Since Aperio offers tax-loss harvesting through both cap-weighted and factor-tilt strategies, we’re often asked, “Do I get more value from plain loss harvesting (cap-weighted) or from factor tilts, and does one affect the other?” To answer that question, Aperio recently published a whitepaper "Tax-Managed Factor Strategies." In that paper, we identify the ways in which factor alpha and tax alpha have historically interacted for taxable portfolios. One of the findings of the paper is that for certain historical periods, a taxable investor may require a higher pre-tax return from a particular factor tilt in order to offset the reduction in the benefits from loss harvesting. Here, we’ll focus on what we’ll define as “Tax Alpha Differential,” the drop in after-tax return caused by switching from cap-weighted to factor strategies.
Figure 1 shows Tax Alpha Differential (TAD) by comparing the percentage of historical tax alpha generated from a factor strategy versus a cap-weighted strategy. We define the TAD as the drop in tax alpha attributable to the switch in the simulation from a cap-weighted strategy to a factor-tilt strategy. The single-color orange column on the left represents a base case of a cap-weighted strategy, which of course shows as 100% of potential tax alpha when measured against itself. Based on a series of back-tests from 1995 to 2017,* the tax alpha for that period averaged 2.23% per annum for US portfolios, although investors may want to use caution in expecting that high a level of benefit since the two bear markets of 2000–02 and 2008–09 are included in the period analyzed. That’s what we see with plain-vanilla loss harvesting (i.e., when using a cap-weighted index like the Russell 3000), which is higher than the tax alpha from the factor strategies we’ve analyzed.
The column on the right shows the tax alpha for a factor strategy we describe as Quality Light (see our "Tax-Managed Factor Strategies" paper for more details). For that strategy over the same time period, the tax benefit of loss harvesting remains substantial at 70.5% of what was earned from the cap-weighted approach. The 29.5% differential (or shortfall) can be viewed as the extra return that an investor would need to earn from the factor bet in order to compensate for the lower tax alpha. While this differential provides a highly useful metric for certain situations, it’s not applicable for all the choices taxable investors face. For the sake of argument, we’ll divide investors into two types, those focused primarily on taxes and those focused primarily on factors. Each type may care about both, but we’ll assume that either taxes or factors still counts as the most important driver in decision-making. For the tax-focused investor, the differential metric provides a useful way to think about whether or not to choose a cap-weighted benchmark or a factor strategy.
What about investors who focus more on factors than on tax benefits? Those investors may be considering the choice as between a factor-tilted ETF (exchange-traded fund), for example, versus a tax-managed SMA (separately managed account). Those investors might focus on the “Remaining Tax Alpha” from the Quality Light strategy in Figure 1. Such investors, who presumably feel high conviction regarding the benefit of the factor tilt and are going to choose a factor tilt either way, would consider the Remaining Tax Alpha as a potential benefit of switching to an SMA because of the tax benefits.
Figures 2 and 3 below show the choices and metrics faced by two hypothetical investors.
Within Aperio, we’ve even debated which metric provides more benefit, the Tax Alpha Differential or the Remaining Tax Alpha. Then we realized, as is often the case with differing perspectives, that each approach works well in different situations, depending on the choices an investor may face and their priorities and assumptions.
How do these two metrics look across a range of factor strategies? Table 1 below shows the two percentages for seven different strategies.
In the simulation, the first four factor strategies shown on the left in Table 1 all reflected a market beta of one and also retained a relatively high percentage of the tax benefit. For lower-risk strategies with a below-market beta, the tax benefits retained appear to be lower due to a smaller number of available stocks and tighter constraints inherent in factor-tilt strategies. In some cases, the Remaining Tax Alpha can even go negative, common in active management. In practice, many such SMAs might be managed less robotically than in a back-test to keep the tax alpha above zero. However, these outcomes remain instructive in terms of the impact on tax alpha of switching to certain factor strategies. Furthermore, even a low tax alpha may still provide improved tax flexibility for managers who seek to change their factor tilts tactically over time, where use of ETFs may trigger booking all gains since inception as opposed to a more tax-optimized transition available within an SMA.
As can be seen in Table 1, global factor strategies retained more of the tax alpha than their domestic counterparts, which may reflect the larger universe of stocks available.
In summary, investors can make better decisions by considering the Tax Alpha Differential or Remaining Tax Alpha metric, depending on the factor tilt(s) they prefer and their emphasis on factors or taxes, as well as whether they’re considering domestic or global strategies.
Send questions or comments to blog@aperiogroup.com.
