After-tax returns are higher when we apply a simple ETF-pair-based tax loss harvesting strategy.
We are not the only ones who can do this, but we have the ability to display an individualized benefit to clients
based on their tax bracket, external holdings, etc.
Background
Tax loss harvesting (TLH) is a passive investing strategy
(i.e. has no opinion on which way prices will go) which increases after-tax returns
in taxable accounts.
The most commonly implemented TLH strategy is a simple one based on pairs of ETFs.
We choose a pair of ETFs (per asset class) which are:
similar enough that they both track their corresponding asset class very well
different enough to avoid triggering the IRS wash sale rule,
which disallows you from claiming a tax loss if you buy a "substantially identical" security within 30 days
of having sold it at a loss.
cheap enough to trade (small spread)
cheap enough to hold (low expense ratio)
A good example is two ETFs that track the 500 and 1,000 biggest US stocks, respectively.
In practice, their prices move almost in unison. However, they are not "substantially identical" per the wash sale rule.
Pairs-based TLH monitors ETFs trading at a capital loss.
When the price of an ETF held in the account has declined enough
where it would generate a tax loss above a (configurable) threshold,
it is sold and replaced by purchasing the similar ETF in the pair.
This maintains the desired asset exposure, while "harvesting" a tax loss.
After-tax returns are likely to be higher because it is almost always better to postpone tax payments until later:
"Time value of money": The money that would have been paid in taxes upfront gets to grow with the market instead.
For investors who will be in a lower tax bracket when they sell their assets later in life, they will owe less tax.
[Note: Currently, assuming the tax law does not change, this benefit would be biggest for short-term capital gains,
which get taxed at ordinary income rates, which can drop a lot during retirement.
However, for a scenario of selling assets in retirement,
the benefit effectively applies only to long-term capital gains,
where the differential would be limited to a possible avoidance of the 3.8% Medicare surtax.]
There are cases where no tax is owed at all, e.g. when assets are passed down to someone's heirs,
or in a charitable contribution.
[Note: Things get more complicated due to an annual limit of tax losses that can be claimed against ordinary income,
but there are often enough capital gains in high-end accounts from other sources (e.g. sale of a home)
that the full amount of harvested losses will reduce tax liability.]
Of course, in the few cases where it is not better to postpone taxes, or if the client does not want it, TLH can be disabled.
Scenario parameters
Held by us: $1,000,000 in cash, which we are free to invest
External: no external assets (to keep the test simple),
but the system is perfectly capable of handling external assets while also performing TLH.
Target allocation: in both cases, a balanced mix of
7 asset classes
The fairest way to compare the results with vs. without TLH is to look at the after-tax liquidation values.
TLH works by realizing losses today at the expense of postponing gains.
We cannot ignore the fact that tax would have to be paid at some point, because it would make TLH look even better than it is.
We display totals using three sets of liquidation assumptions:
Held by us: unrealized taxes never have to be paid (e.g. when passing down to heirs).
Held by us (liq. tax): any tax owed at any given point in time gets subtracted from the final portfolio value.
Held by us (liq. tax; ST taxed LT): long-term tax rates are applied even to short-term gains / losses.
This metric is less realistic; however, we prefer it, because it avoids abrupt value changes around the days when short-term tax
lots become long-term, such as one year after a large investment.
Note how the dark green line (TLH enabled) is higher than the light green line (TLH disabled) using either
Held by us (liq. tax) or
Held by us (liq. tax; ST taxed LT).
If we search for "money-weighted returns" in the data grids for TLH
disabled vs.
enabled, we can see the annualized difference is
0.84% (551.7 - 467.3 bps) for pre-tax and 0.31% (424.4 - 393.3 bps) for after-tax returns, in this particular example.
This demonstrates that even after-tax portfolio returns are higher with TLH, not just the pre-tax returns.
It is also interesting to compare the Net taxes owed tabs, or view the difference in a single chart
here.
With TLH disabled,
taxes owed (remaining to be paid) at the end of the backtest are $30,483.
This indicates that the TLH approach has paid fewer taxes, since more taxes have been deferred.
Of course, all else being equal, it is better to owe fewer taxes than more. However, in this case,
more tax owed is a worthwhile price to pay because the increased portfolio returns more than make up for it.
Conclusion
Most automated advisors offer this simple strategy;
however, we have the ability to display what happens based on a client's tax bracket and other inputs
(deposit patterns, fees charged, external holdings, etc.)
