
We describe a simple model of taxpayer decisions to realize or delay their capital gains and losses. Investors will delay their realizations if the after tax rate of return is sufficiently high. As the holding period for the asset approaches a year and a day (after which capital gains are taxed at a lower rate), taxpayers will tend to harvest their losses and delay their gains. Next, we use a unique data set of capital gains transactions to investigate the behavior of taxpayers with respect to the preferential tax rate for long term capital gains. Our data allows us to examine the shifting of gains across time periods, but eliminates the effect of the large pool of accrued gains that enlarge previous estimates. We find strong evidence that taxpayers respond to the preferential rate by reducing the realizations of gains in the weeks leading up to the point when that rate applies. However the magnitude of the elasticities are small: We estimate a short-term gains elasticity of 0.52 and a long-term gains elasticity of 1.0. We find that high income taxpayers are more responsive with elasticities of -0.75 for short term gains and 1.5 for long term gains. We also find evidence that taxpayers minimize their tax liability by timing their gains and losses in the same week.
| selected citations These citations are derived from selected sources. This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically). | 13 | |
| popularity This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network. | Top 10% | |
| influence This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically). | Top 10% | |
| impulse This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network. | Top 10% |
