Trading in a tax-managed account is largely driven by three factors: tracking error to the underlying strategy, harvesting losses and client-driven requests.

While the concept of tax-loss harvesting—realizing capital losses in a portfolio to capture a tax credit—has been understood for a long time, it has been difficult to implement, widely due to the amount of monitoring and trading that has to be personalized on an individual basis. Traditional tax-loss harvesting might be limited to replacing one ETF or mutual fund with another toward the end of the year—leaving opportunities associated with individual security holdings unused. Increasingly, however, technological solutions for direct indexing and automated monitoring have considerably increased the scope, and impact, of tax management.

While the behavior of a model tax-managed account may be well understood, looking at client accounts is very helpful as unusual dividends or additions by clients to accounts can significantly affect tax-loss harvesting by providing fresh tax lots to manage. With that in mind, as can be seen in the chart below, the median TaxMTM account at Neuberger Berman trades around 250 times a year, although a small percentage of accounts trade much more. Somewhat surprisingly, even the least active accounts (the 5th percentile of least traded accounts in the chart below) trade over 50 times a year. (Note: A sell followed by a buy is counted as two trades.)

Trades Per Year Across Neuberger Berman TaxMTM Accounts

 Inside TaxM: Trading 

Note: Universe of accounts consists of all NB TaxMTM client accounts that traded during the one-year period ending on Jan 31, 2022, that were incepted at least 3 months prior to Jan 31, 2022. For accounts that traded for less than a full year, the number of trades are annualized. Trades occurring during a transition phase at the beginning of the lifecycle of the account are excluded.

We take two broad approaches to tax management: gain deferral, where we seek to only realize capital losses, at the risk of seeing the portfolio diverge from the target over time, and dynamic overlay, where we allow for some realized gains to better match the target portfolio.

Unsurprisingly, when comparing the two, we see fewer trades in gain-deferral accounts over the course of a year, although there will be some that result from loss harvesting, dividend reinvestment, portfolio rebalancing, or when clients add or remove funds from their accounts.

Naturally, a portfolio looking to track a strategy with 30 names will trade less than a strategy with 3,000 names. Likewise, a portfolio that looks to track a very active manager will trade more than a strategy with less turnover. However, the general pattern holds.

Median Number of Trades in Certain Neuberger Berman TaxMTM Accounts by Strategy and Tax Approach

Median Number of Trades in Certain Neuberger Berman TaxM<sup>TM</sup> Accounts by Strategy and Tax Approach

Note: Four active strategies were selected, such that there were at least a few samples for every type of approach.

This level of trading for tax management would be all but impossible for most accounts as recently as a few years ago. However, technological improvements in direct indexing now make tax-loss harvesting much more accessible, and an important consideration in portfolio design and strategy.