One under-appreciated feature of capital gains taxation are the substantial implicit benefits of deferral—capital gains are only taxed once they are realized (i.e., generally when the investment is cashed out), not when in any economic sense they are earned. Effectively, deferral renders any sufficiently long-term capital gain tax-free, regardless of what the statutory rate is.
To understand this counter-intuitive—after all, aren't long-term gain still taxed before you can ever get the cash?— conclusion, consider the following hypothetical:
Two investors, A and B, are similarly situated: Both have $1 million in cash to invest. Both are subject to a 40% marginal tax rate. Both have identified a company the shares of which, currently trading at $100, will appreciate by 10% p.a. indefinitely. (For simplicity, we assume that all the return from the company comes in appreciation, not dividends; splitting the return up would dilute the effect described here, but not eliminate it.)
The difference between A and B is that A buys the shares at the beginning of the period and sells them off after a number of years, i.e., takes advantage of deferral to only pay tax once at the end. Conversely, B sells all shares at the end of each year, pays taxes on his gains, and the repurchases the shares with what is left, i.e., does not use deferral and pays tax on his gains as they are earned.
Based on these assumptions we can calculate the after tax position of both A and B for any number of years. A Google Docs sheet giving the full set of figures and allowing you to play with the assumptions is available here.
The trajectory for A is fairly straight forward. He has 10,000 shares over the entire period and pays no tax until the end. If he cashed out at the end, he'd be left with about $2 million after tax, after 25 with about $7 million, and after 100 with over $8 billion.
The trajectory for B is a little more complicated. As B pays tax at the end of every year, the number of shares held into the following year declines, while their total value continues to increase. After 10 years, B is left with 6,900 shares and about $1.8 million, after 25 with 4,000 shares and $4.3 million, and after 100 with 246 shares and $339 million.
This substantial divergence is caused by a difference in effective tax rates. The pre-tax yield for both is always 10%. And B effectively pays 40% of this yield to the tax man, giving a constant 6% post-tax yield. Not so for A. If A cashed out after one year, he too would have a 6% post-tax yield. But every year after that A's post-tax yield increases, reaching 9.44% after 100 years and asymptotically approaching the 10% pre-tax yield.
The effective tax rate (i.e., the fraction of the pre-tax yield that is taxed away) for B remains constant at the statutory 40% rate. However, the effective tax rate for A, averaged over the period, drops every year, reaching 30% after ten years, 20% after 25 years, 10% after 55 years, and asymptotically dropping to 0%. A's effective annual tax rate drops even more precipitously, reaching 22% after 10 years, 10% after 20 years, and dropping below 1% after 46 years.
Now if you, along with some clever people, believe that all capital gains taxes are excessively distortionary and the correct rate is 0%, you will ascribe this deferral anomaly to the general incoherence of capital gains taxation. You might even be pleased that, thanks to deferral, the sufficiently long-term capital gains tax rate already is 0%.
But as long as we still have capital-gains taxes, the deferral feature will distort economic decision-making and freeze capital into less productive uses. To see how, consider an investor with a ten-year horizon who has the option of either investing and holding a single company over the entire period (i.e., act as investor A) or reallocate annually to what seems the best opportunities (i.e., act as investor B). If either type of investment yields 10%, A will earn a post-tax yield of 6.94%, while B will earn a post-tax yield of 6%.
But what if reallocating every year to new opportunities will be economically superior (i.e., have a higher pre-tax yield)? Because of A's tax advantage, B's yields would need to be at least 1.5% higher for B's post-tax yield to catch up. In other words, superior investments will not be made because of tax distortion. And over longer horizons, the extra hurdle for new investments becomes ever higher, reaching up to 6%.
This distortion seems to be almost self-evidently a harmful policy. But could the tax advantage of deferral be abolished? Not easily, as I will discuss in a future post.
