In-house management can give pension funds greater flexibility and control than they might possess by outsourcing. External managers might be obliged to stick to the same house style through thick and thin, and trustees still have to pay fees even when investments underperform.

But moving from a simpler overseeing role to managing money actively is a huge change, warns Nick Sykes, partner with Mercer’s investment consulting arm.

“There are relatively few schemes large enough to be credible managers of money, except through index-tracking or perhaps fixed income stocks,” Sykes says.

“Suppose your external fund manager charges 25 basis points on a $1bn portfolio, that would be $2.5m. Could you build a credible management team with that? Probably not. The minimum portfolio size for in-house active management is probably several billion.”

In fact, there have recently been a number of moves in the opposite direction, with pension funds outsourcing to external managers—for instance, because it can be difficult to retain or hire staff with the right expertise.

Either way, there is little point in running an in-house investment team unless its performance is monitored. First a pension fund should be clear about what kind of in-house operation it plans to run.

At one end of the scale, in-house teams may act more as in-house advisers than asset managers, by carrying out due diligence procedures, and selecting and monitoring managers, sometimes with the help of pension consultants.

At the other end, some in-house teams run the entire asset management process, from setting investment strategy and policy, through asset allocation, to picking the individual stocks, bonds and funds.

Some funds such as PKA, the Danish labour market pensions organisation, run part of their portfolio in-house. PKA manages part of its fixed income side investments and some passively-managed segments. This is primarily based on the resources needed to run different asset classes and the availability of information.

There can also be a structural problem for pension funds with an in-house team, in that the fund represents the team’s sole client.

“In contrast, a fund management company runs assets for many clients, with revenues and profits to match,” says Sykes at Mercer. “Given that, why would a pension scheme attract the best in the business?”

However, Sykes reckons a passive portfolio as small as $150m to $250m could be run by, say, one employee working for half the day. As well as buying in skilled managers, in-house teams also need to invest in technical support such as compliance, back-up systems and disaster recovery.

Key management issues, says Sykes, are succession planning and maintaining the quality of the in-house team. “However, trustee boards can tackle these issues by accessing independent expertise in the form of an independent director with an investment background or a consultant,” he says.

Pension funds can monitor their in-house asset managers by setting up similar systems as for external managers, says Jeroen Molenaar, senior investment consultant at Towers Watson.

“There should be a good service level agreement (SLA) in place, which includes the basis for monitoring hiring and firing,” he says. “It should also cover performance, risk, investment style, compliance with guidelines, service and transparency. The board should review the SLA formally at least once a year, at a meeting where it can be candid about the team’s overall performance. It can also carry out more informal, ongoing monitoring, depending on the team’s performance.”

Molenaar says the trustee board should also set up the framework within which the in-house team operates, with clear guidelines on risk budgets and what the team can invest in, besides a definition of “adding value” and instructions as to how this is to be achieved.

“The criteria used by the board should be measurable, so they can score the in-house team against overall performance,” he says. “These should include return and outperformance targets, and also risk parameters.” In terms of performance, measurable parameters for assessing the in-house team should include past performance, expected performance and cost.

For passive management, the board should check whether the team is accurately tracking its benchmark and at minimum cost. For active management, the issue is whether the team’s performance is reasonable, and at a reasonable cost.

One potential problem with using an in-house team is that, as a cohesive unit it is likely to possess more expertise than the board of trustees, and could well run its own agenda, leaving the board to rubber-stamp its decisions. It is also important for trustee boards to create a healthy balance of power via a system of checks and balances.

“For a governance framework to be effective, clear delegation of responsibilities is a must,” says Tony Broccardo, chief investment officer, Barclays Bank UK Retirement Fund (BUKRF).

“There are also a number of checks and controls which should be imposed on the operations of in-house and outsourced teams. This is especially the case for the BUKRF, so the scheme operates under a clear governance framework, which specifies a number of parameters such as risk premia ranges and value-at-risk targets versus liability.”

This allows the investment team to know exactly what they can and cannot do, and manage the fund in accordance with the trustees’ wishes, says Broccardo. “In essence, the trustees should set the agenda on the strategic, big picture, long-term direction of the scheme, while the in-house team is accountable for shorter term decisions, day-to-day issues and investment management,” he says.

PKA’s in-house team must stay within a 30 to 40% limit of total assets for equities, while private equity has to be between 5 and 9% of total assets.

In addition, the fund uses a risk model developed internally, run by a separate risk officer who reports directly to the managing director, not only on the overall risk levels, but also on the composition of risk.

PKA’s team has to perform as well as, or better than, external managers. It reports quarterly to the trustee board, showing comparisons between actual performance and benchmark performance for both each asset class and the overall portfolio.

Any underperformance may be taken up by the board, if it is not just a short-term blip.

With the current popularity of fiduciary managers, however, some funds with in-house management teams are likely to be asking themselves whether appointing a fiduciary would mean dismantling their in-house team.

“If the in-house team is only selecting and monitoring managers, this can easily be outsourced,” says Sykes at Mercer. “But if it is carrying out other tasks such as corporate governance, voting and engagement, there may be a case for retaining those functions in-house if the fiduciary manager is not covering them.”

Indeed, earlier this year PME, the Dutch metal industry fund, which uses Mn Services as its fiduciary manager, was ordered by the regulator to improve its internal governance capabilities, and appointed an internal head of risk.

Barclay’s Broccardo agrees that appointing a fiduciary should not necessarily mean scrapping an in-house team.

“But using a fiduciary manager to work alongside an in-house team may cause more harm than good,” he warns.

“Accountability and ownership can become somewhat murky and difficult to assess, and can be an unnecessary additional control and management level that could lead to more conflicts than solutions.”

Instead, he suggests a model of fiduciary-only or in-house team—plus external advisers or consultants—as a simpler and more pragmatic solution.