A common question we receive at Quorus is: Can a tax management program generate tax alpha for concentrated, active strategies? Plenty of research has been conducted on broadly diversified strategies with low turnover, but strategies with fewer holdings are different.
The first question to answer is should you be tax managing active strategies at all. In our view, the answer is an emphatic yes. You can generate meaningful tax alpha, incorporate intelligent gains deferral strategies, support targeted charitable giving, and personalize client portfolios all on active strategies. If you aren’t tax managing your active strategies you are leaving significant value on the table for your clients.
The challenge is that active strategies are like snowflakes; each strategy and manager is different. While the off the shelf option may work for market-cap weighted direct indexes, active is different. The level of turnover, number of holdings, volatility, and covariance of holdings can all impact tax management.
While each active strategy is unique, there are ways to evaluate how changes in holding count may impact the outcomes of a tax-loss harvesting program.
Concentrated strategies: a stylized example
By using after-tax backtests of historical and simulated stock data, we can illustrate how tax-management outcomes change with different concentrated portfolios.
To start, let’s look at a stylized example with portfolios holding 25, 50, 75, and 100 names. These are equal-weight portfolios that rebalance back to target weights each quarter but don’t have any other turnover. Using these portfolios, we simulate 100 different return streams using Monte Carlo simulation. Through this analysis, we take the median tax alpha compared to a benchmark SMA portfolio with no tax-loss harvesting and its associated tracking error.
There are two interesting points to note here. First, tax alpha is positive and meaningful across all four portfolios. Even in a portfolio with 25 names, harvesting losses can provide value to the right taxable investor. Second, tracking error increases with fewer holdings. This isn’t surprising because with fewer holdings there are less places to reinvest proceeds from harvesting trades.
Concentrated strategies: a case study with historical data
The stylized example is helpful to see the general relationship between tax alpha and tracking error, but what if we want to look at a more realistic scenario?
We built a hypothetical active strategy with targets for alpha, turnover, and holdings. The resulting portfolio had roughly 40 holdings and 40% turnover. In addition, we took a large cap index strategy with roughly 500 holdings and 3% turnover. We assumed each strategy could be purchased as an ETF and used the tax-efficient ETF as our after-tax benchmark.
We backtested the two strategies over a seven-year period from 2017 through the end of 2023. Our backtest uses real historical prices and dividends, starts from $1,000,000 in cash, and incorporates daily portfolio optimization using our production optimizer. In addition, our performance calculation properly nets short- and long-term gains and losses and pays the hypothetical client’s tax bill on April 15th of each year. For more detail on our backtester and a discussion of risks, see the disclosures below.
In our analysis, tax managing the hypothetical active portfolio delivers annualized tax alpha of 1.53% and tracking error of 1.78% compared to tax alpha of 1.15% and tracking error of 0.80% for the large-cap blend index.
Our results illustrate that concentrated portfolios can deliver positive tax alpha over an ETF but with higher tracking error.
So, is increased tracking error worth the benefits of tax management? If the client has the correct profile, yes. Retail investors mostly care about the risk of underperforming their benchmark; however, tax management introduces risk which is mostly symmetric (i.e., the risk of underperformance and outperformance is equal). If the portfolio outperforms the benchmark the value of tax management is simply additive to an already good story. If the portfolio underperforms the benchmark the advisor can point to the value generated through gains deferral and tax-loss harvesting.
How to incorporate tax management for active strategies
As advisors evaluate adding tax management to their process, they should look at their stable of investment strategies and clients. Consider the following:
Client profile
Does the client have a history of capital gains, particularly short-term gains? Does the client have a sizeable taxable account? Does the client live in a high tax state and have a high marginal tax rate? Client attributes will drive about 60% of the value from tax-loss harvesting.
Product vehicles
Is the product available in an ETF or only available in a mutual fund or SMA? An ETF will offer gains deferral with almost no tracking error, while an SMA will pass through losses with higher tracking error. Mutual funds and non-tax-managed SMAs will be the least tax-efficient.
Holding period
How long do you expect to hold the investment strategy? Does your investment team change managers as new opportunities arise? Will the client need cash from portfolio soon? Will the account be passed to heirs? All else equal, tax alpha decays over longer holding periods if there aren’t significant contributions to the portfolio. On the flip side, tax benefits are increased if the account benefits from a step-up in cost basis.
Target strategy
As we saw above, the characteristics of the strategy can impact the risk and return from a tax management program. Consider the investment strategy’s holding count, turnover, manager philosophy, expected alpha, and covariance.
Model the specific situation
For an index strategy with 500+ holdings a tax-management program that uniformly and aggressively harvests losses may be appropriate for the client because the resulting tracking error is only 0.80%. For an active strategy, the tax management program may require customized settings for tracking error, position constraints, and tax-loss harvesting aggressiveness.
Advisors may find that they want to limit tracking error at the expense of tax alpha and vice versa. Quorus can help advisors evaluate the after-tax risk and reward of tax management for their clients and strategies with detailed backtests.
Active portfolios require thoughtful design
The variety of methods for delivering pre-tax alpha in active strategies demands that tax management be thoughtfully constructed and delivered through a platform that is flexible. Advisors should work with a provider that can help them design and implement a tax-management methodology designed specifically for their strategy.
At Quorus, we thoroughly backtest active strategies and review the manager’s philosophy, process, portfolio, and performance to construct a tax management program that fits. In addition, our platform is cloud-native, customizable at the strategy, client, holding, and tax-lot level, and offers real-time reporting and client data.
Contact sales@quorus.io to learn more about adding tax management to your active strategies.
Disclosures
For Financial Professionals Only. Backtest results are hypothetical and for educational purposes only. They do not represent an actual investment in an SMA. Returns and portfolio characteristics are estimated based on historical price data, using daily portfolio optimizations and tax calculations. Backtest based on $1,000,000 initial investment and $10,000 monthly contributions. All capital losses are assumed to have offsetting gains with 35% short-term/15% long-term federal rate and 10%short-term/10% long-term state rate. All returns are shown net of trading costs and management fees. Time period for backtest is 12/30/2016 to 12/29/2023.Tax-loss harvesting is highly influenced by the market environment. As a result, portfolios established in different time periods may have vastly different results. Quorus’s relies on a portfolio optimization process to make daily determinations on appropriate trades. Changes in a number of factors, such as tax rate, market returns, cash flow, and tax loss harvesting aggressiveness, can create deviations in after-tax return and volatility.
The Monte Carlo analysis is based on a five-year investment period and assumes a $1,000,000 starting portfolio from cash. Each portfolio is run over 100 different return streams to determine median tax alpha and tracking error. Tax alpha estimated versus a non-tax managed SMA. Covariance matrix estimated using the average variance and covariance for the largest 500 US stocks from 12/31/2021 to 12/31/2023. Stock returns are generated using an expected return of 5% with an applied standard distribution of returns. All returns are shown net of trading costs and management fees. Tax-loss harvesting is highly influenced by the market environment. As a result, portfolios established in different time periods may have vastly different results.
The hypothetical active strategy is for illustrative purposes only, holdings for a hypothetical actively-managed Separately Managed Account were created according to a set of criteria. In order to determine appropriate criteria, strategies within the Morningstar US SA Large Blend peer group with at or below 150 long stock holdings were analyzed, with the upper limits of 150 holdings as a proxy to filter out strategies whose management approach is potentially more passive in nature. Of this strategy universe, the S&P 500 Index is the most prevalent Primary Prospectus Benchmark, and average annual Turnover, number of long stock holdings, and realized three-year tracking error versus the S&P 500 Index were approx.47%, 48, and 5.2%, respectively. Given the generally desirable characteristics of lower valuation, higher return on equity, and higher expected growth rate as exemplified in general models such as the Gordon Growth Model, an alpha score was generated in FactSet for stocks in the S&P 500 Index with equal weights to Next Twelve Months Earnings Yield, the average of the prior four last-twelve-months Return on Equity, and two-year forward implied Earnings Per Share growth, with the following constraints: maximum position size of 5%,minimum position size of 1%, range of long stock holdings of between 30 and 50,maximum annual turnover of 40%, Beta of 0.85-1.15 to the S&P 500 Index, and targeting 5% tracking error to the S&P 500 Index. As of 5/15/2024.
