Investing directly in the stocks underlying an index through separately-managed accounts opens up numerous benefits unavailable through pooled vehicles such as ETFs or mutual funds.  Of these, perhaps the most noteworthy benefit available to taxable investors is the potential for increasing your after-tax return through tax-loss harvesting. In fact, there are almost always tax-loss harvesting opportunities embedded within an index, but the benefits are unavailable to ETF and mutual fund investors. These and other pooled vehicles must treat all pooled investors equally, and often times must implement closely to the transactions required of a particular index.

The fact that directing investing unlocks this opportunity for harvesting embedded losses is well-established. What is less established is whether or not this technique yields benefits consistently across market conditions. Our analysis suggests that these benefits can be significant, and have historically been material across all market conditions, even in strong bull markets.

Direct stock indexing through separately-managed accounts (SMAs) is more realistic than ever at any account size, thanks largely to the emergence of fractional share technology and retail asset-based pricing within the investment management industry. This gives individuals (high net-worth or retail), or smaller-sized family offices or foundations exciting new alternatives to get broad-based stock exposure.


Tax-loss harvesting – how it works

Note: For an introductory discussion on Direct Indexing, see our eDocument available here

Tax-loss harvesting works by identifying portfolio holdings that could be sold in order to convert unrealized losses into realized losses. These realized losses are offset against taxable realized gains. By the judicious use of such offsetting trades, an astute investor can dramatically reduce or perhaps even entirely avoid capital gains taxes in a given year.

For a typical high-bracket investor, various studies have estimated tax alpha from stock ownership to range between 1.5% – 2% annually. While this number has gained wide acceptance as a general estimate, it presumes that the portfolio holds losing positions that can be traded to generate realized losses to offset the taxable realized gains. This raises the question of the extent of such offsetting trades that are available in different market environments.


Tax-loss harvesting’s value-add crosses different market environments

Intuitively, the payoff from tax-loss harvesting is most attractive when there are large losses in the market and high dispersion of returns between individual stocks. Conversely, a year with strong returns and low dispersion between stocks (as in 2017) should have a much lower payoff. In other words, tax-losses are easily harvested in high return/high dispersion years and less attractive in low return/low dispersion years.

In order to determine the scope for harvesting, we look at four different years that represented clearly distinct scenarios represented by these two variables, using dispersion and returns data for the S&P500®.

The years 2008 and 2009 saw extremely high dispersion between stocks (over 30%). By contrast, in recent years, the spread of stocks has been much more contained, with dispersion of less than 20%[1]. In recent years, only 2015 was a (very slightly) down year, and as everyone probably recalls, 2017 was a strong up year. So we looked at the following years:


Year Return Dispersion Environment
2008 -38.5% 33.3% Low Return/High Dispersion
2009 26.5% 30.1% High Return/High Dispersion
2015 -0.7% 19.0% Low Return/Low Dispersion
2017 21.8% 17.7% High Return/Low Dispersion


Next, we quantified the tax-loss harvesting opportunities with respect to the S&P 500® in these four years. The “Unrealized Loss %” is the percentage of the portfolio’s starting value each year that was available to harvest as a “loss” as of year-end. This represents a combination of the number of losing stocks, and their weights within the portfolio, and therefore represents the potential “tax-loss harvesting opportunity.”


Year Return Dispersion % of S&P 500 Stocks Down Avg Return of Down Stocks in S&P 500 Unrealized Loss % for the year
2008 Low High 95% -43.9% -39.4%
2009 High High 15% -15.0% -2.5%
2015 Low Low 58% -21.9% -9.0%
2017 High Low 27% -15.7% -2.7%
Note: Assumes no rebalancing within the year. Stocks delisted/removed from index mid-year are excluded from analysis

Note 2: All returns shown are price-only


One can see that even in the most bullish of years (2009), 15% of stocks had a “down year”! This is a year in which (like in any up year) holding an ETF or mutual fund would have yielded no opportunity to harvest any losses. While the Unrealized Loss % was certainly narrow in 2009, even this year saw opportunities for harvesting. Of course, conversely, 2008 saw incredible potential for tax-loss harvesting any gains, as 95% of the stocks in the portfolio were down for the year, with an average loss in those stocks of 44%.

Most importantly, in each of these years, there was always some availability of tax-loss harvesting potential in the form of unrealized losses. These ranged from nearly 40% of the portfolio to 2.5% of the portfolio that could have been used to offset taxable gains.

Anywhere from approximately 23% (if offsetting long-term gains) to ~43% (if offsetting short-term gains at the highest bracket) of the Unrealized Loss % for a given year is the potential boost to that year’s after-tax return. Also, depending on the strategy, there may be realized short-term gains that can be tactically offset by harvesting losses, thus improving the tax profile of the overall strategy (net gains are largely if not entirely targeted to be realized at the lower long-term capital gains rates). In addition, an investor may be able to take advantage of lower tax-brackets in the future (e.g. post-retirement, low-income year, inheritrance), thereby maintaining some control over when best to realize losses or gains.

Of course, eventually taxes are paid on accumulated gains. However, the ‘time value’ benefit of realizing losses early and deferring gains until later are significant. This added value generated from the harvested savings is compounded over subsequent years.


Tax-loss harvesting brings added after-tax returns to an investor that, unlike performance, is not subject to speculation or to “beating the benchmark.” The precise amount of added after-tax return will certainly vary based on market circumstances. However, it is worth noting that whatever that amount, tax-loss harvesting’s potential is entirely unavailable to an investor holding an appreciated mutual fund or ETF.

Optimal believes that direct indexing and tax-loss harvesting represent the next advancement in investing, and given its increasing viability at all account sizes, is something that all taxable investors and advisors should keenly explore.



[1]          Based on dispersion data received from S&P Dow Jones Indices, for which we gratefully acknowledge Fei Mei Chen and Craig Lazarra. See “The Dispersion-Correlation Map”,, June 2016

Monty Joshi, CFA

Monty Joshi, CFA has an MBA from the Haas School, UC Berkeley and a BA (Econ) from the University of Chicago. He has served in various of investment-related roles such as risk manager with the Chicago Board of Trade, equity options trader with Hull Trading (now part of Goldman Sachs), long-short hedge fund manager, and pension consultant with AD Biller and Associates. Previously, he managed finance/operations for several start-up companies in Silicon Valley.