Three-way process controls risk and maximises retu

The Stockholm-based AP3, part of the Swedish public pension system with €15.7bn of the system’s buffer fund capital under management, has adopted a three-way approach in translating its liabilities into a concrete portfolio structure with detailed risk mandates. This comprises looking at asset-liability modelling (ALM), the fund’s manager structure and its risk budgeting. AP3 says this process allows it to understand and proactively take control of return potential both in terms of market and active returns as well as continually monitor risk.
The main decision-taking stage in the process is the ALM, as this enables AP3 to determine not only the weightings per asset class and currency hedge ratios but also the overall level of risk. By undertaking extensive market simulations and stress testing, AP3 says it understands better how the way the portfolio is constructed affects the pension system. But simulation results have an inherent element of uncertainty, caused by the finite sample size of simulations and uncertainty of their variables, such as asset price, covariance assumptions and demographic developments. AP3 says it therefore prefers to think of strategic asset allocation in terms of optimum weight ranges. Apart from the core asset class weighting determined by the asset-liability modelling, AP3 also sets fluctuation limits to allow the weightings to vary over time.
The ALM process itself can be divided into two intertwined steps. Firstly, AP3 takes a long-term equilibrium approach to determine the average strategic allocation over time and secondly, asset pricing considerations in the medium term, that is, between three and five years and lead to a deviation from the long-term average. Such was the case in 2002 when Japanese bonds were excluded from the benchmark and the decision was taken to maintain a high hedge ratio with respect to US dollar exposure because of its weakened position.
At manager structure level, AP3 uses the abstract benchmark portfolio as a starting point to define the portfolio’s eventual microstructure. Whereas the ALM focuses solely on overall market return and risk – that is, beta and its covariance structure – manager structure deals with the analysis of active return, namely alpha and tracking error. AP3 says the two structures are never fully integrated even if both are open to the same methodology. Instead, the two sources of return are analysed in separate but interrelated parts of the investment process with different time horizons. Whilst theoretically getting paid for carrying long-term beta risk can be justified, including a skill-based, if not transitory, alpha in the long term is less convincing. AP3 prefers analysing alpha in the manager structure with a medium-term, two- to five-year focus. The manager structure process is designed to lead to a portfolio that maximises the fund’s long-term return potential within the framework determined by the ALM. The process is based on a thorough analysis of the market structure to identify areas of potential active return and to determine typical risk information ratios of skilled managers in submarkets. This will lead to the portfolio being divided into actively, semi-actively and passively managed components to be managed either inhouse or externally. Overlay mandates will also be defined at this stage. AP3 highlights US equities as a means of illustrating this procedure. It says natural subdivisions of US equities are based on market cap. By analysing various segments, one might conclude that large caps should be split into one passive and one or two enhanced mandates to be managed externally. In the mid-cap segment, it might be that two external managers with a certain style mix and a tracking error of 5% might be the best way forward. A similar process is used for determining the number of small cap mandates, though not who will actually manage them, with a broader mandate definition with respect to risk/information ratios.

The risk budgeting step represents the stage where the manager structure is consolidated by the selection of asset managers and where risk limits and returns targets for the mandates are determined. This includes a process that sets and monitors current active return targets and risk for each decision taken concerning the construction of the portfolio. The objective here is to maximise active returns within risk limits by allocating risk to those mandates that reveal the highest active return potential.

Highlights and achievements
AP3’s proactive stance in maintaining control over transforming its assets and liabilities analysis into a well-constructed portfolio and continuous risk management has led to a winning three-way formula being adopted – asset-liability modelling, asset manager structure and risk budgeting.
These are intertwined and AP3 says the exercise has given it has a clearer understanding of how portfolio construction affects the pensions system overall.
AP3’s flexible approach to strategic asset allocation by setting fluctuation limits to its core weightings and its recognition that many different factors, not just financial, can have an adverse impact means it successfully maintains its long-term profitability.
Looking at alpha and beta separately, even though the two may ultimately share the same methodology at different asset management institutions, ensures AP3 implements the best manager structure it can with respect to the results of its ALM studies.
Developing a process that considers the risk for each individual asset investment decision means AP3 remains abreast of risk exposure and can effectively consolidate its manager structure whilst allocating risk in the best possible way to ensure a high return potential.

Have your say

You must sign in to make a comment


Your first step in manager selection...

IPE Quest is a manager search facility that connects institutional investors and asset managers.

  • QN-2444

    Asset class: Trade Finance.
    Asset region: Global.
    Size: USD 10m.
    Closing date: 2018-06-25.

  • QN-2452

    Asset class: Hard Currency.
    Asset region: Emerging markets.
    Size: USD 300 mil..
    Closing date: 2018-06-22.

Begin Your Search Here