At last month’s IPE Real Estate European Investor Forum, IPE Real Estate editor Mark Cooper spoke to three leading real estate investors and a representative from benchmarking group Investment Property Databank about their views on benchmarking
The questions we asked
n Why benchmark in the first place? What are you comparing with, what are your goals and what is the exercise designed to achieve?
n What benchmarks do you use in your real estate portfolios?
n How effective and reliable are the various benchmarks tools available?
n Has improving benchmarking contributed to increased real estate allocations?
n Would you like indices for indirect real estate funds to show the performance of individual funds?
n How do you deal with the fact that most of the indices outside of the UK are only updated annually? How do you mark to market monthly or quarterly?
n How do account for leverage and differences in investment style when benchmarking unlisted real estate funds?
n How do differences in underlying valuation methods in different countries affect the reliability of benchmarks?
Does this undermine real estate indices and make it hard to compare between countries?
n Does more data lead to more volatility and a move away from the ‘smoothing’ that characterises property returns
based on valuation?
n We have indices for European and global listed real estate. How quickly are we moving to having benchmarks for
European direct real estate or global direct real estate
n Would you prefer a European index which has some questionable elements or wait until all the data for all the countries are equally reliable?
As part of the IPE Real Estate European Investor Forum, held in Barcelona in February, the investor forum section the magazine went live, before an audience of more than 100 pension fund managers, sponsoring fund managers and guests.
As another change to the normal course of events, a non-investor was invited on to the panel. Hans Op’t Veld of Investment Property Databank was added in order to give the panel and the audience a perspective on what developments were taking place in the world of property benchmarking and answer their questions about the IPD indices.
The questions above came from both the audience and the panel moderator, IPE Real Estate editor Mark Cooper
Dan van Aert
Stichting Pensioenfonds ABP
Dutch civil service pension fund ABP is a long standing investor in real estate, with investments in both listed securities and unlisted funds. Portfolio manager Dan van Aert says the most important thing an investor should consider in the context of benchmarking is to know what is required from the benchmarking. “You need to know what your goal is,” he says. “As a pension fund, we have our liabilities to pay - which is our ultimate goal. We have our allocation for real estate, so we benchmark the real estate returns against what we want to achieve for the pension fund.
“When we allocate we are looking at absolute return targets; looking forward about how much we allocate depends on our view on long term absolute returns.”
However, he points out that the day-to-day management of real estate investments is judged relatively. “We are in real estate and we want to be outperforming our competitors,” says van Aert.
As the ABP portfolio is divided into listed securities and our non-listed real estate private equity investments it uses different types of benchmarks. The listed section of the portfolio is benchmarked against the EPRA indices (now the FTSE EPRA indices) while the non-listed funds portfolio is benchmarked against the Dutch IPD benchmark.
As there are is a variety risk/return profiles on offer in unlisted funds, ABP takes this into account with its benchmarking. “For example, if you are investing in an opportunity fund you have to outperform the benchmark by a huge number of basis points and if you invest in a value added vehicle you have to outperform by a lesser degree but more so than for a core fund. If you have a lot of development and high leverage in a vehicle then you have to outperform core-style investments,” says van Aert.
The frequency of index reporting is an issue for ABP, in two different ways. “It’s hard to mark-to-market because lots of benchmarks only report annually,” he says. “Which is why we look at Dutch IPD because it is quarterly. All investments in our portfolio are valued quarterly.”
With unlisted real estate, van Aert argues that if the market develops more data, more transparency and more benchmarks then the result will be more volatility. “If you put that increased data into an asset allocation model then at the end there will be more spikes in the data - implying volatility and therefore higher risk. This could make you decrease the allocation to unlisted real estate as it alters its risk/return profile.”
At present, with unlisted vehicles, there is a smoothing effect from the lower frequency of data, which can be assessed as being lower-risk. Of course investors have the capacity to ‘unsmooth’ the data which will give them the best of both worlds. Van Aert says he agrees that monthly valuations in the UK IPD index will lead to less smoothing and more risk.
As one of the biggest institutional investors in real estate, ABP would be very keen to see a pan-European direct real estate index, says van Aert. “An index for all European direct property is something we would love to have, but the problem is that you don’t have the same level and quality of information in every country. However, we see that this situation is improving and in a rapid way.
“We also see organisations like INREV, which are making efforts to create more transparency. Also, getting better valuation methodology in some countries will improve the picture and make indices for those countries a lot more reliable. All of this will take time though.”
ATP – Danish labour market scheme
The Danish labour market pension fund ATP has been an investor in direct real estate in Denmark for 20 years but has only recently begun investing outside of its home market. It is in the process of establishing an indirect portfolio overseas and has so far invested e370m in 12 funds and expects to have invest e550m in 15-17 funds by the end of this year. It is now seeking a real estate investment advisor to work with the real estate team to work on a global real estate strategy.
Portfolio manager Mads Rude says: “Our Danish portfolio is well-established and uses the Danish IPD index as a benchmark. The international portfolio; that’s still in a build-up phase so we have more of an absolute return benchmark. This is something like 10-12%. Real estate in ATP is classified as an alternative asset, like private equity, so it is on the more exciting side of things where higher returns are expected. So with a 10-12% absolute return we are getting more into value-added territory.
“Our 10-12% absolute on the indirect comes from looking at what the risk free rate is, what inflation is, applying the effect of leverage and other things like that which lets us see where we can get to with the portfolio.”
Rude says ATP does not worry too much about the lack of a quarterly benchmark for its real estate portfolio. “With ATP being a long term investor, we don’t really do benchmarking on a quarterly basis; the mark-to-market on a quarterly basis is not that important to us. There is a problem that we don’t have a benchmark which accounts for differences in fund style and gearing. That’s partly the reason why, in this build-up phase we have used an absolute return benchmark,” he says.
“Hopefully when INREV introduces a benchmark over the next few years we will get something for those types of returns, so we at least have some kind of proxy to see whether we do well or badly.”
ATP is not “especially keen” to see returns for individual vehicles within INREV indices. Rude says: “We know the funds we are investing in and how they perform; we have a portfolio view and a benchmark for that but I think we are in too early a phase, as a market, for that kind of disclosure.
“The UK and Irish markets are used to long series of data, while on the continent we have not had that. Now more benchmarking is available on the continent we are seeing more UK investors move over here.”
Rude believes that short term performance of unlisted funds relative to the benchmark is unlikely to result in more liquidity, especially as the market is at an early stage in Europe. “In the direct market if you see properties underperform you can sell them or do something different with them in the short term,” he says. “It’s different with a fund because you have ceded control to the manager and for well-researched reasons. If you have been in a fund for only one or two years it’s too early to say, especially if the fund is a young one. In the longer term we will probably trade holdings in unlisted funds though.”
Rude is keen for organisations such as INREV and IPD to get on with developing pan-European indices, despite there being questions over the amount and reliability of the data from some areas. “The decision by INREV to create something for Europe came out of investor demand. In trying to get something going there’s a lot of issues that come up, but we take the view that you need to get some data out. Then you can see what is available and encourage people to move in the direction of providing more and better information. In time this should converge into a dataset that can provide a nice like-for-like comparison There’s also nothing really out there, which also made a difference!”
Irish Life investment manager
Irish Life has been a long term investor in direct real estate in its home market on behalf of its own pension fund and retail and institutional clients, and for a long while in the UK as well. Chief investment officer Hooman Kaveh is currently looking at expanding the focus to continental Europe.
He says: “I look at real estate from an asset allocation perspective, so it is one asset class we can consider in our investment portfolio. So as with equities and bonds and more recently with hedge funds we are looking for the characteristics of the asset class. That’s typically the main reason we use benchmarks. We want to get as long a return stream as is available in order to measure the asset class’s volatility and most importantly the correlation with other asset classes.”
He cites some points which make real estate more complex: firstly that real estate funds tend to have some degree of leverage and it produce an average of property fund performance, rather the performance of direct properties.
“We’ve been investing in our home property market, Ireland, for many years - 40 years in fact,” he says. “That predates the creation of the IPD benchmark by quite a long way. Our unitised property fund was set up in 1973 and people used to ask us for monthly data to help them benchmark themselves!”
Kaveh says the group and pension fund are comfortable with investing in the UK and Ireland. “We are only just beginning to look at continental Europe and one of the reasons we’ve been less comfortable in the past is the lack of benchmarks. Also, we decided in 1990 to invest in continental European real estate - in Benelux and Germany. Now that was one of the worst times to do that and we went home in 1994 with a big loss!
“The key thing is that there doesn’t seem to be a lot of transparency out there and we have been looking at the IPD indices to try to get a better idea of what the picture is. We get monthly data in Ireland and that’s one of the reasons we have largely stayed in our home market. On the continent it would help allocations if there were more benchmarks available.”
Kaveh does not believe that more frequent benchmarking for real estate and thus increased statistical volatility will lead to allocations being lowered. He says: “I agree that you get more risk with more data. People do say property volatility is greater than is actually visible and I think that’d right. Nonetheless, I’m not sure that would lead to lower allocations. More data tends to mean more confidence, which means a greater allocation. A modelling situation might imply a lower allocation but in the minds of people it would probably lead to more enthusiasm for the asset class.”
He argues that an investor has to look at both absolute and relative benchmarks. “It does make sense to have an absolute structure and you should have this absolute target in mind. However, at a lower level a single asset or manager might rely on sticking to the abosolute benchmark in the medium term and not doing as well as they might.
“For example in the UK and Ireland for the past few years real estate has been doing very well with double digit returns. So no investor would be pleased to have an asset or a manager producing 8-9% and then being told: ‘well that that hits your absolute target, it’s OK isn’t it?’, when IPD is showing the market getting 18-19%. So it’s OK to think in absolute term at the top level but not all the way down.”
Kaveh is aware of the limitations in producing a pan-European direct real estate index but says: “I know that most indices are built from the bottom up but that could be very time consuming in the context of European real estate. You could take samples and construct from that instead. I know I’d prefer one sooner than later.”
Hans Op’t Veld
Investment Property Databank
IPD is the world’s leading benchmarking service provider for direct real estate. It began with a UK index in 1980 and has since developed indices for 16 countries, mainly in Europe. It is also working towards a pan-European index.
Op’t Veld recongnises that the limitations of valuation practice in the underlying markets will always affect the perception of an index. “There is one obvious example in Europe: Germany, where valuation standards are dictated differently from the way cross-border investors would prefer and I’m sure the German market believes that’s the right way for them to value. However, it makes it hard for them to match up with the International Valuation Standards Committee’s normal valuation standards in other countries.”
He says there are also some European countries, especially southern European countries, the valuation code is not as developed as in countries such as the UK, which have a more mature real estate market. “What some people don’t realise is that there are differences in the ways valuations are carried out from country to country. You can’t just take a benchmark at face value. You have to think about what’s in there and what makes it tick,” he says.
IPD is increasingly finding that investors are no longer just focusing on the direct assets but on vehicle level returns, so what IPD is now spending a lot of time on reconciling asset level returns with vehicle level returns; bringing gearing and other issues into the equation. “This will give investors more of an idea what happens in between asset-level and vehicle-level returns,” says Op’t Veld. “Investors are demanding that, as they blend different elements into their portfolio, that we take account of that.”
Op’t Veld knows that many investors would like to see individual fund returns – “to name and shame!” he says. “Transparency has to be the ultimate goal, but to report the performance of individual funds at the moment would require taking account of the fact that funds are at different stages in their life. So that means that you would have a ranking of apples and pears and you must try to avoid that.”
He says that publishing individual fund performance only works in an environment of comparable funds, where there are a significant number of similar mature funds to compare. The UK’s Association of Property Unit Trusts indices are able to do this. “The rest of Europe is moving in that direction but it will take time.”
A pan-European index for direct real estate is a big challenge for IPD, he says. “IPD is now looking at most European markets, although we have a few blind spots. So we can get a view of the maturity of the European markets. Really, where investors have gone, we follow.
“For example, many investors have been moving in to southern Europe in recent years and so we are getting the data back from them which helps us construct the benchmarks. We will have a fully-fledged European index by the end of this year, which we will add countries to.
“At the moment we have rather short data series for some countries but we are building on that and this will obviously take time. But we have to start somewhere and it will evolve. If everyone says ‘let’s wait’ then an index might never happen.”