Chunyu Yang

Management science 64(9) p. 4157-4176 Doi: https://doi.org/10.1287/mnsc.2017.2733

We study portfolio choice with multiple stocks and capital gain taxation assuming that capital losses can only offset current or future realized capital gains. We show, through backtesting using empirical distributions, that optimal equity holdings over an extended period are significantly lower on average than benchmark holdings suggested in the literature. Using Value and Growth or Small and Large portfolios, the backtests show that allocations remain persistently under-diversified. Carry-over losses have large economic significance since they can dramatically shrink the no-trade region. Finally, the backtested economic cost of incorrectly modeling capital losses is at least 8 percent of lifetime wealth.

Journal of Computational Finance 18(1) p. 121-143 Doi: https://doi.org/10.21314/jcf.2014.279

Journal of Financial Economics 109(3) p. 775-796 Doi: https://doi.org/10.1016/j.jfineco.2013.03.016

Empirical studies of household portfolios show that young households, with little financial wealth, hold underdiversified portfolios that are concentrated in a small number of assets, a fact often attributed to behavioral biases. We present a potential rational alternative: we show that investors with little financial wealth, who receive labor income, rationally limit the number of assets they invest in when faced with financial constraints such as margin requirements and restrictions on borrowing. We provide theoretical and numerical support for our results and identify the ratio of financial wealth to labor income as a useful control variable for household portfolio studies.