Devising an asset allocation strategy

Asset allocation strategies aim to meet investor objectives based on the interaction of expected returns, volatility and correlations across different sectors.

Average sector returns are known as ‘beta’. If an investor can achieve superior returns to the market average, then these returns are known as ‘alpha’.

Future expected returns

Past market returns (beta) are a function of previous pricing levels, leasing conventions and growth rates achieved in past economic conditions. Future returns will depend on current pricing levels and growth rates achieved in economic conditions and on leasing terms in the future.

Market returns for asset allocation should therefore be projected from current price levels, using assumptions for future growth, letting periods, tenant rollover rates, costs etc.??As real estate is inherently opaque, projections will be improved if the assumptions are discussed with market specialists in the investment team. Agreeing expectations for each of the drivers of return also engenders a sense of ownership from the investment team and ensures an understanding of what expected returns, and therefore the portfolio plan, are based on.

An additional benefit to estimating market returns using such ‘bottom-up’ assumptions is the ability to identify and quantify the impact of sources of alpha and incorporate these into the estimates of portfolio risk, such as selecting higher growth property or achieving shorter vacancy periods.

It is important to ensure that higher asset level returns are not achieved through taking higher risk (conversely, or that lower asset returns are delivered from taking less risk).??The portfolio plan can include allocations to more risky types of property such as developments or lower risk types of property such as very long-lease property.?The asset allocation plan should however reflect the return characteristics of such assets. After all, if the properties acquired differ significantly from those used in the asset allocation strategy, then the whole portfolio plan will need re-calibrating. This undermining of the portfolio plan is known as ‘style drift’.

The outcomes of each of the assumptions should then be tracked against expectations, creating a ‘feedback loop’ to detect unexpected outcomes and also documenting the sources of the alpha achieved. This institutional knowledge can then be fed-back into the process to improve the process and the modelling.

Volatility of returns

Just as past market returns are a function of previous pricing levels, leasing conventions and growth rates achieved in past economic conditions, so was the volatility.

To estimate the future volatility of sector returns an upside and a downside scenario, defined as +/- one standard deviation respectively, can be projected. Where available, the calibration can be made on historic rental growth and vacancy periods adjusted for any structural changes in occupier or investment markets.??The range from these upside and downside scenarios can then be used as an estimate of future market risk.

Accounting for specific property variations

Individual properties will perform differently to their sector average based on the outcomes of lease events and the natural variation in growth and letting periods across properties.

The more assets in a portfolio the more diversification will neutralise this asset specific or unsystematic risk. The residual specific risk can be estimated using the number of properties held in each sector.

Of course, it is possible that concentration risks emerge in a portfolio, such as a cluster of lease expiry dates, which will reduce the benefit of diversification and raise the exposure of the portfolio to a particular tenant, year or location. These concentration risks should be monitored as they can occur unexpectedly through the normal management of a portfolio.

The RES Asset Allocation Model

The RES Asset Allocation Model quantifies expected returns from both beta and alpha sources.

The projected returns are based on current pricing and use transparent ‘bottom-up’ assumptions for growth, capital costs, vacancies, irrecoverable costs and price movements to project returns from a ‘risk-neutral’ property.??The assumptions should be agreed with an investment team and also used throughout the asset analysis process to ensure alignment of investments made with the asset allocation plan.??The specific risk is incorporated in the portfolio risk estimate based on the number of properties held in each sector.

The volatility of future returns is estimated based on current pricing and lease terms, using estimates for the future volatility of growth, vacancy periods etc. Unless there are thought to be reasons not to do so, these assumptions can be based on historic outcomes.?

Using these projected returns and risks, the RES Asset Allocation Model calculates the likelihood of achieving an investment objective or of falling below a particular level of return.

Alternative portfolio structures can then be compared to the current portfolio.