Are United States Household Portfolios Efficient

Abstract: The theoretical mean-variance efficient portfolio model was modified to incorporate human wealth and net primary residence. Eight traded assets were selected to represent the set of risky assets available to the household investors: combined stock index, large stocks, small stocks, the average return series for individual stocks in the CRSP Decile 10 (smallest) stock portfolio to proxy business ownership, corporate bonds, long-term government bonds, intermediate government bonds, and Ibbotson AssociateŁŒs real estate return series. Treasury bill represents the risk-free rate in this study. Simulation programs were developed to identify the efficient portfolios by finding the portfolio weights in risky assets that result in the minimum-variance frontier for the total portfolio. The results of the simulation programs give the efficient asset allocations to different household investors with different human wealth ratios, net primary residence ratios, and planned investment horizons, once the diversification of investment portfolios are related to the perceived stability of future employment income. The simulation results show that when rational household investors have a high human wealth ratio (e.g., those with ages between 30 to 40 years old), and a long investment time horizon (e.g., 15 year before their retirement), their efficient frontier is a combination of intermediate government bond, real estate, large stocks, small stocks and business ownership. People with high risk aversion should invest in intermediate government bonds and real estate for a 15-year horizon. People with low risk aversion should invest in real estate, small stock funds, and business ownership for a 15-year horizon. People who have risk aversion between these two points should choose a combination in the order of intermediate government bonds, real estate, large stocks, small stocks, and business ownership. The efficient portfolios from the simulation results are compared to the current portfolios of U.S. households estimated from the 1998 Survey of Consumer Finances. In the formal efficiency test of householdsŁŒ current portfolios, about one-third of total households hold inefficient mean-variance portfolios, compared with the same characteristics as those used to produce the simulation results in this study.

Until recently, researchers in economics and finance paid relatively little attention to household portfolios. Reasons included the tendency of most households to hold simple portfolios, the inability of the dominant asset pricing models to account for household portfolio incompleteness, and the lack of detailed databases on household portfolios in many countries until the late 1980s or 1990s. Now, however, the analysis of household portfolios is emerging as a field of vigorous study. The eleven chapters in this collection provide an overview of current theoretical knowledge about the structure of household portfolios and compare predictions with empirical findings. The book describes the state-of-the-art tools of analytical, computational, and econometric investigation, as well as some of the key policy questions. It provides an original comparative analysis of household portfolios in countries for which detailed household-level data are available (the United States, the United Kingdom, Italy, Germany, and the Netherlands). Finally, it uses microdata for an in-depth study of the portfolio composition of population groups of special policy interest, such as the young, the elderly, and the rich.

We use data from the Federal Reserve Board's Survey of Consumer Finances (SCF) to explore how household asset portfolios in the United States evolved between 1989 and 2016. Throughout this period, two key assets - housing and financial market assets - drove the household balance sheet evolution; however, we find a great heterogeneity in the balance sheets that averages and aggregates conceal. We observe that ownership of assets has become more concentrated over time, and we show that nearly all of the time series variation in financial vulnerabilities in family balance sheets is due to middle-income families, who hold most of their assets in housing and are often the most highly leveraged income group in the housing market. Tracking the evolution of wealth over time among birth-year cohorts, we observe the standard life-cycle asset accumulation processes among low-, middle-, and high-income families.

Theoretical and empirical analysis of the structure of household portfolios.

We address the issue of the efficiency of household portfolios in the presence of housing risk. We treat housing stock as an asset and rents as a stochastic liability stream: over the life-cycle, households can be short or long in their net housing position. Efficient financial portfolios are the sum of a standard Markowitz portfolio and a housing risk hedge term that multiplies net housing wealth. Our empirical results show that net housing plays a key role in determining which household portfolios are inefficient. The largest proportion of inefficient portfolios obtains among those with positive net housing, who should invest more in stocks.

"This paper proposes a method for predicting the probability density of a variable of interest in the presence of model ambiguity. In the first step, each candidate parametric model is estimated minimizing the Kullback-Leibler 'distance' (KLD) from a reference nonparametric density estimate. Given that the KLD represents a measure of uncertainty about the true structure, in the second step, its information content is used to rank and combine the estimated models.