Kiat Ying Sky Seah

Estimating and decomposing changes in the White–Black homeownership gap from 2005 to 2011

This study evaluates the effects of the recent US housing bust on the White–Black homeownership gap by estimating and decomposing the changes in the distribution of the gap between 2005 and 2011. Our analysis shows that the housing bust did not affect the homeownership gap uniformly. In fact, we find that the gap decreased for households that were the least likely to own and remained unchanged for households that were the most likely to own, and that Black households with around a 50% probability of homeownership were especially vulnerable to the crisis. We also find that the contribution of the residual gap was modest. Changes in the White–Black homeownership gap over the sample period are mainly attributed to changes in household income, whether the household earned dividend, interest or rental income, and marital status, with the extent of their respective influences varying over the homeownership distribution. Our empirical approach reveals distributional information on the determinants of the changes in the homeownership gap at the household level. Such insights have valuable policy implications that would otherwise be concealed in analyses that look only at the conditional mean.

'REITarization' Effects on Global Real Estate Investments

The creation of real estate investment trusts (REITs) facilitates the conversion of illiquid commercial real estate assets into liquid tradable securities in the public market space. The process increases liquidity and generates significant growth in cross-border real estate transactions. This study uses a large sample of proprietary commercial real estate transactions data from more than 73 countries within the sample period of 2001 through 2015. We investigate the impact of REIT creation on international real estate capital flows. Using non-REIT countries as the control group, we show that REITs open a new channel to attract institutional investments in real estate. New investments in real estate by both local institutional investors and foreign investors increase in REIT countries relative to non-REIT countries. The results imply that there is no substitution effects of REITs on private real estate market activities. We also find no evidence of “crowding out” effects on local investors, as we do not observe significant changes in the outflows of real estate capital. The relatively large increases in domestic real estate investments have caused declines in the shares of cross-border real estate capital flows in the post-REIT regime.The creation of real estate investment trusts (REITs) facilitates the conversion of illiquid commercial real estate assets into liquid tradable securities in the public market space. The process increases liquidity and generates significant growth in cross-border real estate transactions. This study uses a large sample of proprietary commercial real estate transactions data from more than 73 countries within the sample period of 2001 through 2015. We investigate the impact of REIT creation on international real estate capital flows. Using non-REIT countries as the control group, we show that REITs open a new channel to attract institutional investments in real estate. New investments in real estate by both local institutional investors and foreign investors increase in REIT countries relative to non-REIT countries. The results imply that there is no substitution effects of REITs on private real estate market activities. We also find no evidence of “crowding out” effects on local investors, as we do not observe significant changes in the outflows of real estate capital. The relatively large increases in domestic real estate investments have caused declines in the shares of cross-border real estate capital flows in the post-REIT regime.

Individual Payoffs and the Effect of Homeownership on Social Capital Investment

This paper examines the role of homeownership in a household’s decision to invest in social capital. Because homeowners are more likely to improve the quality of their community and less likely to move, many housing advocates believe that homeownership brings about better citizenship. The empirical literature on the relationship between social capital and homeownership is, however, plagued by two particular difficulties. First, homeownership is an endogenous variable that is correlated with unobservable characteristics that describe good citizenship thus making instruments and more sophisticated econometric specifications necessary. Second, datasets of social capital investment choices are hard to come by. Using confidential and detailed household-level data, we are able to model the interrelated decisions of social capital investment and homeownership with an endogenous switching model that controls for selection effects. With appropriate instruments, we find strong evidence that homeownership increases the rate of social capital investment.Are all social capital investments equal in the eyes of homeowners and renters? Presumably, social capital investments that lead to increases in home values provide stronger incentives for homeowners than renters. In contrast, for social capital investments that do not directly impact home values, one would not expect homeowners and renters to differ in their investment rates. In this paper, we test this hypothesis using confidential and detailed individual-level panel data from Los Angeles county. We estimate the effect of homeownership on social capital investment, i.e., participation in social-capital creating activities, using a bivariate probit model and fixed effects models that control for individual-specific, time-constant heterogeneity that would otherwise cause omitted variable bias. Each model addresses the endogeneity of homeownership differently with identification arising from different sources. We find strong evidence that homeownership increases the rate of participation in block meetings, a social capital investment that should affect property values, and find no homeownership effect on three other social capital creating activities that likely do not: volunteerism, participation in a local political organization, and participation in a civic group. The results suggest that the effect of homeownership on social capital investment depends on whether the returns to such investments accrue solely to homeowners.

Changes in the White-Black house value distribution gap from 1997 to 2005

This paper examines the white-black house value gap in the United States. Instead of using standard conditional mean analysis and decomposition methods (via OLS regression), we estimate and decompose the changes in the white-black house value gap from 1997 to 2005 using quantile regression. We find that studying the entire distribution of house values is important. Our results show that racial differences in house values are mostly explained by differences in the characteristics of white-owned and black-owned houses but that the variation in the racial differences across the house value distribution are explained by racial differences in the marginal pricecs of housing characteristics. Moreover, between 1997 and 2005, we find that black-owned houses increased in value faster than white-owned houses in the lower value percentiles while the opposite was true for high-valued houses. In this case, we conclude that these changese were not caused by racial differences in house characteristics but by differences in the appreciation rate of the marginal prices of housing characteristics.

Fiduciary Investment Under Uncertainty: Examining Pension Plan Investment Behavior When Trustees Have Reputational Utility

Pension funds investment behavior have found to be persistent over time. Empirical findings of such persistence support conjectures that belief biases such as overoptimism and reputational concerns affect fiduciary investment behavior. Standard mean-variance analyses do not consider the implications of prudent-man laws on pension fund investment. We suggest an alternative way to derive pension portfolio choice by explicitly allowing pension trustee to make a joint belief-allocation choice in a utility function which takes an entropic form and considers investment objectives in a fiduciary environment. Our results demonstrate that pension trustees are inherently optimistic in forming projections but such optimism increases with the degree of risk-aversion. Further, the higher the level of risk-aversion, the more likely trustees will conform to peer group consensus and the less likely the consensus will change over time. This pattern of investment behavior seems consistent with empirical observations.