A cash cure for the curse of success

Tax-Smart SMAs can generate meaningful tax benefits for investors over time. However, their success in the early years means that these benefits generally decrease as an account ages, which is known as “ossification”.¹ This decline in potential tax benefits is ultimately due to two positive drivers: 1) the tendency of stocks to appreciate over time and 2) effective tax-loss harvesting, in which tax lots with a higher cost basis are sold and replaced by those with a lower cost basis. Over time, this combination tends to increase unrealized gains, which eventually become the larger portion of an account's value (Exhibit 1).

Fortunately, clients can combat ossification by contributing additional cash to their accounts, effectively resetting a portion of the account’s cost basis. In prior pieces, such as Stay one step ahead: Leveraging Tax-Smart SMAs throughout the investing lifecycle (April 2025), we discussed the potential impact of contributing an additional 5% of an account’s market value per annum to a Tax-Smart SMA as a way to extend the loss-harvesting potential of the account.

Having determined that cash contributions can help maintain an account’s tax-loss harvesting potential, we now explore if the amount of cash, the timing of the contributions or any other factors might have a further impact on the potential for tax benefits. In short, our analysis determined that:

• The amount of cash added has the greatest potential to combat ossification.
• The frequency of contributions has much less impact and is highly influenced by other factors.
• A strategy of combining cash additions with gifting can be an efficient option.

To test these options, we simulated a series of index-tracking Tax-Smart SMA portfolios (or “vintages”) that are initially funded with $1 million in cash and this time subsequently receive additional cash contributions under a variety of scenarios. We consider vintages dating as far back as January 1995, with each vintage proceeding to tax-loss harvest over a 10-year period, while seeking to align with the S&P 500 index in terms of sector exposures, stock exposures and predicted tracking error.

More cash, less ossification

We analyzed how much cash clients might need to contribute to achieve various levels of tax-loss harvesting opportunities. Generally, the more cash added over time, the more potential for tax-loss harvesting opportunities. We looked at three measures of success, in terms of combatting ossification and detail the findings below.

1. To maintain a constant pace of tax losses harvested in total dollar terms, contributing 5% – 7.5% may be sufficient (Exhibit 2).

As the account grows, the required contributions would grow in tandem and the efficiency of the account would decrease. For example, without even considering market appreciation, a $1 million account that receives contributions equal to 7.5% of its value on an annual basis for the next 10 years would have a total value of roughly $2 million (the initial $1 million plus about $1 million in cash contributions). Keeping a constant pace of loss harvesting in dollar terms as an account becomes twice as large may not feel sufficient.

2. Keeping the Tax Benefit² increasing at a constant rate may require as much as a 20% addition.

Another way to measure the efficiency or efficacy of a Tax-Smart SMA would be to think about the potential Tax Benefit that can be generated as a percentage of an account’s value, which considers both the tax avoided by realizing of capital losses and the benefit of re-investing this value rather than paying it out.

As a baseline for this objective, we observed that Tax-Smart SMA vintages that receive no additional cash contributions exhibited an annualized Tax Benefit of around 1% of account value over their respective 10-year investment horizons in this simplistic analysis, with a Tax Benefit ranging between 0.4% and 2.0% across the worst and best vintages.³

While we view this as an attractive level of potential Tax Benefit, the sequencing is front loaded, with the annualized amount averaging from 1% – 4% in the earlier years, but declining to less than 0.5% per annum in years 8 – 10. 

To see how much cash may be needed to keep the Tax Benefit increasing at a constant rate, we tested the impact of annual cash contributions ranging in size from 2.5% – 20% per annum and found that around 20% of an account’s value may need to be contributed in cash on an annual basis (Exhibit 3).

3. Keeping the Tax Benefit at 1% requires a 10% – 15% addition

If a near linear increase in the Tax Benefit is too high of a bar, a more modest target might be the level of cash needed to keep the average annualized Tax Benefit at or above 1% per annum. Exhibit 4 shows that between 10% and 15% of an account’s value may need to be contributed in cash each year in order to provide an account with enough “fresh tax lots” to take advantage of opportunities in the later years and maintain a Tax Benefit at around 1% of an account’s value per annum over a 10-year period.

Flexibility may be more helpful than frequency

Having concluded that continually adding cash to a Tax-Smart SMA helps stave off ossification—and the more cash the better—we then looked at whether the frequency of cash contributions made a meaningful difference in two ways: frequency of regular contributions and a lump sum vs. consistent contributions. We found that frequency did not have much of an impact and when it did, it was highly dependent on other factors. These results suggest that it is challenging to incorporate the frequency of cash contributions into a strategic decision from tax-loss harvesting perspective, and that clients should instead focus on what works best for them from an investment perspective more holistically across their portfolio.

Starting with the frequency question, Exhibit 5 shows that regardless of how much cash is added, the frequency of cash contributions does not appear to matter. Although the analysis did not find a material difference in the Tax Benefits from quarterly, semi-annual, or annual contributions, all else equal, less frequent contributions reduce constraints or restrictions on future trading activity, which could be beneficial from a tax-loss harvesting perspective.⁴

We also considered the scenario for clients that only have a fixed dollar amount to invest. One approach would be to invest all of the cash up front in a lump sum and jumpstart the loss-harvesting process, even if that leaves fewer opportunities later in the account’s lifecycle. The other option would be to hold back a portion of the initial investment, contributing cash over a longer horizon so that an account is continuously topped up with fresh tax lots that might generate greater loss-harvesting opportunities.

We looked at vintage simulations and considered the differences between a client investing $1.6mm all at once vs. incepting an account with $1 million and adding $600,000 over the course of the account’s investment horizon (contributing $15,000 in cash on a quarterly basis). Our analysis showed that both scenarios are impacted by a number of important variables, making it difficult to incorporate the results into a strategy.

The lump sum approach yields more losses harvested, on average than the more gradual approach but the pace of loss harvesting tends to decline over time. For the more gradual funding approach, the average level of losses harvested is lower over the 10-year period, but the path is steadier from year to year (Exhibit 6). Of course, there may also be investment implications to holding investments back versus deploying all at once, and we recommend that clients make these types of decisions from that investment perspective.

Gifting can contribute to a tax-loss harvesting strategy

Lastly, in seeking to maximize the efficiency of a Tax-Smart SMA, we explored the idea of combining cash contributions with gifting securities in the account that have appreciated. Gifting securities reduces the portion of an account that is ossified by unrealized gains, thereby improving the potential loss-harvesting outcomes as a percentage of account value. In addition, though beyond the scope of this paper, gifting stock to a qualified charity may help reduce an individual’s overall tax liability.

Going back to the 10-year vintage analysis from Exhibit 1, after running a Tax-Smart SMA for 10-years, on average, the account would have grown to $2.4 million with a cost basis of $1.1 million and $1.3 million of its value represented by unrealized gains (about 54%). At that point, if the investor contributed $250,000 in additional cash (slightly more than 10% of the overall account value), then the cost basis of the account would increase by $250,000 and the unrealized gains as a proportion of total account value would decline from approximately 54% to 49%. If the investor could contribute $250,000 in cash and also gift highly appreciated shares worth $250,000 from the account, the resulting proportion of unrealized gains in the account would decline all the way to roughly 44% (Exhibit 7). This strategy may not improve tax loss harvesting outcomes for the account in total dollar terms because the level of fresh tax lots is the same as in the cash contribution-only scenario, but it would improve the outcome as a percentage of (now reduced) account value. 

When considering the interaction of cash plus gifting, perhaps surprisingly, we observe that a 5% per annum cash contribution combined with a 5% per annum gift may lead to similar levels of Tax Benefit as compared to 10% per annum cash contributions on a stand-alone basis (as shown in Exhibit 8). Importantly, this conclusion speaks more to the overall efficiency of the account than the dollar amount of losses harvested. In fact, losses harvested in dollar terms from a 5% cash contribution scenario are quite similar to a 5% cash plus gifting scenario because the Tax Benefit improvement from gifting can be explained by a similar level of losses harvested from a reduced account value. We would still consider this a good outcome because it makes the account more efficient and may bring additional Tax Benefits outside of tax-loss harvesting.

Cash and gifting keep on giving

Ultimately, cash may be king in preventing tax-loss harvesting ossification. Across a wide range of market cycles and portfolio vintages, the most consistent takeaway is straightforward: the more cash added to an account over time, the greater the potential for tax-loss harvesting. In addition, we do not see an optimal frequency for cash contributions so long as contributions are spaced out enough to leave room for loss harvesting. However, the interaction of cash contributions and gifting can bring important Tax Benefits measured as a percentage of account value.

As always, the right approach depends on an investor’s broader objectives and portfolio context. We’re here to help clients evaluate these trade-offs and design a strategy that supports long-term tax efficiency while staying grounded in their unique goals.

¹ On average, our research has shown that about 70% – 80% of the tax-loss harvesting benefits of an account occur within the first five years. However, we see additional benefits to Tax-Smart SMAs in the gifting or withdrawal phase of an account’s lifecycle

² Tax Benefit (or “tax alpha”) is a measure of the difference between pre-tax excess returns and post-tax excess returns. Tax benefit is synonymous with “tax alpha”, as is commonly referenced in the industry and other JPMAM communication. 

³ This analysis assumed a simple, monthly loss-harvesting approach, however, the level of annualized tax alpha is in line with academic research and general rules of thumb across the Tax-Smart industry.

⁴ We did not include a monthly cash contribution scenario due to the potential for wash sale transactions. The wash sale rule pertains to purchases of substantially similar securities both 30 days before and after a security is loss harvested, which means that monthly cash contributions could keep an account continuously restricted if the levels of cash contribution requirement immediate investment of cash each month.