The Use of Benchmarks for Real Estate Portfolio Performance by U.K. Financial Institutions

Abstract

The paper examines the application of benchmarks in the United Kingdom primarily through the use of semi-structured interviews with 17 major investment houses, holding domestic real estate assets under management of nearly £180bn with in excess of £515bn in other countries. The MSCI/IPD database is the predominant reference point for peer and relative benchmarking, but funds also apply an absolute benchmark approach. Many fund houses indicated a reluctance to change benchmarks. However, increasing short-termism and the transformation of market fundamentals after the GFC have led to a reappraisal of the nature of existing benchmarks and there is a continuing move toward use of alternatives.

Keywords:

• Benchmarks
• portfolio construction
• indices
• global financial crisis
• UK

CASE STUDIES

Case Study 1: The Benchmarks of a Long-Established Unit-Linked Property Fund

Case Study 1 is of a unit-linked property fund formed in the early 1980s. Over its first 37 years, it has been managed by only three individual fund managers. All three were interviewed for this case study. The fund was subdivided internally into a life assurance fund and a pension fund. From the property manager’s point of view, the fund was regarded as a single fund with a single cash flow. For the first six or seven years, when benchmarking was still in its infancy, the fund had no official benchmark. Attempts at benchmarking used the marketing department to research competitor funds’ unit prices and compare them with the fund’s holdings. That comparison focused more on performance rather than any true benchmarking exercise.

Benchmark A

Formal benchmarking arrived with the establishment of IPD, and the fund was the second one to be placed in the IPD Universe. The first genuine benchmark was therefore adopted around 1987–1988. It used the embryonic data from the newly published All Funds Index from IPD and was known as All funds in IPD Annual Index.

Benchmark B

From 1990 a more rigorous approach was applied, aided by the rapid acceptance of performance measurement by institutions and of IPD in particular. It was then possible to construct a meaningful benchmark that combined funds of a similar size and of similar investment objectives, namely, All unit-linked funds in IPD Annual Index between 50% and 150% of the value of the fund assets. At this time the fund’s benchmark report was not published until the end of the second quarter of the following year, and pressure built up from both the fund management side and the marketing/risk teams for a more frequent comparison. Consequently, the practice grew informally of using both the IPD Monthly Index and the monthly unit price performance (from the likes of Lipper or Statspack) as a proxy for assessing the fund’s performance during the year.

Benchmark C

By the end of the last century, the unit-linked universe and value of the funds being captured and measured by IPD was sufficiently large to enable the targeting of a more relevant benchmark. As a result, a further change occurred in 1999 as the fund continued to increase in size, when the benchmark was changed to All unit-linked funds in IPD Annual Index. However, Benchmark C still suffered from the fact that the index was published only once a year. The fund manager, the marketing department, risk departments, etc., were all seeking more frequent comparisons. The IPD Quarterly Index that was about to be launched would address many of these concerns.

Benchmark D

In the late 2000s, the fund moved to its current set of three benchmarks: For direct property, the IPD All Funds Quarterly Index; for UK shares (REITs): the EPRA/NAREIT UK index; and for cash, the cash index. The change in the benchmark was driven by the change in investment policy brought about by the global financial crisis, notably the need to have more liquid assets to cope with sudden redemption requests. Although the three separate components were appropriately benchmarked, the overall fund was not, meaning that, while all three asset components could outperform their benchmarks, the overall fund may under- or overperform at a unit price level. Overall fund performance (the unit price) continued to be monitored informally through Lipper.

The change in investment policy and the consequent change to the choice of benchmark to address liquidity issues implies that investors in the unit-linked fund indirectly influenced the choice of the benchmarks at that point. Notwithstanding this, no other benchmark change over the years came about through investor influence.

Case Study 2: The Benchmarks of a Balanced Multi-Asset Fund

Case Study 2 is of a balanced institutional portfolio within a substantial multi-asset investment house that has invested in commercial real estate over a long time period. This real estate portfolio has been in operation for decades and is a segregated mandate that accepts no external investment. It retains substantial holdings across the three traditional United Kingdom property sectors of retail, office, and industrial. The portfolio also invests directly in some of the more recently established alternative sectors. The portfolio now sits within an expanded real estate offer that includes retail investment products, and specialist and international investment vehicles.

In the course of this study, the authors have spoken directly with members of staff who had responsibility for the development and implementation of benchmarking policy, leading towards its formal establishment with the adoption of IPD in 1985. Prior to that date monitoring was very rudimentary, comprising crude cash flows of the portfolio. The fund has participated in IPD benchmarking since its foundation. Since that time the portfolio has had only five managers, which has contributed to the stability of the fund with consistent preference for relative benchmarking.

Benchmark A

From the outset until 2003 the benchmark applied was the returns of Non-unitized life funds in IPD.

Benchmark B

By the late 1990s and early 2000s, the miss-selling ¹ of endowment policies produced serious challenges for life assurance companies, particularly in relation to their with-profits business. These policies proved not to be value for money and fell out of favor with the public (Jones et al., 2017). There was also increased competition driven by the establishment and success of lower-cost unit-linked vehicles. This led to a major reduction in the number of non-unitised life funds, so the benchmark was redefined in 2004 as Life funds greater than £1bn in the IPD database.

Benchmark C

In a similar vein, the benchmark had to be adjusted again to reflect changes and consolidation in the market after the global financial crisis. As a result, the benchmark was changed to extend it to encompass pension funds while reducing the threshold size required. In addition, one IPD market segment was deleted from the benchmark because the view was taken that the sample contained in that data was not representative of that particular subsector. The benchmark was then defined from 2011 as Life & Pension Funds >£500m, with specific benchmarking data removed from the IPD database.

Benchmark D

In 2015 a decision was taken to remove the exclusion of benchmarking data on the subsector agreed upon when implementing the previous change in 2011. The benchmark therefore was simplified to Life & Pension Funds >£500m.

Benchmark E

In the middle of 2018, the benchmark was changed to the IPD Quarterly Universe. The principal driver behind the change was again the declining sample of the database.

Notes

1 Mis-selling is a British term that refers to selling a product on the basis of misleading advice.