The Costs and Benefits of Tax-Loss-Harvesting (TLH) Versus an ETF

Recently, we have experienced a rush of questions from investors/clients who seek our opinion on direct indexing/tax-loss-harvesting (“TLH”) and how it compares to the potential tax benefits of an ETF. Unfortunately, there is no easy answer because there are no silver bullets in finance: Direct indexing + TLH, ETFs, Mutual Funds, Insurance wrappers, SMA, Hedge Funds, DAF/foundations, and so forth, like everything in life, present cost/benefit trade-offs and there is no perfect solution for every situation.

In this post, we present a framework for thinking about the various “tax-efficient” solutions out there. We will focus on the specific comparison of tax-loss-harvesting in an SMA vs. investing via an ETF vehicle since that is where we have received the most questions.

Stepping back, when considering an investment we generally recommend a 2-step approach:

• Step 1: Understand what you own and why you own it (the most important step!).
• Step 2: Understand which investment vehicle will deliver you the most benefit for your unique situation (one can determine this on a DIY basis or by consulting their financial advisor).

Our firm mission is to empower investors through education so we are always open to help investors work on step #1, however, we are agnostic on step #2. In fact, in our own practice, we leverage multiple technologies and investment delivery mechanisms to help investors achieve a particular goal. For additional context, we started off our RIA/asset management careers as a tax-optimization direct indexing business when we started our business in 2010. Eventually, we transitioned our strategies (which are high turnover relative to a passive broad-based index) into the ETF structure. We believe the ETF structure served our client base more effectively via tax efficiency, easy access, and 100% transparency. All that said, we still manage a large managed account business that leverages direct indexing and tax-loss-harvesting capabilities. A good example is our 1042 Qualified Replacement Property business.

What follows is a simple framework for thinking through the expected benefits of tax-loss-harvesting relative to ETFs.

SMA Tax-Loss-Harvesting Versus ETF Cost/Benefits

First, a quick summary on the ETF structure and conducting a tax-loss-harvesting program within a separately managed account. In both situations, we will assume a long-only systematic US stock portfolio owned by a U.S. high-net-worth individual.⁽¹⁾ This analysis is meant to be simple and relatively easy to interpret. One can engage in heavy “analysis paralysis” when it comes to this topic, but always remember Warren Buffett’s advice: “Beware of geeks bearing formulas.”

Let’s first cover the primary potential tax cost/benefits of an ETF vehicle (a more detailed discussion is available here):

• Benefit: ETFs can leverage the custom create/redeem process to rebalance with very little, if any, immediate tax consequences for shareholders. Hence the reason most ETFs have no capital gains distributions.
• Benefit: ETF fees are tax-deductible because ETFs distribute net income, which is income + dividends minus expenses and fees. In contrast, financial advisory fees are generally not tax-deductible.
• Costs: ETFs, which are organized as registered investment companies, cannot distribute capital losses. In contrast, capital losses do pass through in a managed account.
• Costs: ETF portfolios cannot be easily customized to accommodate customized tax situations (i.e., dealing with a low-basis single stock position). In contrast, with a managed account, you can do whatever you want.

Next, here are the potential tax cost/benefits of conducting tax-loss-harvesting in an SMA:

• Benefit: Losses can be passed through to the account holder in a separately managed account.
• Benefit: The portfolio can be customized to the client account owner’s unique situation and/or preferences (e.g., ESG).
• Costs: TLH cannot manage portfolio rebalances or M&A situations (i.e., a cash buyout offer) without tax consequences.
• Costs: Fees associated with managed accounts are not tax-deductible.

The table below summarizes the basic costs and benefits for an SMA with TLH and an ETF structure.

Quantifying the Benefits of TLH versus an ETF

Here are the elements that feed into the equation investors will need to make when determining the relative tax benefits of TLH versus an ETF. NPV stands for net present value. We introduce this concept in this context because the costs and benefits of TLH vs. ETF have different horizons. By ‘discounting’ these costs/benefits to the current time period we are able to account for the time value of money and conduct an apples-to-apples comparison.

TLH versus ETF Net Benefits = NPV (short-term tax shields of TLH vs. ETF) + NPV (long-term tax shields of TLH vs. ETF) + NPV (trading costs differents of TLH vs. ETF) + NPV (index slippage of TLH vs. ETF) + NPV (fee differential of TLH vs. ETF) + NPV (relative transaction costs of TLH vs. ETF) + NPV (customization of TLH vs. ETF) + NPV (brain damage of TLH vs. ETF)

Where,

• NPV(short-term tax shields) = benefits of passing losses via an SMA.
  • ETFs cannot distribute losses (although fin engineering can create something similar via an ETF and leverage in an SMA).
  • TLH wins in this category, by design — ETFs can’t pass losses through, they can only carry them forward.


• NPV(long-term tax shields) = benefits of avoiding tax leakage from future activity (e.g., M&A; XYZ buys ABC in a cash offer).
  • ETFs win in this category because of in-kind create/redeem capability, which TLH can’t access.


• NPV (trading costs) = the relative transaction costs tied to TLH programs versus ETF programs, taking into consideration impact costs, payment for order flow costs, and commissions (which are now ~0).
  • ETFs win in this category because TLH requires additional trading activity to facilitate tax-harvesting.
  • We recommend you read Maneesh’s piece on this topic for more details.


• NPV (index slippage) = benefits of adapting to changes in index constituents over a long period (i.e., the SP 500 today is not the same as it will be in 20 years from now).
  • In general, there is some economic cost to slippage and/or a lack of flexibility that comes with TLH programs.
  • ETFs win in this category because they can transition to updates in a portfolio/index without tax consequences.


• NPV (fee differential) = benefit of lower management fee.
  • TLH programs are typically more expensive versus equivalent ETF strategies. ie. <5bps vs 35bps+.
  • Remember, this cost is huge on an NPV basis, e.g., a 30bp fee differential at a 10% discount rate equates to a 3% lump sum cost (perpetual annuity).
  • ETFs generally win in this category because the management fees are typically lower than managed accounts.


• NPV (customization ability) = benefit of customizing a portfolio to a unique situation such as a single stock position with a low basis.
  • Managed accounts are much more customizable than ETFs, which have portfolios that are ‘fixed in place.
  • Managed accounts can vote their own proxies and don’t delegate this duty to an ETF advisor.
  • Managed accounts win in this category because ETFs are less flexible in their ability to accommodate unique situations.


• NPV (complexity & brain damage costs) = benefit of managing 1 ticker versus 500 tickers in an SMA.
  • There is a real cost to the time consumed in understanding and digesting an investment strategy. This cost is amplified if one is managing assets on behalf of their family and/or descendants who may or may not be financially savvy.
  • Managing proxy votes and corporate actions can be a pain and/or costly if it is delegated to a 3rd party.
  • Complexity and brain damage also serves as an indirect “work program” for experts/advisors who seek to create ‘stickiness’ with a client’s account.
  • ETFs win in this category.

Is there a winner? TLH or ETFs?

Not really.

If one has unique circumstances where customization is highly valued (e.,g., ESG or managing taxes around single stock positions), TLH can be an awesome strategy. From the perspective of advisors, TLH is also interesting because it can offer them a tool to solve client problems and can serve as a method to ‘enhance stickiness.’

However, outside of unique circumstances, if we are seeking to maximize our after-tax after-fee after-brain damage investing benefits, the ETF is often the winner. This statement is truer for investment portfolios that require any level of turnover (e.g., factor strategies, active equity, tactical asset allocation). And this claim is often true for ultra-low-turnover strategies…the simplicity of buying a Vanguard/iShares ETF for very low costs with insane liquidity and tax efficiency is tough to beat in most cases.

One piece of good news is that you don’t need to only bet on one approach. You can dabble in both and/or mix them. For example, one can also consider hybrid approaches where they manage ETF strategies within an SMA. When one adds low-cost leverage to the mix, there are lots of interesting ideas to explore. (example here).

Good luck out there!