Stephen Ryan argues that non-listed real estate is a good match for DC pensions
Occupational pensions are growing. Defined contribution (DC) assets are worth €17trn and represent 48% of the world’s pension assets, according to Willis Towers Watson. However, DC is growing faster than defined benefit (DB), and in time will dominate occupational pension provision.
However, there are three main obstacles facing non-listed real estate in a DC world: liquidity, high cost (actual or perceived) and product availability. All three obstacles can be overcome, often by blending non-listed real estate with other asset classes.
In the US, the preferred solution is the daily priced fund of funds. In the UK, there is no consensus, but the approach taken by the government’s National Employment Savings Trust (NEST) – 70% domestic core property and 30% passive exposure to global REITs – has clear attractions. Australia relies on high net cash flows and an in-depth understanding of both plan cash flows and market illiquidity. In Ireland, Denmark, France and the Netherlands, non-listed real estate is blended into lifestyle strategies in different ways.
The benefits of investing in non-listed real estate – low correlation, high and stable income, low volatility, inflation-related returns, high total returns and efficiency – are as relevant to DC plans as they are to other investors. For DC, specific benefits are needed at different points in the plan member’s life:
• Younger members of DC plans need high total returns;
• Older members of DC plans need low volatility;
• Retired members need income and inflation-related returns.
DC investment strategy is closely tied to the member’s age. This explains why a lifecycle approach is usually adopted for the DC default investment strategy, which accounts for the vast bulk of DC assets. Lifecycle means that a member’s savings are invested aggressively when the member is younger and retirement is far away, and progressively more conservative as the member nears retirement. Lifecycle is a risk-management tool. Figure 1 shows an example.
Nobody really likes risk. Risk is the price that investors pay to get returns. In the world of pure DC, investment risk lies with the plan member, not the employer. There is no risk sharing in DC, which sets it apart from DB. To use a simplistic analogy, DC is like savings while DB is like insurance. Non-listed real estate combined with lifecycle helps reduce risk. That is why this combination can be found in DC plans in the UK, the US and in Europe.
The £1.8bn (€2bn) NEST sees potential in real asset classes, with real estate a good fit for its young membership, having strong positive cash flow. The bulk of NEST’s assets are in growth-phase funds, targeting annual investment returns equivalent to inflation plus 3%, after all charges.
The key reasons for NEST’s significant allocation to real estate of up to 20% are:
• To improve portfolio diversification;
• To provide a partial inflation hedge, which suits NEST’s inflation-plus return objectives;
• To take advantage of the illiquidity premium in direct real estate at a time when markets are potentially too sensitive to this issue, in the full knowledge that NEST will be cash-flow positive for the foreseeable future.
The real estate allocation is implemented through a 70% weighting to the UK direct commercial real estate market (via a non-listed core fund) and a 30% weighting to the global listed real estate sector, which is passively managed. This element tracks the FTSE EPRA/NAREIT Global Developed Real Estate index. When opting for the 70/30 ratio, NEST considered several factors, including return enhancement, liquidity, costs, risk-adjusted returns, tracking error, currency impact and weight of real estate in the overall portfolio. The non-listed and listed elements are run by the same manager. The fund is not available to NEST members on a standalone basis.
In the US, DC strategies account for about 60% of total pension assets. The most popular default strategy is the target-date fund (TDF), where members with the same target date for retirement are grouped together in the same fund. The Defined Contribution Real Estate Council (DCREC) estimates that about $25bn (€20.8bn)-worth of non-listed real estate assets are invested in TDFs across more than 100 DC plans in the US.
DC plans need liquidity but there is a clear difference between the routine liquidity associated with regular tasks (such as investment of monthly contributions) and the extraordinary liquidity associated with ad-hoc events (such as manager terminations, strategy changes and plan mergers). Routine liquidity is largely predictable in terms of volume, while extraordinary liquidity is unpredictable but tends to be signalled well in advance.
The typical DC-investment strategy has two sections: the default strategy and self-select funds, which the members can choose for themselves. Within the self-select section there may be some pre-mixed, multi-asset funds, often with risk profiles such as ‘conservative’ or ‘high risk’. There may also be some building-block funds (equity funds, bond funds, and so on ), which more confident members can use to blend their own portfolios. Non-listed real estate can add value in most parts of the typical DC line-up. Figure 2 shows how.
Non-listed real estate works best for DC when embedded in a multi-asset strategy, whether that is the plan’s default strategy or a self-select ‘pre-mixed’ strategy. The multi-asset setting makes liquidity management easier. It also leads to lower overall total expense ratios, if some of the other assets are managed on a low-cost passive basis. Lifecycle strategies are a suitable place to house non-listed real estate. Within the broad category of lifecycle strategies, target-date funds are more flexible than lifestyle strategies. Open-end core funds, with either a domestic or international strategy, are a good fit here.
In conclusion, there is a case for non-listed real estate to refine and reinforce its role in DC, as this would be beneficial for plan members and managers. Product innovation is necessary, and some useful product blueprints are already available.
Stephen Ryan is research manager at INREV