In October 2019, Charles Schwab began offering zero commission trades. TD Ameritrade, E*TRADE, and Fidelity soon followed. In December 2019, Robinhood introduced fractional share trading. Fidelity and Schwab both followed in the first half of 2020. At the same time, customized software-driven portfolio management was gaining ground, and by 2020 it was commonplace.

As these developments in portfolio management came into focus, investors were migrating away from actively managed funds to passively managed funds, or index funds. Index funds seek to closely replicate the performance of a specific market index such as the S&P 500.

The popularity of investing in index funds has risen steadily, and Morningstar reported that, in January 2024, investments in passive funds exceeded investments in actively-managed funds for the first time.

The Rise of Direct Indexing

The combination of software-driven portfolio management, fractional share trading, and zero commission trades created the possibility that investors who wanted to invest in an index could do so directly rather than by buying a mutual fund or ETF. An investor could own each of the specific securities comprising the index in the proportions held by the index. This is known as direct indexing.

Direct indexing has been widely marketed as the next evolution in personalized investing. Proponents claim it offers lower cost, better alignment with investor values, and tax efficiency.

At Rain Dog, we believe that, while the theory is attractive, the practical execution results in higher costs and failure to achieve the other objectives.

This white paper outlines three key problems with direct indexing and explains why traditional index investing remains a superior choice for most investors.

Mechanics of Direct Indexing

When an investor direct indexes the S&P 500, the investor will buy stock in 500 individual companies. For an investor with $1 million, Table 1 shows the amount that would be invested in the five largest and five smallest S&P 500 holdings as of June 16, 2025.

Table 1 Selected constituents of the S&P 500

 

Effect of Index Reconstitution

The S&P 500 is reconstituted quarterly. Some stocks are added and some are removed. The S&P 500 is also reconstituted in response to mergers and acquisitions. For the past several years, approximately 15 stocks have been added and 15 have been removed each year. The direct index investor must ripple the effects of these changes through their portfolio each time the index is updated, which is at least quarterly.

For stocks that are being added to or removed from the index, there is typically a spike in trading activity on reconstitution day, which has an effect on price, and that is usually temporary. Institutional funds with billions of dollars invested may take steps to neutralize the effects of these temporary price movements, but the direct indexer may not be able to employ the same sophistication with the relatively small amounts invested.

Effect of Sales

When an investor reduces a position in a direct index portfolio, each position is sold proportionately.

Suppose an investor has a $1 million portfolio and sells $10,000. That would result in sales of individual securities ranging from MSFT at $631.67 to CZR at $0.97. The effect of adding $10,000 to the portfolio is similar.

In practice, direct indexing doesn’t usually try to hold every stock, which avoids the noise in the smaller holdings. However, that introduces tracking error. A Wharton study modeled holding the 100 largest stocks had a standard deviation of 145 basis points (1.45%)–and that’s standard deviation, not absolute error.

Higher Costs

Although direct indexing is marketed as a low-cost solution, it is typically more expensive than simply buying a comparable mutual fund or ETF.

Maintaining the direct index portfolio by hand is labor intensive since it involves updating the proportions of hundreds or thousands of stocks at least quarterly. Because this is impractical, advisors use model portfolios that do that work for them. The model portfolios charge model fees that average 0.40% of assets under management (AUM), not including potential trading costs.

If the direct index fund attempts to hold all the stocks in the index proportionately, that introduces bid/ask trade inefficiency for trades like the $0.97 CZR trade mentioned above. If the direct index fund holds only a subset of the index’s stocks, that introduces tracking error, which can be significant.

In contrast, index funds from providers like Vanguard and Schwab routinely carry expense ratios of about 0.03% for their S&P 500 funds, and they can trade in large blocks that minimize bid-ask spreads and achieve tracking errors that round to 0.00%.

The S&P 500 is the most widely known index but it is not the only index, and there are funds based on many other indexes. These examples that are based on the S&P 500 likely understate the inefficiency of direct indexing insofar as most indexes track more than 500 stocks. The Russell 1000 tracks 1000 stocks; Russell 3000, 3,000 stocks; MSCI ACWI, 2,900 stocks; and CRSP US Total Market, 4,000 stocks; for example.

In summary, direct indexing increases costs in several ways:

• Fees for use of model portfolios,
• Friction from bid-ask spreads, especially on small trades, or
• Losses due to tracking error if the direct indexing avoids small constituents of the index, and
• Losses due to buying securities that are entering the index and selling securities that are exiting the index on or near reconstitution day.

These issues are exacerbated with indexes that hold thousands of securities rather than only hundreds.

Poor Tax-Loss Harvesting

Tax-loss harvesting is promoted as a core advantage of direct indexing—but its effectiveness is constrained by the nature of an “index” and by wash-sale rules.

Effect of “Index” on Tax Loss Harvesting

Small positions in indexes. Indexes include hundreds or thousands of individual securities, and the vast majority of individual stocks make up less than 1% of the portfolio each. In the S&P 500, only 15 stocks currently make up more than 1% of the index each. 420 of the stocks make up less than 0.25% each.

Losses on such small positions are rarely meaningful at the portfolio level. Harvesting a 20% loss on a 1% position yields only a 0.20% portfolio-level loss and an even smaller tax benefit.

For example, ORCL is the security that currently makes up closest to 1% of the S&P 500. If an investor started with a $10,000 position in Oracle in a $1 million portfolio and lost 20%, that would constitute a loss of only $2,000 and a tax benefit that is a fraction of that.

To appreciate the dilution of individual positions, consider the number of stocks and average weight per stock in common indexes shown in Table 2.

Table 2 Average weighting per member for selected indexes

Tax loss harvesting at the individual stock level produces only de minimis benefits.

Difficulty finding peers for tax-loss harvesting. When a stock is sold at a loss, wash-sale rules prevent repurchasing the same security for ±30 days. Investors must take one of the following actions:

• Hold cash for 30+ days and stay out of the market, which is a behavior with well-documented risks, or
• Find a substitute security with a similar expo-sure, which is a challenge when the index al-ready holds every major peer.

Consider the current top 5 holdings in the S&P 500. What peers would investors hold as substitutes for MSFT, NVDA, AAPL, AMZN, or GOOGL? The answer is that the top 5% of positions in the S&P 500 do not have any meaningful peers, while the bottom 95% of positions in the S&P 500 are likely to be too small to make a material difference in tax loss harvesting.

Impracticality of large scale tax-loss harvesting with direct indexing. How would an investor harvest losses in a market that is down 10% across the board?

Focusing on the largest stocks in the S&P 500, an investor could identify substitutes for the largest 22 stocks (which make up approximately 50% of the index in total). The investor could then sell those stocks and buy the substitutes. Then after 30 days the investor could sell the substitutes and re-buy the stocks. That approach would result in harvesting 50% of the possible loss while not truly being in the index for 30 days because of the unavailability of truly comparable substitutions.

The net effect of the largest-stock approach is that the investor would take on the market risk of not holding the most significant stocks in the index for at least 30 days, just to pursue a 5% loss (50% of the 10% across-the-board loss). The investor would also incur tracking error vs. the index.

The alternative direct-index approach is to focus on the smallest stocks. With that approach, an investor would go through the same process with about 475 securities—trading out of each of the 475 securities into comparable securities that are not already in the index—just in order to realize the other 50% of losses. This is both impractical and wildly inconsistent with the idea of direct indexing when 475 of the 500 stocks in the index are not being held.

Asymmetry of benefit from tax loss harvesting vs. after-tax returns. In the example above, assume the investor is in the 20% long-term capital gains tax bracket. For a $1 million portfolio, a 5% loss is $50,000, and the net after-tax benefit at a 20% tax rate is $10,000.

To realize a tax benefit of $10,000 on a $1 million portfolio, the investor must remain out of the market for 30 days with 50% of the portfolio. If tax loss harvesting causes the investor to miss out on more than $12,500 in gains, that erases the after-tax benefit of $50,000 in tax loss harvesting.

Contrast with ETF Tax-Loss Harvesting

ETF-based indexing offers a simple and elegant alternative. If the market drops 10% across the board, an investor can harvest the entire loss by, for example, selling their Vanguard S&P 500 Index Fund (VOO) and immediately reinvesting in the Dimensional US Equity Market ETF (DFUS).

Under current guidelines, the wash-sale rule does not apply. The investor realizes the full loss and stays invested at all times, and the portfolio remains true to the index.

Direct indexing complicates tax-loss harvesting at the micro level while producing minimal benefits at the macro level—and introduces tracking error at the same time. Index ETFs enable tax-loss harvesting cleanly and efficiently, at larger scale, with minimal or no increase in tracking error.

Drift Toward Active Management

Direct indexing is marketed as a passive investment strategy. However, in practice, it introduces elements of active management:

• Customization leads to deviation from benchmarks. Applying ESG screens, sector exclusions, or other personal preferences will cause portfolios to diverge from their intended indexes. The greater the degree of personalization, the greater the degree of tracking error that is introduced.
• Tax-driven trades introduce tracking error. Frequent buying and selling to harvest tax losses (while substituting not-really-comparable securities to avoid wash sale rules) can result in portfolios that do not track their benchmarks and do not mirror their performance.

This drift from passive to active management is concerning, especially given the historical performance of active funds. According to the SPIVA® U.S. Year-End 2024 Scorecard, a significant majority of actively managed funds underperform their benchmarks on a risk-adjusted basis:

• In 2024, 65% of all actively managed large-cap US funds underperformed the S&P 500.
• Over the 3-year period ending in 2024, 85% of all large-cap funds underperformed the S&P 500.
• Over the 10-year period ending in 2024, 92% of all large-cap funds underperformed the S&P 500.

The funds cited in the SPIVA scorecard might have underperformed the S&P 500, but at least those portfolios were constructed according to a coherent investment strategy.

Starting with a set of stocks that comprise an index and then selectively removing certain securities due to personal preferences or substituting others is not a rigorous portfolio construction technique. It is idiosyncratic at best, and it doesn’t take too many changes before “direct indexing” becomes a euphemism for haphazard stock picking.

Conclusions

Direct indexing has been presented as a next-generation investment solution, promising personalization, tax efficiency, and low cost all at the same time. But beneath the marketing veneer lurks a costly strategy that drifts toward de facto active management, which has been found to deliver reduced returns in most cases.

At Rain Dog, we believe that long-term success comes from rigorous asset allocation, a strict focus on minimizing costs, and thoughtful tax planning—which is best implemented via index funds and ETFs as the primary investment vehicles.

This Field Note was originally published as a Rain Dog white paper in June 2025.