Performance measurement and benchmarking has always been a challenge for the property asset class. Rachel Fixsen talks to six investors and managers
Sees improvement in international benchmarks Performance benchmarking lacks focus on risks taken
At Schroders, benchmarks are used for performance measurement in the property multi-manager business, but this is generally only the case for UK mandates and funds. “This is mainly because there are fewer established international benchmarks with long histories and the ones that do exist don’t have comprehensive coverage of all countries,” says Graeme Rutter, head of property multi-manager at Schroders.
Rutter manages segregated accounts and funds of funds investing in unlisted property vehicles, listed funds, joint ventures, property equities and other vehicles. “The international benchmarks tend to have good coverage of the more established northern and western markets but are more patchy elsewhere.”
That said, he sees an improvement in the available international benchmarks, and believes that in time, people might adopt them.
The Schroders team compares its predominantly UK multi-manager mandates to the AREF/IPD UK Pooled Property Fund Index All Balanced Fund Weighted Average, against which it has an outperformance target of 0.5-1.5% net of all fees over rolling three-year periods, Rutter says.
International mandates, on the other hand, have total return targets of 7-8% per annum over rolling three-to-five-year periods.
Even though most of Schroders’ property mandates compare performance relative to a benchmark, absolute return does matter here too. “Our clients wouldn’t thank us if we said we had outperformed the benchmark by 1% but had delivered a minus 10% absolute return, so both are important,” says Rutter.
Rutter sees a danger of disappointing clients if they accept mandates with absolute return targets – unless the monitoring period is very long to cover the extremes of market cycles.
“There is a natural conflict because many clients want the reassurance that their portfolios are outperforming short term,” he says.
One problem with performance benchmarks is a lack of focus on the risk taken to achieve returns. In strong market conditions much of a fund’s outperformance may be driven by gearing, but this could easily be given back as markets decline. “We try to position our mandates to outperform, taking the minimum risk possible and with minimal volatility,” he says.
At Danica Ejendomme, the real estate investment arm of the Danske Bank subsidiary Danica Pension, benchmarks are used to gauge property performance, and are based on many aspects.
They take account of factors such as the internal rate of return, current and forecast vacancies, the age of buildings, cost per square metre, the average remaining lease period, and the geographic location of buildings, says Christian Olsson, head of property at Danica Ejendomme.
“Performance measurement is very important, especially for the board,” Olsson says. “It gives a good benchmark on whether the company is doing well compared with the market.”
At the moment, Danica’s real estate portfolio is made up purely by direct holdings, and the aim is to achieve absolute rather than relative returns.
Olsson explains that the value of property investments is calculated according to the discounted cash-flow model, based on a 10-year horizon. In this model, future cash flows are estimated and discounted to give their present values.
But he is only partially satisfied with the property benchmarks that exist right now.
“The property management benchmarks could be better,” he says. “In Denmark, only some market participants actually deliver data to providers. The problem with creating benchmarks for buildings is that the individual properties have a lot of anomalies and problems which are specific to them.”
According to Danica Pension’s real estate investment strategy, the pension fund aims only to own property within Denmark, and all of this should be held directly, not via funds.
Current total property investment amounts to around DKK22bn. This includes 16 shopping centres – or around 41% of the total retail market in the country. On top of this, Danica Pension owns 137 office buildings and has a relatively small investment in the residential sector. It holds 500 apartments, but is building 1,000 units in Copenhagen.
However, the pensions subsidiary is planning to start investing in property abroad soon, and this time via funds.
APG, the €325bn Dutch pension manager, runs a portfolio of strategic long-term real estate investments in non-listed and listed real estate securities. “We have an integrated approach, whereas most investment managers have separate portfolios,” explains Patrick Kanters, global head of real estate at the organisation.
To measure performance, APG uses a composite of IPD indices, and the US NCREIF index. The resulting comparison benchmark is not perfect, but does reasonably reflect the investment universe APG’s property investment operation works in, Kanters says.
As ‘bricks and mortar’ benchmarks, these do not include fees and other asset management costs. Leverage is another factor is not accounted for in the benchmarks. “It’s not a true reflection of the underlying investment,” says Kanters.
The comparison with the benchmark becomes more meaningful when it is on a longer-term basis, particularly when measuring listed property investments, he says. “To get a better feel for the underlying investments and a better yardstick for performance, we also use INREV and ANREV benchmarks and listed regional benchmarks.”
APG is looking for absolute returns with long-term stability, including a relatively high level of income as well as capital appreciation. “In our strategy we have structured smart beta benchmarks,” Kanters says. “We don’t just copy the market-weight indices that exist, but we work with an optimised approach.” This involves using long-term targets for the Americas, Europe, Asia (ex. Japan and Australia), and deviating from the market-cap weight of standard indices.
Performance measurement gives a good insight into the ratio of risk to return as well as the level of volatility. It also shows whether the organisation’s investment strategy is working well, whether true operational value is added, as well as shedding light on how the manager is doing relative to others.
“Real estate benchmarks need to evolve further,” Kanters says. “Large parts of Asia are not covered by appropriate benchmarks. It is important that the geographical coverage will increase further and that the penetration into the region increases.”
Marieke van Kamp
ING Insurance Benelux
ING Insurance Benelux uses IPD indices for benchmarking its direct, non-leveraged property portfolio, and INREV fund indices for its leveraged funds.
“We use these indices – and others – to measure the risk-adjusted returns of real estate in the longer term,” says Marieke van Kamp, head of real estate investments.
The insurance company is aiming for relative, rather than absolute returns from the asset class. “The purpose of investing in real estate is to gain access to the specific risk-return characteristics of real estate,” Van Kamp says.
Investment managers are expected to give ING Insurance Benelux access to this real estate return. Van Kamp also says the institution expects its managers to be able to outperform their peers by its active management and stock picking.
“Benchmarking is key to measure both,” says Van Kamp. “We do try to achieve an absolute income return.”
The gradually changing perception of property within overall portfolios means benchmarking has become more important. Van Kamp says real estate is increasingly seen as one of the asset classes, rather than a separate part of the portfolio with its own rules and characteristics.
“Benchmarking of the total portfolio is very important – no changes there,” she remarks.
“And because real estate has to comply to the same requirements as the rest of the portfolio, benchmarking for real estate becomes more and more important.”
This benchmarking is less about current performance versus current markets, and more a question of the attribution of real estate over time. “The more accurate and longer the benchmarks or indices, the better the attribution of real estate within a wider portfolio can be shown,” Van Kamp says.
This, she says, is important for optimising the overall portfolio, but also in light of capital adequacy regimes such as solvency regulations.
The market for real estate indices has come a long way, according to Van Kamp, but she still finds the current range of benchmarks far from ideal. “We’re not there yet,” she says. “I think it is imperative to the property industry to further develop transactional and valuation indices to enable better risk-return modelling and measurement.”
Aviva Investors uses a variety of benchmarks, reflecting the client mandates in cases where those mandates are targeting relative returns.
These reflect the nature of the mandate and the geographic allocation, but generally relate purely to the unlisted real estate funds market in different jurisdictions, explains John Gellatly, head of European real estate multi-manager.
For example, the IPD/AREF Balanced/Pooled Indices are used in the UK, a variety of INREV benchmarks across Europe, IPD Mercer in Australia and NCRIEF ODCE in the US.
“Our client base, by number, is roughly evenly split between those with absolute or relative return performance objectives,” he says.
Clients either adopt absolute return strategies because they are looking for specific return profiles to meet their liability requirements, or because there are no suitable benchmarks available. This latter issue is particularly problematic for global mandates.
“Though IPD does produce a global index, it is still relatively immature, has considerable skews within it given the nature of the contributing entities – for example, US multi-family weighting on a global basis – and is only produced once a year, in mid-Spring,” Gellatly says.
As well as this, it is hard for Aviva Investors to construct robust benchmarks for mandates where it provides integrated real estate allocations solutions, he says, such as portfolios composed of a mixture of unlisted funds and public securities. In those cases, one of the defaults is to use absolute measures.
Gellatly sees a growing dissatisfaction with some of the indices, as people become aware of the distortions that happen in benchmarks, which can lead to distorted portfolio construction and investment strategies. “Hence the increasing application of so-called smart-beta strategies – or customised benchmarks,” he notes.
While the UK benchmarks for unlisted funds are generally quite good, Gellatly says, those across the rest of the world are relatively weak, although they are improving. “More disaggregation of the benchmarks would be beneficial so as to enable more attribution analysis across the real driver of portfolio returns.”
Relative return benchmarking using a peer group benchmark, such as IPD, is the most common approach for UK pension funds, observes Richard Cooper, senior consultant at Aon Hewitt.
Outperformance targets tend to be set at modest levels, typically 50-100bps above the benchmark return, which means managers are not under pressure to take significant positions away from the benchmark.
“The main advantage of relative return benchmarks is that measurement of the manager’s performance is unequivocal, even over relatively short periods, and performance attribution relative to the benchmark is simple to calculate,” he says.
But there are some downsides to this approach, which is why some pension fund clients have adopted absolute return targets, for example with a specified target above inflation. Given the cyclical nature of the property market, it can be virtually impossible to produce positive absolute returns over all time periods,” says Cooper. For this reason, absolute targets have to be long-term – measured over a property cycle.
“For these mandates, we use a range of metrics to assess and attribute short-term performance for clients,” Cooper says.
Performance measurement is an important part of an investor’s fiduciary responsibility and this has driven the improvements in property performance measurement over the years.
“Calculation methodology is a key issue for us, for instance where performance measurers such as custodians and IPD use different methodologies, which can result in inconsistencies of the returns reported,” he says.
The sharp swings in property prices over the past five years have exposed some shortcomings in valuation-based indices, Cooper notes, where market-clearing prices and secondary-market pricing for funds have been quite different from valuations. “Some clients have chosen to take specialist advice on fair value and adopt actual, realisable pricing,” he says.
But Cooper finds that while transaction-linked indices can help, commercial property trades are not similar enough or frequent enough for indices like this to be completely representative as benchmarks.