What do property funds actually achieve, and how should we measure that achievement? Ian Cullen reports

The manner in which property returns are delivered to end investors has changed dramatically over the past 10 years. In the mid 1990s, segregated life and pension funds, which dominated the institutional property map at that time, held less than 1% of their portfolios indirectly. Moreover, those unlisted pooled funds, which offered real estate exposure to segregated accounts, employed no gearing at that time. So returns on net asset values delivered to end investors, while diluted by cash and skimmed of fees, were essentially returns on the underlying investments.
The manner in which property returns are delivered to end investors has changed dramatically over the past 10 years. In the mid 1990s, segregated life and pension funds, which dominated the institutional property map at that time, held less than 1% of their portfolios indirectly. Moreover, those unlisted pooled funds, which offered real estate exposure to segregated accounts, employed no gearing at that time. So returns on net asset values delivered to end investors, while diluted by cash and skimmed of fees, were essentially returns on the underlying investments.
Today, life funds hold over 22% indirectly and pension funds over 19%. What is more, the IPD UK Pooled Property Fund Indices demonstrate that most indirect offerings of property returns are now geared, some at well over the 50% level, which means that debt is now spread to greater or lesser extents throughout the institutional property market.
From an IPD perspective, this raises many challenging measurement questions, which render problematic our bottom-up methodology - the cornerstone of the IPD approach to offering even playing field transparency and comparative analysis to real estate investment markets worldwide.
This paper re-addresses some of these basic measurement and analysis questions, using the simple and ever more widely deployed financial technology of gearing as an example of the problems we face. Our work is far from either complete or comprehensive in its coverage of the rapidly expanding range of financial technologies available to fund managers.
Current excitement at the opportunities offered by derivatives would have been another obvious one to choose as it raises a raft of measurement questions. Gearing has, however, a relatively long track record, is essentially simple at least in concept, and is more than sufficient to challenge the conventional granular IPD approach to compiling and measuring gross returns.

The decomposition of fund performance

Conventionally, IPD undertakes the unpacking and explanation of a fund's relative performance in two separate ways. First, the gross asset level return margin is split to identify and quantify the weighted contributions of different types of asset management activity - buying, selling, development/refurbishment and the more passive forms of management involved in retaining what are termed ‘standing' or ‘stabilised' investments.
Second, a parallel analytical approach takes that same gross return margin and apportions it to a mix of strategic and stock specific contributions - the classic attribution method.
The difficulties with this approach are first, that it introduces a somewhat arbitrary separation of two closely linked levels of portfolio management - strategy formation and investment activity - and second, that it leaves open the question of how to treat financial management; as a part of strategy formation or as yet another and disconnected type of investment activity.
Figure 1 illustrates the standard separation of the two sorts of analysis. Gross asset returns reconcile at all portfolio levels, but not at any finer disaggregations.
Financial overlays, such as gearing, lend themselves to the weighted contribution approach, since it permits their simple aggregation with other financial layers on top of the direct return components, so taking a gross relative return to a triple net bottom line measure. Moreover, such overlays can often be thought of as extensions of the base range of investment activities - cash retention sits comfortably beside asset retention just as indirect investment parallels direct investment.
This is a perfectly legitimate way of decomposing and reporting real estate gross asset through net fund level return differences, and is now included in quarterly IPD Benchmark reports in the UK. The weighted contributions of indirect interests, cash, gearing and fund level costs are overlaid upon the basic gross asset returns to approximate the return on net asset value (NAV), which is actually received by end investors.
Precise reconciliation will normally be impossible - even if all fund level activity is fully and accurately reported - because periodic (monthly) performance analysis will always at best approximate the continuous cash-flows that determine the bottom line NAV. In practice, we are still some way off the comprehensive fund level data capture target set more than two years ago.
When we do reach that target, however, the power of this analysis to reveal the links between asset-driven returns and the distortions of financial activity will significantly extend the utility of IPD's comparative reporting.
This framework can be extended to encompass the impacts of total return swaps, and other synthetic interests, which may be exploited within real estate funds.

The reintegration of financial and asset analyses

Is this as far as we can push fund level performance analysis? To address this question, we are currently revisiting the basic attribution framework to see if that model is capable of extension in some way to encompass debt analysis. Pure asset level attribution clearly misses out on much of the excitement that drives bottom line returns. Applied as it normally is to composite segment total returns, it also misses some of the asset level subtleties of strategic and stock selection impacts upon performance.
The most obvious limitation is the absence of any income/capital decomposition of the returns, which are attributed to property or strategy factors. When markets are roaring forward and are being driven by high rates of capital growth, this limitation may not be noticed. As markets decline, the income component of total return becomes of increasing importance, and in principle there is no reason why attribution methods should not be applied separately to both the capital and income drivers of returns.
We are currently pursuing exploratory work with a view to decomposing asset level attribution analysis so that it can separately identify the stock and strategic elements of both income and capital profiles of the assets.
The key relevance of this refinement lies also in that it offers a much more powerful platform for the addition of some of the crucial financial management strategies to the core analysis. Cash and debt management stand out as being of critical importance at the financial level. Both can be analysed together, since cash balances dilute leverage and focus attention on the crucial net debt ratios and flows.
One of the most important features of debt analysis is that it has both capital and income components, which may be sensibly embedded into fund level attribution, alongside the flows from and value movements of the ungeared assets.
The analytical possibilities of extending the model are indicated in figures 3 and 4. The first crudely summarises the outputs and functions of conventional attribution analysis applied to commercial property investments. Though crude, the summary does accurately cover the focus on composite total returns, and identifies the fundamental analysis distinction - between the impacts of strategic balance within the portfolio and those associated with stock specific factors.
Figure 4 shows one way of extending the base attribution framework to capture both the allocation and selection contributions to a fund's gross performance as they flow separately from capital and income return components. The method then decomposes the gearing impacts upon bottom line relative performance into three parts:

The relative costs of servicing the debt; The contribution from the benchmark relative gearing ratio;  The contribution from the gearing of fund performance differentials directly.

The impacts of debt in a rising market

The question of the relevance and utility of this extension of the traditional approach is essentially an empirical one. On the face of it, the need seems minimal. A comparison of the direct property returns to unlisted pooled funds, with their full net bottom line results over the past few years, suggests a very simple interpretation - the holdings of indirect interests made very little difference to asset returns, the inclusion of fund level fees had the obvious diluting impact, and very little else is noteworthy.
A closer look at these series, however, reveals a subtly different story. The first order differences between these Indices show an initially more or less undiluted negative impact from some mix of cash and fees upon gross asset returns. These penalties are then at least partially offset as gearing cuts into high teens returns over the last three or four years. Figure 6 reports these marginal effects.
The key remaining question, therefore, is whether we can explore and explain these broad relationships at fund level, using the extended attribution framework outlined above.

A worked example

A reasonable sample of the unlisted pooled funds in the UK are now reporting sufficient details of their financial activity - as well as their gross asset flows and values - to warrant at least a demonstration analysis. What we have done to date is to create a fictitious fund by aggregating the actual data from three separate funds (all relatively highly geared, averaging around 80%), and then construct a benchmark from a mix of unlisted funds whose typical gearing ratio is much lower (averaging around 10%).
We can best demonstrate the approach using two very different sets of quarterly results - one at the height of the market boom (Q2 2006) and the next at the (current) low point (Q4 2007).
The bulk of the research analysis pre-dated our full capture of the latest 2007 Q4 results, and so is based on October/November estimates. In the event, the outturn proved more painful than our fairly bearish early estimates. If the analysis were rerun today on the basis of actual results, the contrast with 2006 would be even more stark
The two figures paint a fascinating and painfully plausible account of the fate of a highly geared fund as the market goes from high to low. In Q2 2006 the Benchmark gross return was 6.2%. The fund actually underperformed this result, despite adopting and applying a beneficial strategy - with an overweight position in provincial offices - one of the key high performing segments of the market.
The problems were stock specific, again in the all important provincial office segment, and were felt exclusively on the capital side of the return equation, with a strong income return partly offsetting the pain.
What this meant for the fund's bottom line was equally fascinating. That part of the gearing impact that was related to the fund's relative return was also negative. However, given a solidly positive fund gross return and a debt ratio roughly eight times the benchmark average, the net relative position flipped from an asset underperformance of 1.3% to a NAV outperformance of over 1%.
The conversion of asset underperformance to fund outperformance through gearing depends, of course, on a healthily growing market. Figure 8 offers a vivid demonstration of what happens when that prop is removed. The asset level fundamentals have not changed and so offices (and now industrials) are still performing less well than their peers. But incomes are relatively strong and the fund is still well placed strategically. So the asset relative return represents less than a 1% deficit.
This margin is, however, relative to a market return of -4.6%. The figure shows how this escalates the NAV deficit to a full 5.5%. The gearing impact associated with the fund's relative return was once again small (at around 0.5%), but the 80% gearing of the absolute return wiped a further 400 basis points off the bottom line, damaging the relative margin by a similar amount.
A great deal more needs to be done before we can routinely analyse the performance of property funds, from top to bottom line, in a seamless manner.
 This will require the adoption of a single standard for the recording and measurement of all financial levels of fund performance, along the lines currently in place for the IPD measurement of returns to directly held real estate.

Ian Cullen is co-founding director of IPD