معافیت های ساختاری و تنوع بخشی: تاثیر بحران مالی آسیا در سال 1997 بر ادغام آسیا و اقیانوس آرام در بازار دارایی واقعی

Currently, there exists relatively little research on the influence that the 1997 Asian financial crisis has had upon capital flows within the property market and the associated long-run implications of it. This paper examines the impact that the crisis has had upon the integration and dynamic links between a number of Asia-Pacific real estate markets. The results show that Asia-Pacific property markets are integrated, despite a structural shift occurring at the time of the crisis. These results are a particularly important finding for fund managers concerned with the impact of globalization on the performance of their real estate portfolios, showing that in the Asia-Pacific region diversification benefits are actually less than that suggested by an analysis incorrectly ignoring the crisis.

International diversification in real estate has recently become a more important issue among academics because the evidence is not entirely clear on the benefits from diversification across property markets. Portfolio managers are faced with the persistent problem of maintaining investment returns while simultaneously reducing risk. This outcome is achieved through portfolio diversification – allocating resources across a number of asset classes and sub-classes as well as across countries. Ideally such managers seek investments in markets that are insulated from each other so that, in particular, the effects of a collapse in one market, or one segment of the market, are not transmitted to investment holdings in other areas. Hence the notion of market segmentation is of prime importance in property portfolio management. The outcomes from academic research on property markets over the last few years, however, have been unable to reach a firm conclusion on whether international diversification in real estate is beneficial. Much research has centered on the issue of market segmentation/integration. For instance, if international property markets are well integrated then little gain in risk reduction may be achieved through holding internationally diversified investments. Well-integrated property markets can also imply that such markets may respond to the same economic stimuli, thereby providing little gain in diversifying across these markets. On the other hand, if markets are clearly segmented and respond to different economic stimuli, then it is important for portfolio managers not only to diversify, but to be able to allocate resources in a dynamic fashion so as to take advantage of changing conditions in each market. Moreover, even if markets are integrated, the impact that a shock transmission may have upon these markets may differ. This can be an important issue if there are structural shifts in one or more of the asset series making up the portfolio, since the existence of a structural break may disguise the true nature of any potential relationships between assets within the portfolio. This may be particularly crucial to those portfolio managers concerned with strategic asset allocation, i.e., the diversification strategies to be pursued over the long term. This study aims to analyze the effect that the 1997 Asian financial crisis had on the interdependence among several Asia-Pacific real estate markets. In particular, the paper is concerned with whether the benefits of diversifying across such real estate markets may have altered because of the crisis. Not only will this contribute to the research on whether property markets are integrated, but also provide a very useful look at how these markets reacted to the Asian crisis.1 The finding is that failure to take into account the events of 1997 disguises the true nature of the long-run inter-linkages between these property markets. Specifically, if no consideration is given for the 1997 crisis, the real estate markets in the study group show no signs of integration. However, they are found to be significantly cointegrated when allowance is made for the crisis. This finding of cointegration has important implications for property portfolio managers dealing with these markets, as not only do they have to be aware of influencing events within one specific market, but also events in the other cointegrated markets. The structure of the rest of the paper is as follows: Section 2 briefly reviews some recent literature on diversification in real estate markets; Section 3 considers how the methodology of cointegration can be used to indicate the likely diversification benefits that exist between markets; Section 4 describes the data and results; while Section 5 offers some conclusions.

The literature to date examining the benefits to diversifying real estate holdings internationally is mixed. To contribute to the debate, this paper set out to examine the question of whether a select group of real estate markets in the Asia-Pacific region were inter-linked, whether such inter-linkages might be influenced by the Asian crisis, and what the implications of the finding would mean for diversification into Pacific-Rim real estate markets. The tool chosen to examine this was cointegration analysis, which allowed us to query whether Asian real estate markets were integrated over the long term. When conventional Johansen procedures were applied to the full data set, the results suggested that the property markets were not cointegrated, implying benefits to diversifying in the region. This finding was supportive of some previous studies indicating property markets are fairly independent of one another. Nevertheless, the outcome was rather surprising given current trends in globalization and the extent to which most financial markets are now becoming inter-related. Therefore, suspecting that the 1997 Asian crisis may have had an impact on these results, Zivot and Andrews unit root tests were conducted on each series and unit roots identified, with a possible break at about the time of the crisis. To account for the impact of the potential break, two-period Johansen tests, along with Inoue (1999) and Johansen et al. (2000) cointegration tests were then applied. These tests confirmed that, contrary to the initial analysis, this group of property markets in the East Asia Region was, indeed, cointegrated. It seemed that the Asian crisis generated a structural break that caused a shift in the model parameter values around mid 1997, and this notion was supported by a Johansen parameter constancy test. In addition, we used exogeneity and exclusion tests to show that Japan, at the 1% level, and Singapore, at the 5% level, could be considered important in influencing the long-run equilibrium within these markets. From a portfolio manager's perspective the influence from these markets, particularly Japan, cannot be ignored when examining local property market performance. Moreover, the above results not only show that there are both long and short-run linkages between the property markets within the Asia-Pacific Region, but that the impact of the crisis has had little impact upon the long-run degree of integration within these markets. Finally, although this study clearly shows the need to be cautious when drawing conclusions about the independence of real estate markets, this paper did not utilize direct property data. A topic for future research would therefore be to determine whether the same results apply if direct property series were examined. Recent research by Kallberg et al. (2002) has shown that the correlations between direct and indirect property returns for Asian real estate markets are relatively strong, although (consistent with other research) the standard deviations are generally higher for real estate securities. Unfortunately, the direct property data examined by Kallberg et al. was only available at quarterly and semi-annual intervals over relatively short time periods, making any analysis somewhat cursory. In the future, when suitable data become available, a similar analysis to this study will be possible and desirable. It is anticipated that the outcome of such research would be supportive of earlier work by Myer et al. (1997) and Case et al. (2000) in demonstrating not only how closely related the underlying national property markets are, but will also illustrate the long-run similarities and differences that may exist between the direct and equity-listed property markets.