Evaluating the valuers
Valuations are used as an alternative to actual prices as, even in highly liquid markets such as the UK, the quantity of transactions averages only 14-15% of the portfolio each year. The basis of valuation is market value and this represents the exchange price that would have occurred had the property actually transacted. It is not the role of the valuer in these circumstances to suggest that prices are either too high or too low and adjust the valuation accordingly, the role is simply to, as accurately as possible, record the estimate of what price the property would exchange for at the given date. Therefore, we can only trust valuations if they are attempting to record that price and do it well; the test for this is how close they actually come to prices.
The most transparent market for this test is the UK as it has the longest run of data on prices and prior valuations and the Royal Institution of Chartered Surveyors and Investment Property Databank do an annual analysis (RICS/IPD, 2005). Major analyses have also been undertaken on US and Australian data and they are also beginning to emerge in mainland Europe through the work of IPD and its partners. However, due to constraints on space I'll focus only on the UK data.
Some major caveats must be made. A straight comparison between prices and valuations can only be interpreted if the market process surrounding those sales is also understood. If the valuations have an impact on that process then they are not independent of each other and so the level of a prior valuation might influence which properties are sold and also the price. In the UK, Baum, et al (2000) have looked at the influence of valuations and found that some pension fund trustees will not allow or are less than happy about sales at less than prior valuation or book cost and that the level of valuation is considered by some fund managers when deciding which stock to sell.
This also raises another issue which is asset manager influence over the valuations. Baum et al (2000) also reported cases of client influence such as:
n Where pension fund trustees did not allow sale at less than valuation, valuers were pressured to down-value at year-end valuation to make the subsequent sale easier;
n Valuers were pressured to support a purchase price at the next year-end valuation so no performance measurement downgrade in the first year;
n Valuers were pressured to move end-year valuations where the asset manager bonus payment was based on tracking total return benchmarks. This issue can occur at each year-end with results in valuations being ramped higher and higher through time;
n Where a new asset manager was appointed, he pressured valuers to reduce valuations in the first year so his performance started off a lower base and he was more likely to hit subsequent targets coming off this lower base.
Not all of these have a direct impact on the valuation/price debate but it means that the expectation for any price/valuation comparison is not that price and valuation will be equal, it is more likely that prices will exceed valuations on average as a number of properties which would come from the lower end of the price distribution and higher end of the valuation distribution will not find their way to the market or will be taken off the market as the "reserve" price will not be met. These issues are important for the German open-ended fund debate where they operate under rules not to sell at less than valuation (within 3%).
The actual figures for UK valuation accuracy are illustrated in Figure 1 and show that the level of accuracy was not good in the 1980s but things have improved significantly since then. Absolute accuracy has been under 10% for most of the last 15 years (ie, average difference between valuations and prices ignoring whether it is an under- or over-valuation). In rising markets such as the late 1980s and in the period since 1995, valuations tend to lag prices, in the most recent past, bias has been increasing as markets have been rising (bias is the valuation/price comparison taking into account whether they are over or under valuations). The most interesting observation is that in poor market states, as occurred in the period 1990 to 1995 in the UK, valuations tend to catch up prices and, if the poor market is long enough, exceed them. This also has implications for the German situation.
Overall, the accuracy figures, subject to the caveats, tend to suggest that valuers do not respond to market change quickly enough and tend to reduce the gradient of that change, knocking off the peaks and troughs of the market cycle.
Hence the view that market indices based on valuations tend to lag real price markets such as equities and bonds and have less volatility. Comparisons of the correlations between equities and bonds and property in order to make the case for property as a diversifier are therefore naïve unless this process of smoothing is understood. This smoothing occurs in a market where no attempt is being made by the valuers to actively smooth the troughs and peaks of the market; these effects would be magnified if the system included this tendency from fund managers and valuers.
To summarise, valuations done with the objective attempt to mimic the market price regardless of what the valuer thinks about the level of prices, and where the fund managers are also only interested in true performance rather than manipulating it for their own purposes, can be trusted as long as the limitations of valuations are understood. Valuation is not a precise science it is therefore difficult for the valuer to resist persistent attempts to move the valuations by small amounts so that fund policy can be implemented. Is there a valuation issue with the German open-ended funds? Not according to the German open-ended fund managers association (BVI). In a press release of January 24 2006 it stated:
"The BVI is confident that the current use of a committee of experts results in realistic valuations. Property sales completed in recent months confirm this
view. The association therefore regards public criticism around valuation as unjustified."
Unfortunately the BVI has one small problem. No one appears to believe it and perception is important in all capital markets, especially those with illiquid assets and an instant call on their funds. Not that the valuation issue was the main concern that led to the mass exodus of investors but it is certainly part of the poor perception. Any hint of mis-valuation to hide the true state of the market value of the investments in the fund would help persuade investors to off-load units, especially if that mis-valuation helped keep the prices high.
Protestations that individual sales out of the fund are close to valuation are not evidence that the fund as a whole is not over valued. Fund managers will pick those properties that have valuations more closely aligned to true market value and the previous discussion suggests that price/value comparisons need to be treated with caution. So this assertion adds more fuel to the fire, are the assets left in the fund the most over-valued properties? Added to this is the suspicion that German open-ended fund managers have more motivation than most to manipulate valuations to match their strategy.
Their strategy is based on a clear commitment to offer a product that has very low risk and volatility. The BVI ‘sales' document sub-titled, ‘An investment in solid value' is packed full of statements like "a low volatility alternative", "steady growth in value at a low risk" and "stable profitability and the absence of wrenching moves in market price". How is that policy delivered within a property fund?
There are diversification and stock selection strategies for investing in low-risk cash flows and individual funds can have lower volatilities than others, but to create a whole market that appears to ignore the volatility which appears to exist in every other part of the developed world appears either very clever or very suspicious. A comparison of the volatility of the IPD indices around other parts of Europe and the world with Germany is stark (figure 1). With the possible exception of the Netherlands (which also may have a case to answer), all the other markets are much more volatile. As the vast majority of volatility within total returns comes from capital growth volatility, and capital growth is assessed through valuations within performance measures, there is one other way to deliver a low volatility measure. Are the valuations objective accurate reflections of the market or are they something different?
There is not the space in this article to do a comparison of the valuation systems in different countries around the world. However, Germany does differ from some of the other mature commercial property markets such as the UK, US, Australia and New Zealand in many respects in that there is significant state regulation rather than self-regulation, the education of valuers has been in the past based on the construction side of the business and only recently has it been taken on by the business schools and the number of valuers accepted by the funds appears more restricted than for investment funds in the other mature economies.
Market data is less freely available than in some of these other countries and it appears that some fund valuers do not work in the market place. However, there is more than one valuer for each property but each valuer is allowed to earn a substantial element of their total income from one fund (which could mean even more from a parent company which runs more than one fund) Germany is also the home of "sustainable value" concepts, mainly applied in the lending valuation arena, a concept that has been strongly attacked outside of Germany for its lack of economic basis or rational application.
From these differences a number of questions arise but they can be rolled into one overall result. Given the strategy of the funds and the education and background of the valuers, were the valuations heavily smoothed by a combination of both valuers and fund managers comfortable with the "need" to keep the volatility of the fund as low as possible? In the long term, the overall returns to these funds would be based on income and would no doubt match other property funds but shorter-term fluctuations were smoothed out giving the illusion of low volatility. This only became a problem to long-term investors when the bear market lasted for much of the last 10 years in Germany, as the real values didn't come back to the valuations; as they had in the past. What should be done?
There is no monopoly on the truth as far as systems are concerned and each jurisdiction has to learn from both the good practice and mistakes of others. In the UK, for example, the bank lending valuation process has gone through significant change since the UK property crash of 1990 and has had major discussions between client and valuer representative groups, two major reports on the valuation process (Mallinson, 1994; Carsberg, 2002) and a number of major changes to valuation standards.
Despite some major changes for the good, a number of processes were introduced which proved to be stunningly stupid (a new basis of valuation called Estimated Realisation Price is probably the best example) and not all have been totally discarded. The problem of client influence on valuations has been a major issue leading to some monitoring procedures from RICS which may or may not help the valuer resist the more inappropriate client contact. These initiatives need monitoring to see if they do lead to more transparency in the process, which is the key to greater trust in the post-Enron environment.
As the UK has gone through this process, it may prove a model of both what works and what doesn't. In Germany, some of the initiatives already taken by the BVI should improve the situation and the globalisation of property education and training suggest that in the longer term Germany will not be perceived in the same way. However, in the short term, the publishing of individual valuations might highlight the valuation issue even more.
The worldwide improvement in commercial property values will also help any valuation/value mismatch, but this may be temporary, and the core requirement is for all valuations to be undertaken to objective estimates of the asset exchange price.
Regardless of the actual truth, German valuers need to get rid of any perception that they are adopting average market values, sustainable concepts, understating potential void rates or void periods, not using capitalisation rates based on market transactions, relying on cost-based appraisals or keeping a cartel of valuers allowed onto expert committees earning significant percentages of their total fee income from a single fund of funds from the same stable. In my opinion that is the agenda for them to address and there are signs that some movement is taking place. But they may need to adjust quickly. Some commentators feel that the current level of capitalisation rates are not sustainable and a correction is coming; let's hope all fund managers, including German open-ended funds and their valuers, are ready to mark it accurately if it occurs. If not, the markets might make them regret it.