handful of consultancies, of which Hymans Robertson is the most popular, advise the majority of the 89 local authorities.  

Amid this division, only the rules of the scheme are currently centralised. Any fundamental changes have to be set by parliament. But for a number of years now, national government has wanted more: proposing that a cut in the number of administering authorities, or at least pooling the assets they steward, would both bring economies of scale and permit the scheme to act with the clout of a global investor. 

The Hymans Robertson report focused on the first topic: economies of scale. The clout a global investor might have – in terms of influencing investee companies, co-investing with megafund peers abroad or beefing up internal resources – is not much investigated. In the report’s own words, its remit was to “quantify the potential for additional cost savings across the LGPS using the best objective evidence and data currently available and to assess how those cost savings might be accessed most readily”.  

The biggest and easiest cut would be to eliminate active management and instead pool securities management in index tracking. Although it is reckoned that this would mean £215m in transition costs, the annual saving would be £230m, should all the assets come together. A further £190m would come off the annual bill as a result of lower transaction costs from running assets passively. 

The third great cut would come in alternatives, where an estimated £240m a year could eventually be found if fund-of-funds structures in the likes of private equity, hedge funds and alternatives were avoided. This saving would take longer to realise than the first two

because Hymans Robertson foresees the need to unwind any existing fund-of-funds arrangements in due order and not incur penalties for unnecessarily hasty exits.  

Regional resistance
To put all the above in place would theoretically not require primary new legislation. (Hymans Robertson does model an option where the number of administering authorities falls to 10 or fewer, which would involve central government. The immediate savings would be lower, although the quality of governance and thence returns might eventually rise – the report does not investigate governance.) In practice, reform would not happen without legal change – at least not to achieve the kind of savings in the headlines quickly. And most regional pension officers are against the proposals. As one local government wit put it: “Turkeys don’t vote for Christmas.” 

The government cannot simply direct local pension funds to invest in particular assets. But it can close or merge them by law. 

Widespread rejection would have a major impact on the figures. If 10% rather than 100% of actively-managed equities and bonds are switched to passive, Hymans Robertson estimates that the aggregate bill would fall by only £30m a year. Likewise, if only 10% assets in alternatives fund of funds are withdrawn, then an estimated £20m a year would be saved.  

The largest few administering authorities could probably achieve £20m of savings under their own steam, with no need for nationwide agreement or a massive transition programme – involving £120bn of bonds and equities, should the full savings be sought. (A quarter of LGPS equities are already passively managed; much of the 6.7% of assets in property would not be affected; and the alternatives would take a decade to centralise). 

So the £120bn question now is whether this report somehow gets transformed into government policy and thence reality, or remains merely a useful reckoner of costs that sits on council office shelves from the Isle of Wight to Lerwick for the next decade. 

In either case, the answer will have much to do with which set of estimates one wishes to select – not just from the analysis by Linda Selman and John Wright, the authors of the report from Hyman Robertson, but from all the respondents. No one knows exactly what the savings will be. 

For example, on transaction costs, the figure   Hymans Robertson quoted in the report of £190m is for 2012-13 from equities only. Had the data been averaged over five years to 2013, then the saving rises to £240m. Should market impact be included, this brings over £50m extra a year. Then there are costs on bonds and alternatives.  

Not only are the latter not estimated, but the report does not include performance fees from alternatives managers. This summer the UK’s industry-wide scheme for railways, Railpen – which is not an LGPS authority – revealed its total costs to asset managers were up to five times the headline fees. Railpen has greater exposure to alternatives than almost all LGPS authorities (Clywd might be an exception). But the ratios are worth bearing in mind because they are so rarely reckoned; there is currently only one country in the world where asset owners are required by their regulator to know such detail on total costs, and that is the Netherlands. 

In the debate on active versus passive, Edinburgh-based Baillie Gifford, one of the most popular investment managers among the LGPS, reports it has earned 35 percentage points more than index performance, net of costs, for all its mandates longer than five years with English and Welsh authorities aggregated – and that each discrete mandate has outperformed. That is over £1bn that switching to passive would have forsaken.  

Baillie Gifford adds that savings on turnover costs in the Hymans Robertson Report are estimated at 0.17 percentage points, whereas Baillie Gifford claims its clients’ experience is that “a long-term, low-turnover equity strategy need only cost 0.04% per annum, meaning that around three-quarters of the £190m savings identified by Hymans Robertson could be achieved without giving up the potential for superior investment performance in the future”. 

Many authorities take the same line of reporting their outperformance and then questioning why their local taxpayers should pay more simply because other authorities haven’t managed their investment strategy as well.  

Cambridgeshire County Council does not want mediocrity to dominate the LGPS, as it said in its response: “Numerous academic and empirical studies.… show that the ‘average’ active manager has, after fees, underperformed the index over most periods of time. Supporters of passive management point to this as evidence that active management does not add value. The flaw in this argument is that it refers to the ‘average’ manager. Apply the same logic to other walks of life: the average golfer does not win the Open Championship, the average writer does not pen a classic and the average painter can’t be expected to produce a masterpiece. So why should anyone anticipate high investment returns from the average fund manager?” 

Such criticism has foundations: the Hymans Robertson Report aggregates performance by authority funds to reach an orthodox conclusion that their efforts have not been much better than various regional and asset indices. Viewed top-down from Westminster that might be a case for unification and managing assets in line with the index. Viewed from the regions, the conclusions are shallow. Respondents would prefer the weaknesses of underperformers addressed rather than the success of outperformers lost. 

The London Borough of Wandsworth says that laggards should be merged or suffer intervention. Cambridgeshire calculates that, should returns in the lower three-quarters of authorities be raised by 0.5 percentage points a year, then the ensuing appreciation would comfortably exceed the £660m proposed by the Hymans Robertson report. 

The difficulty with such arguments is that they are less certain that the radical switches proposed in the Hymans Robertson report. Moreover, politicians achieve prominence these days by saving money in ways that do not bother the electorate – and these proposals satisfy that criterion.  

Lack of data
That would still leave the consequences of cutting and centralising. In all cases, the difficulty is in understanding the quality of pension fund management, absent much-needed data. For example, the risk dimension was notably absent in the Hymans Robertson report (much of the analysis for which was actually carried out by CEM Benchmarking). Hymans Robertson admits that its findings do not account for the volatility of underlying managers – that all gets aggregated to the fund level.  

For Peter Wallach, head of the £6.2bn Merseyside Pension Fund, this is an oversight. 

“Without knowing how much risk was taken to achieve results, I think the picture is incomplete,” he says. “Our current active strategy was built to achieve optimal returns. One of Merseyside’s better performers has been minimum variance specialist, Unigestion, which has lowered market beta of the overall equity portfolio.” 

Neighbouring Lancashire makes a similar point. If correlation between separate funds is lost due to nationwide pooling on a passive basis, then “in down markets and as a result of so called ‘black swan’ events, the likelihood and expected size of any systemic LGPS failure (and the potential for the need for central government ‘bail-out’, calling on the implicit Treasury Guarantee which has never yet been tested) is likely to be increased”.  

Steamroller 
Bear markets can be mitigated by different allocations between passively-managed assets, of course. Here again, however, is an important matter on which neither side can provide convincing data. John Finch, senior investment consultant at JLT, responded that asset allocation is “the vital component in the reduction of risk when the behaviour of assets is compared to the behaviour of liabilities”.  

This is a sound argument but Finch goes on to claim that it is clear that research evidence supports the case for dynamic active asset allocation – even though one of the papers JLT cites, by Blake, Lehman and Timmerman, raises serious questions. Wallach at Merseyside says he believes in active allocation but acknowledges that the fund hasn’t managed to add value from this process yet. 

Finch is right that the activity has to be linked to liabilities, but liability management – like so much else about pensions – lies outside the scope of the report by Hymans. Liabilities might yet prove the greatest influence on the eventual government decision. Since 2009, many authorities have gone cashflow negative. A change in the benefits structure has also persuaded some current members to leave. One estimate reckons that one-third of all authorities are now paying out more than they receive in contributions. 

This pain can be deferred, not least because a tricky general election looms next May. But should pension liabilities rub up against austerity budgets thereafter, a new government – of any hue – could be tempted to save cash quickly. Among those in favour of mergers are influential economists such as Michael Johnson at the Centre for Policy Studies on the right, and unions on the left.   

“By refusing to act, the government is failing in its stewardship obligations to scheme members, employers and ultimately the taxpayer,” says Colin Meech, national office for capital stewardship at Unison, the biggest union in local government. “It appears to be favouring the investment managers and other contractors who continue to enjoy rising income from our members’ pension funds.” 

Local authority pension chiefs still believe government is likely to introduce a ‘comply and explain’ policy rather than radical mergers. This would be a weak outcome. Pension schemes in the UK have already had 13 years since the Myners Review to comply and explain. But perhaps it reflects a lack of energy in Westminster six months before a general election to steamroller the local authorities into submission and establish centralised funds. 

For pension fund managers and analysts, the Hymans Robertson Report only shows the need for deeper investigation of which elements of the LGPS work well on a sustainable basis and why. For politicians, they may simply be resting until after the general election, and the next pressing demand to ‘free up’ a few billion pounds in local government expenditure.