Continue active refreshing of this index's data?

Continue active refreshing of this index's data?

Blog Posts — May 1, 2020

Frequently Tax Optimize or Drift Away and Lose Tax Alpha

When clients invest in tax-managed investment strategies, their goal is to track the model portfolio, while harvesting as many losses as possible. The challenge, of course, is that these two requirements are at odds with one another. The act of loss harvesting requires selling some of the securities used to track the model portfolio. Hence, to reduce drift away from the model portfolio, other securities with similar characteristics must then be purchased. 

Since clients seek to harvest as many losses as possible, loss-harvesting trades must be made frequently. The question is, how frequently? And since clients require good tracking of the model portfolio, what should the risk level target be?

To answer the questions above, we must address the questions below:

  • What is the ideal tracking error to reduce model drift and allow for loss harvesting?
  • How frequently should the portfolio be rebalanced to achieve this?

As a first step, we need a way to measure the quality of potential solutions. But what? Total gains (or losses) realized? Some combination of total gains (or losses) and active return? Or, perhaps, some way of measuring the intrinsic value of short- and long-term gains (or losses)?

While total gains (or losses) realized is a valuable measure, it treats short-term and long-term taxes equally, which does not differentiate the tax impact between them. So, instead, how about treating the tax value of losses harvested as a credit, in much the same way as we treat taxes on capital gains as liabilities? This credit is basically equivalent to the taxes one would avoid paying in the future, thus increasing the portfolio’s return. For simplicity, we will call this tax alpha

Specifically, tax alpha incorporates the short- and long-term tax rates, combined with their respective net gains (or losses). In subsequent blogs, we will explore in detail the math and methodology to calculate tax alpha.

Now to the first question: What is the best rebalance frequency for a tax-managed investment strategy?

To answer the above, we ran numerous backtests to see what different rebalancing frequencies accomplish.

A total of 27 backtests was run using the Axioma Portfolio OptimizerTM from 12/31/2014-12/31/2019. The tests were conducted for nine different rebalancing frequencies and three different risk targets, using the STOXX® USA 900 Index as our model portfolio, and starting with cash. 

The three charts below summarize the nine rebalancing frequencies for the tracking risk targets of 10 bps, 25 bps and 50 bps. The frequency trade named Monthly After Wash Sale is a monthly rebalance the first day after the wash sale period.

  • Blue is the pre-tax active return of the portfolio relative to the STOXX USA 900
  • Orange is the tax alpha

As the rebalancing frequency increases, tax alpha increases, but then starts to flatten between monthly After Wash Sale to weekly frequencies. Monthly After Wash Sale and Weekly rebalancing seems to be generating the highest tax alpha in all three risk level targets, but it is marginal.   

In short, the best rebalancing frequency is difficult to determine, but it is clear that it is more often than bi-monthly and less than daily. A prudent starting point for any tax-managed strategy would be to rebalance about once a month, right after the wash sale period ends.

Now to the second question: What is the best tracking error to use while also loss harvesting? Let’s focus on the monthly after wash sale frequency for this question to compare the results and see. 

In the example below, the 10 bps risk target resulted in a yearly tax alpha of 69 bps; the 25 bps risk target resulted in a tax alpha of 1.34%; and the 50 bps resulted in a tax alpha of 1.62%. Taking on 15 bps of tracking error for 65 bps of tax alpha definitely makes the 25 bps risk target more attractive. The 50bps risk target is far less attractive, since it requires taking on 25 bps of extra risk to get 28bps of tax alpha. Although the 50 bps in risk results in more tax alpha, the extra risk isn’t providing strong enough returns to justify that extra risk.

So, the ideal risk target for this strategy is likely between 25bps and 50 bps. That said, using a monthly after wash sale rebalance frequency with a 25 bps risk target results in a very healthy level of losses, while also tracking the model portfolio well.

Nine Rebalancing Frequencies for Tracking Risk Targets of 10 bps

Source: Axioma Portfolio Optimizer, STOXX USA 900 Index

Nine Rebalancing Frequencies for Tracking Risk Targets of 25 bps

Source: Axioma Portfolio Optimizer, STOXX USA 900 Index

Nine Rebalancing Frequencies for Tracking Risk Targets of 50 bps

Source: Axioma Portfolio Optimizer, STOXX USA 900 Index

Table 1 below summarizes the yearly pre-tax active return and tax alpha for the monthly rebalance with 25 bps of tracking error starting with cash. Not surprisingly, the largest tax alpha is in year 1. But the exciting result is in year 5, when a 42 bps tax alpha is achieved, while the benchmark returned over 31%.

Table 1

Source: Axioma Portfolio Optimizer, STOXX USA 900 Index

As shown in this example, tax-sensitive investors have many components to work through in order to adhere to client’s specific mandates and provide strong returns. Minimizing portfolio drift while also harvesting losses is a challenge, but it is possible to achieve with the right tools. Learn more about how our tax-managed investing solution can help you automate and scale your tax strategies.