How the old order is changing
Consider these two quotes: “After five years of underperformance the UK’s ‘big five’ fund managers are losing pension business to their specialist rivals”, and “Merrill Lynch, Schroders, and Gartmore are among a group of the world’s best-known fund managers who suffered a dramatic fall in their UK pensions business last year”. The picture they paint is clear; the ‘stranglehold’ of the big five is over and the UK pension management industry is becoming more fragmented and more like the US model, with new entrants to the market place, boutiques springing up, and innovative approaches aplenty.
Or is it? Let’s look back at the two quotes. Firstly, they were written nearly two and a half years apart – January 2000 and June 2002. Both were by Simon Targett, writing in the Financial Times. Furthermore, the accompanying table in the more recent article showed that almost half of the asset class covered (UK segregated pension funds) continues to be managed by just five firms. Indeed three of those five firms, MLIM, Schroders, and UBS occupied first, second and fourth positions respectively as long ago as 1985, albeit in two cases under different names. To those outside this charmed circle, changes in the UK industry’s structure must look glacially slow and frustratingly belated.
So what is really happening in the UK market? It turns out it is surprisingly difficult to assemble good longitudinal data, collected on a consistent basis over an extended period, and which for a bonus is easily comparable with the US market. That said, there is both statistical evidence and some logic to the argument that the UK’s Big Five are losing market share, at least to the next layer of managers, if not to start-ups.
The Financial Times publishes from time to time a table ranking the top managers of UK segregated pension funds. Davies and Steil1 have collated data for four separate years between 1985 and 1998, which we can easily update for end 2001, using the latest report in the Financial Times of 3 June, 2002. The evidence appears to be that the peak concentration of assets in the middle to late 1990s is definitely behind us, with the industry structure back to where it was in the late 1980s(2).
Perhaps the most surprising aspect of the table is how the market shares have changed since end 1998. The top five appear to have ceded 16% of market share to the next 15 in the space of three years. Previously, the moves between these two groups were of the order of 1% to 2% per annum. In money terms, the top 20 firms managed £373bn (E580bn) in segregated accounts at end 2001. It is not fanciful to imagine that the figure was similar three years ago, at the end of 1998. If so, the big firms lost something like £20bn per year to the next group.
The loss of market share was accompanied by, and indeed caused by, performance problems over an extended period for all members of the Very Large Investment Manager (VLIM) group. It is not obvious why, given their different styles, all four active managers ran into performance problems at roughly the same time. The exception is BGI, which is predominantly an index fund manager. Perhaps the performance problems were predictable for the active managers; as they took market share, they eventually became locked into their positions, and simply ran out of liquidity. As an index manager, BGI’s need for liquidity was not so great. Certainly UBS Global Asset Management, the successor firm to PFDM, has worked hard since then to manage its market impact and transaction costs more effectively.
The pain of widespread poor relative performance has occasioned some serious finger-pointing. Many think that the investment consultants deserve criticism. Did they not advise the appointment of the already big managers, naively extrapolating recent good numbers while ignoring or underestimating the issue of each firm’s capacity to run huge amounts of UK equity money on a consistent basis? Julia Hobart of Mercer Investment Consulting believes the blame is overdone and misplaced. In her view the consultants are not to blame, as they would often put forward medium and smaller-sized managers, but trustees in the early and mid-nineties found the combination of previously good performance and the comfort factor of a recognised name to be irresistible, and understandably so.
George Urquart, executive director of WM Company, the Edinburgh-based performance measurement company, is not so sure: “The majority of trustees are heavily influenced by their investment consultants.” He points out that although the consultants insist that past performance is not a (or even the) major part of their manager selection methodologies, it is very rare to see a top performing manager removed in favour of a less well performing manager. He notes further that only now and only in part is the manager selection skill of the investment consultant to be subjected to quantitative and independent third party analysis. (Mercer’s buy-list is to be audited by WM).
Industry structure however is rarely simply a question of strong brands and good performance, and there are additional reasons why the VLIM group may face a ‘disappointing” future. In 1990, barely 5% of UK pension funds had a sensible benchmark. All but a small minority of funds measured their managers against the peer group. In spite of major differences in the nature of their liabilities, and in the strength of the sponsor’s covenant, trustees often ignored, ostrich-like, the individual character of their own fund in favour of a “one-size fits all” investment policy. Today, that has changed; over 60% of pension schemes now have a properly specified benchmark. This at least gives the trustees and their advisers the opportunity to appoint a different manager for each asset class, or indeed several different managers for each asset class, one of whom may be an indexer. Thus the demise of the peer group benchmarks favours specialist managers who can demonstrate excellence in a single asset class, who previously would stand little chance of getting on to the short list for a balanced mandate. Looking back, the early and middle 1990s were perhaps a transition period, where the benchmark arrangements were changing, but not yet the choice of managers. Thus a couple of ‘big five’ managers continued to manage the assets, but against specific benchmarks, not against the peer group average.
These changes mimic the evolution of the US market with a delay of a couple of decades. The often higher levels of education among its pensions officers and the greater resources of the typical US pension plan have long facilitated a more fragmented and some would say more competitive investment management industry. The US is of course a much larger economy, and as the Myners Review points out, there are subtle but important differences in trustee law, which put greater burdens on trustees in the US to become familiar with investment matters. All these factors will shift power and influence away from the suppliers of investment management services towards the consultants and the clients themselves, and in doing so, will tilt the balance towards a greater fragmentation of the industry. Innovations such as Russell’s and other’s manager-of-managers approach both encourages and is encouraged by the fragmentation we are beginning to see here. A manager-of-managers is surely after all going to be far less influenced by brand and reputation?
Industry structure and the reasons for it are of more than academic interest. They inform and help guide the setting of a sensible expansion strategy for newcomers. Many overseas firms tried to establish themselves in the UK market over the last 20 or more years. Only Capital International and Fidelity have become significant pension fund players without making a major acquisition, and in both these cases the expansion has been relatively recent, reflecting the changes in structure seen in recent years. What many of the early would-be entrants did not fully appreciate was the extent to which the trustees’ lack of expertise drove them to a herding strategy, with a preference for conventional solutions(3).
So how does the US compare with the UK figures in our table? Calculated in the same way, taking the top 20 managers of US pension assets, the Herfindahl index stands at 0.07, and the ‘numbers equivalent’ figure is therefore 14, just as in the UK. But there the similarities cease. Only two managers, SSgA and BGI, have more than 5% market share, consisting of a mainly indexed $1trn (E1.5trn), about the value of all UK pension assets. That leaves more than 500 other managers to fight over a further $6.6trn. The landscape is different.
1 E Philip Davis and Benn Steil, ‘Institutional Investors’, MIT Press 2001
2 The FT data relies on segregated funds only. However, this may cause an understatement of changes in the industry’s structure. Legal & General alone now has over £60bn in pooled pension fund indexed funds, some of which came from the disgruntled former clients of the Big Five, but which is not reflected in this analysis
3 See Chapter 6 of Gordon Clark’s ‘Pension Fund Capitalism’ Oxford University Press 2000. He explores trustee decision making as influenced by three frames of reference – habits of prudence, rules of proprietary conduct, and norms of relationships – which collectively favour conservatism. Davis and Steil in the earlier quoted work, believe that this is especially the case where there is (or was) a more severe asymmetry of expertise, as in the UK