Japan is witnessing some important developments in the area of real estate indices. Jeff Wynkoop examines what this means for investors

Real estate indices serve an important function for real estate investors because
they provide an objective standard for assessing the performance of a specific asset and its asset manager.

Markets necessarily fluctuate over time, making it hard for investors to know to what extent the performance of an asset should be attributed to its manager, and to what extent performance is due to the market itself. Thus, having a broad-based real estate index or set of indices is vital to the creation of an efficient and transparent market for professional investors.

In addition, real estate indices can be used for creating real estate derivative securities, which can also be useful for investors seeking direct real estate exposure without actually owning the hard assets.

There have been several recent interesting developments in Japan in the area of real estate indices. In April 2011, the Tokyo Stock Exchange launched an index based on changes in residential asset prices, the TSE Residential Price index, modelled on the US S&P/Case-Shiller Home Price index. The index uses actual transaction prices for 470,000 deals involving existing residences in the greater Tokyo Metropolitan area, beginning from June 1993 and compiled monthly. In addition to benefiting consumers, this residential price index should provide a good yardstick for macroeconomic comparisons with other countries with similar indices, such as the US.

Another development is the ARES J-REIT Property index, which first came on line in April 2006. This index of Japanese real estate investment trust (J-REIT) performance tracks movements of prices based on ¥1.8trn of securitised Japanese real estate assets. However, the Association of Real Estate Securitisation (ARES) is currently in discussions with various real estate players with the intention of increasing the size of the assets subject to the index over three times by mid-2012. If successful, this will mean the index will encompass approximately 40% of all securitised real estate assets in Japan, including, for the first time, a number of privately securitised assets not traded on a public exchange.

The ARES J-REIT Property index is an ‘average price’ index, whereas the above-mentioned TSE Residential Price index is a ‘repeat sales’ index. The latter is based on an historic sample of prices (at least two actual sale prices) for the subject properties. The other major type of real estate index is a ‘hedonic regression’ index, such as the IPD/Recruit Residential Price index. An average price index is based on an aggregated average price of a certain measure (for example, price per square foot). The problem with this type of index is it is not always the best measure for investors because it does not take into consideration how property prices have changed between two periods of time.

Since properties sold in one period never coincide exactly with properties sold in another, changes in price may be due to specific characteristics of the properties changing hands (for example, different average age of the buildings transacted), rather than any aggregate change in the market per se. That is, comparing different properties of varying locations with different construction dates and other building characteristics may not accurately reflect actual market conditions (if prices go up from one period to the next, it may be because the average quality of the properties changing hands increased or the properties were in better locations, rather than any change in the market fundamentals themselves).

The repeat-sales and hedonic-regression indices were both developed to try to overcome this basic problem of not compar ing like-for-like. The repeat-sales index is based on groupings of individual properties that have been transacted a multiple number of times. This aggregation is used to control differences in price due to differing locations, attributes, and so forth, of properties. Since this type of index varies due to actual changes in prices paid by buyers for the same properties, performance is comparable to the performance of stock indices (which change in price over time only due to changes in what buyers are willing to pay for the same securities).

The hedonic-regression index takes another approach to solving this comparing apples with oranges problem. Rather than tracking changes in price in the same properties (which may not occur for large intervals of time), a hedonic-regression index is calculated by using a model of individual property and transaction characteristics (for instance, location, age, size, building quality) and tracking the changes in these characteristics over time.

In this way, a hedonic-regression index is able to isolate outside factors that may
be affecting price and distil any variance as due only to actual market changes, provided the characteristics used as constants in the model remain constant, and are complete enough to provide a good assessment of price changes for each property.

Jeff Wynkoop is a principal at Ichigo Real Estate Investment Advisors