PDF Ebook Portfolios of the Rich
Ever since the pathbreaking work of Pareto more than a century ago, economists have known that wealth is extremely unevenly distributed. More recently, survey data have revealed that portfolio structures are also very different for households with different levels of wealth. While the portfolios of the rich are complex, the portfolio of financial and real assets of the median household (at least in the U.S.) is rather simple: a checking/savings account plus a home and mortgage, and not much else. Overwhelmingly, the data tell us that if we wish to understand aggregate portfolio behavior, it is critical to understand the behavior of the richest few percent of households, both because they control the bulk of aggregate wealth and because their portfolio behavior is much more complex than that of the typical household.
Though the foregoing arguments may seem to provide a compelling rationale for studying the portfolios of the rich, there has been little recent academic work in this area. The goal of this paper is to provide a summary of the basic facts about portfolios of wealthy households in the U.S. (and how the facts have changed over time) in a form which allows comparison of their behavior both with the rest of the population in the U.S. and with portfolio behavior among other groups and other countriessurveyed in the recent volume Household Portfolios edited by Guiso, Haliassos, and Jappelli (2001 (Projected)), and to make a preliminary attempt to understand the characteristics that will be required of any model which hopes to be consistent with the observed behavior.
The principal conclusion will be that the most important way in which the portfolios of the rich differ from those of the rest is that the rich hold a much higher proportion of their portfolios in risky investments, with a particularly large concentration of net worth in their own entrepreneurial ventures.
After the empirical conclusions are presented, the paper informally considers how these results relate to theoretical models of portfolio behavior. The starting point will be a standard stochastic version of the Life Cycle/Permanent Income Hypothesis model. That model will prove inadequate, however, because it implies that the rich should look like scaled-up versions of everybody else. They should have neither the extreme wealth-to-income ratios observed in the data, nor the unusual portfolio structures. The goal of the theoretical discussion will be to consider whether any of three potential modifications to the standard model might explain the observed combination of facts.
The first idea is that perhaps there is exogenous, immutable ex ante variation in risk aversion across households. In that case more risk-tolerant households would take greater risks and on average would earn higher returns. If owning a private business is the form of economic activity that offers the highest risk and highest return, one might expect that the most risk tolerant households would gravitate toward entrepreneurship, and on average would end up richer (though the failures might end up poorer). The paper will argue that this story has several defects, ranging from the fact that the empirical evidence fails to find a correlation between wealth growth and initial (expressed) risk aversion to the fact that, taken alone, the story provides neither an explanation for the lack of diversification of entrepreneurial investments nor for the tendency of wealthy households to hold much of their net worth in heir own entrepreneurial ventures.
These points lead to the second possibility: that the observed patterns are entirely a consequence of capital market imperfections, as suggested recently by Gentry and Hubbard (1998) and Quadrini (1999). Those authors argue that adverse selection and moral hazard problems require entrepreneurial enterprises to be largely self-financed. They further assume that there is a minimum efficient scale for private enterprises and that this minimum scale is large relative to the wealth of the typical household. The combination of these two assumptions can explain why households with low or moderate wealth or income are less likely to become entrepreneurs. Furthermore, this story requires no differences in tastes among members of the population, and in principle can explain both the high saving rates of the rich and the high portfolio shares in their own entrepreneurial ventures. However, this story too has problems. The first is that, in the absence of differences in preferences between the rich and the rest, the standard model implies that those households who have invested heavily in their own entrepreneurial ventures should try to balance the riskiness of these investments by holding all other assets in very safe forms. Instead, the non-entrepreneurial investments of rich entrepreneurs are much riskier than the portfolios of non rich nonentrepreneurs. A second problem with this story is that even the model with imperfect capital markets implies that as the rich get old, they eventually begin running down their wealth. In contrast, empirical data reveal no evidence that wealthy elderly households ever begin to run down their wealth.
The final possibility is that the model’s assumption about the household utility function needs to be changed in a manner similar to that proposed by Carroll (2000), who simply assumes that wealth enters the utility function as a luxury good in a modified Stone-Geary form. Because Max Weber (1958) argued that a love of wealth for its own sake is the spirit of capitalism, Bakshi and Chen (1996) and Zou (1994) have dubbed such models ‘capitalist spirit’ models. Carroll (2000) proposed this modification to the standard model as a way to explain the high lifetime saving rates of the rich, and argued that many different kinds of behavior, ranging from philanthropic bequest motives to pure greed, would result in a formulation of saving behavior that would be well captured by the modified model. An unanticipated consequence of the model is that it implies that rich households have lower relative risk aversion than the nonrich, which in turn could explain why the rich hold riskier portfolios than the rest, and why high-wealth or high-income young households are more likely to begin entrepreneurial ventures.
The one feature of the data that the ‘capitalist spirit’ model taken alone cannot explain is the tendency of entrepreneurs to invest largely in their own entrepeneurial ventures, which appears to require some form of capital market imperfection. The paper thus concludes that the main features of the data can probably be explained in a model which combines capital market imperfections of the kind emphasized by Gentry and Hubbard (1998) and Quadrini (1999) with a utility function like that postulated in Carroll (2000).
Download
PDF Ebook Portfolios of the Rich
Posted in :
