Managers talk of outperformance but how do investors identify whether this is the result of genuine skill rather than pure market exposure? Andrew Baum and Kieran Farrelly report
Since the mid-1990s there has been huge growth in the aggregate size and number of global property funds in both listed and unlisted formats. Fund managers have been able to raise significant capital, particularly for unlisted funds which potentially reward them with performance fees without the manager necessarily being able to provide clear evidence of out-performance against any market benchmark.
In a more challenging, mature, and increasingly transparent market this is unlikely to continue to be the case. It will become increasingly possible to assemble performance records, and following this there will be more detailed analysis of those records. Potential performance analysis systems will include the traditional attribution methods employed by Investment Property Databank, but will also cover the risk-adjusted performance concepts of alpha and beta widely used in other asset classes.
By employing the capital asset pricing model (CAPM) the impact of manager skill or alpha can be estimated. A higher risk portfolio should out-perform a lower risk portfolio but this does not itself demonstrate any skill. Alpha measures the excess return earned by the manager over and above that required to compensate for the risks being taken.
In property fund management, managers can exercise skill when structuring their portfolios from a top-down perspective (allocating to markets and sectors) and at the stock level (sourcing and managing their assets). Outperformance from portfolio structure is delivered by managers who, ceteris paribus, allocate relatively more to outperforming sectors or geographies. This implies that the manager has a forecasting capability which is their source of edge in this respect.
Outperformance at the stock level is very different to that of traditional securities fund management, and this is due to the ‘private equity' characteristics of property. Properties are selected by investors/owners and require ongoing asset management which encompasses a number of activities. Alpha in property management can thus arise from operational cost control, tenant relationship management, asset maintenance, leasing strategy, marketing and asset enhancement/refurbishment via capital expenditure. Alpha can also be generated when assets are bought and sold. Managers, who are able to purchase assets at discounts, recognise latent value, negotiate attractive prices, and who have the ability to execute more complex deals and thus face less competitive pricing, will, ceteris paribus, outperform.
Property investment risk (beta), like alpha, can be broadly separated into both portfolio structure and stock. Within the constraints of a domestic benchmark ‘structure' beta arises from allocations to more volatile sectors such as CBD office markets. When mandates allow for global investment, exposures to more risky geographies such as emerging markets are then a source of additional risk. Defining structure risk in a purely quantitative manner is difficult in some situations because certain aspects are difficult to quantify. For example, differences in transparency and property rights may not be reflected in the relative performance of market data.
Stock level beta is an area of potential confusion. For example, development can often be referred to as a source of alpha in a given portfolio. This is incorrect, as development in itself is a more risky property strategy and should be reflected by a higher beta, whereas development alpha is obtained by outperforming development managers. Thus there is a continuum of asset level risk ranging from ground rent investments, to assets with leasing risk and high vacancy, to speculative developments, all of which should have a hierarchical range of assigned betas.
Fund structure is a factor specific to property held in a vehicle or wrapper. This factor will have an impact on the returns from listed property companies, and from unlisted funds, alike. There are two main drivers of the fund structure impact: fund expenses and management fees; and leverage. Pure leverage is a beta (risk) generating activity. Expenses and fees simply limit the impact of that beta contribution.
Finally, closed-ended unlisted funds draw capital from investors over a period of time that could be as much as three to four years. The timing of the drawdown is within the manager's control, meaning that using the money-weighted return or IRR approach is appropriate for performance measurement. However, benchmarks are typically calculated on a time-weighted basis, therefore the difference can be attributed to investment timing and fund drawdowns.
Thus a four-tier return attribution framework for property funds can be used which attributes performance to portfolio structure, stock, fund structure and timing. None of the above is intended to suggest that isolating and measuring alpha or beta will be easy or non-controversial. The choice and/or availability of benchmarks, in particular, are limiting factors. Judging whether greater risk is being taken at the portfolio structure or stock level will be a matter of opinion and it therefore requires a pragmatic approach.
We examined a fully liquidated closed-ended value-add UK-focused unlisted fund, which employed around 60% financial leverage over its life. Compared with the IPD Index the fund produced relative outperformance, at the direct property level, of 1% per annum over a five-year period. The manager under-performed due to portfolio structure by almost 2% per annum, but outperformed due to stock selection. The fund had concentrated segment exposures, thus we could conclude that the manager had outperformed by concentrating in segments where they had particular expertise.
The CAPM model showed an alpha of zero and a relatively high beta. This suggests that much of the outperformance via specialisation delivered was a result of incremental risk-taking or higher beta. The high beta reflects the level of gearing at the fund level, and the asset level and portfolio structure risk.
Other studies in this area have also been unable to find significant and consistent alpha, except in a very small number of cases. However, we would also add that executing higher risk or value-add/opportunistic strategies is a skill in itself and finding managers that can demonstrate that they can do this consistently is not easy.
We would suggest that a full analysis of fund performance will require the use of both performance attribution approaches and both a professional and pragmatic interpretation of the results. Just understanding how a manager delivered past performance is a crucial aspect of any prospective fund's due diligence process. For instance, "did the manager simply benefit from the market return or did they execute their stated business plans?" is an important question that needs to be answered in assessing the capability of a particular manager.