Factor ETFs vs Direct Indexing: Cost, Exposure, and Tax Tradeoffs

Factor ETFs vs Direct Indexing: Cost, Exposure, and Tax Tradeoffs

Factor investing has evolved from an academic concept into a mainstream portfolio construction tool. Yet investors pursuing factor exposure now face a fundamental implementation choice: buy a packaged factor ETF, or use a direct indexing platform to construct a custom factor portfolio at the individual security level. Each approach carries distinct tradeoffs in cost structure, factor exposure purity, tax efficiency, and capacity. The academic evidence, combined with industry data from 2020 through 2025, suggests that neither approach dominates unconditionally. The optimal choice depends on portfolio size, tax situation, and how much factor exposure dilution the investor can tolerate.

This article compares factor ETFs and direct indexing across four dimensions that matter most for net-of-cost factor returns: total cost of ownership, factor loading fidelity, tax-loss harvesting alpha, and scalability. The goal is to provide a quantitative framework for deciding which vehicle delivers better after-tax, after-fee factor exposure.

The Cost Structure: Fees Are Only the Beginning

The headline expense ratio of a factor ETF tells only part of the story. Total cost of ownership includes the explicit management fee, trading costs embedded in the fund (bid-ask spreads, market impact during rebalances), and the tax drag from the fund's internal turnover. Direct indexing platforms charge a management fee (typically as a percentage of AUM), plus the investor bears trading commissions and market impact from holding hundreds of individual securities.

The critical insight is that tax drag differentiates the two approaches more than headline fees. Factor ETFs, particularly those targeting momentum or quality, exhibit annual turnover rates of 50% to 100%, generating capital gains distributions that the investor cannot control. Frazzini, Israel, and Moskowitz (2018) documented that trading costs alone erode roughly 50% of gross factor premiums for high-turnover strategies. When tax drag is added, the net premium shrinks further.

Direct indexing sidesteps the tax drag problem through individualized tax-lot management. Because the investor owns each security directly, losses can be harvested at the individual stock level without altering the portfolio's overall factor exposure. Berkin and Ye (2003) estimated that systematic tax-loss harvesting adds 100 to 200 basis points of annual after-tax alpha for high-bracket investors, depending on the volatility of the underlying universe and the investor's marginal tax rate.

Factor Exposure Purity: How Much Factor Do You Actually Get?

A core challenge with factor ETFs is exposure dilution. The theoretical factor portfolio, such as the Fama-French HML (value) factor, is a long-short construct with concentrated exposure to the desired characteristic. Commercial factor ETFs, by contrast, are long-only, diversified, and constrained by tracking error limits, liquidity screens, and index provider methodologies. The result is that a typical value ETF delivers a fraction of the factor loading that the academic factor definition implies.

Several forces drive ETF factor dilution. First, long-only constraints eliminate the short leg, which in many factors contributes meaningfully to returns. Arnott, Harvey, Kalesnik, and Linnainmaa (2019) showed that the long-short spread for value has historically been driven roughly equally by the long and short legs, meaning a long-only ETF captures only about half the theoretical premium before costs.

Second, diversification constraints limit concentration. Academic factor portfolios typically hold the top and bottom 30% of stocks by the relevant characteristic. ETFs often include hundreds of stocks across deciles 2 through 4 to reduce tracking error relative to the parent index, diluting the factor tilt.

Third, rebalancing frequency and buffer rules introduce staleness. Most factor ETFs rebalance quarterly or semi-annually, meaning the portfolio's factor characteristics drift between rebalance dates. A stock that has appreciated 40% since the last rebalance is no longer a value stock, yet it remains in the value ETF until the next reconstitution.

Direct indexing addresses these issues through continuous, customizable factor exposure. The investor can set tighter factor loading thresholds, exclude specific securities, rebalance more frequently (daily or weekly), and combine multiple factor tilts in a single portfolio without the layering costs of holding multiple ETFs. The tradeoff is greater complexity and the need for a robust optimization engine that balances factor exposure against transaction costs and tax efficiency.

Tax-Loss Harvesting: Where Direct Indexing Shines

Tax-loss harvesting (TLH) is the most frequently cited advantage of direct indexing over ETFs, and the empirical evidence supports the claim, with important caveats.

When an investor holds 300 to 500 individual stocks, there are always some positions trading below their cost basis, even in an up market. Systematic daily TLH identifies these positions, sells them to realize short-term or long-term capital losses, and replaces them with correlated substitutes to maintain factor exposure. The harvested losses offset capital gains elsewhere in the portfolio or up to $3,000 of ordinary income annually, with unlimited carryforward.

The TLH benefit is front-loaded. In the first one to three years, the portfolio has the most embedded losses to harvest. As the portfolio ages, cost bases reset lower, and the opportunity for loss harvesting diminishes. Berkin and Ye (2003) found that the TLH benefit decays to roughly one-third of its initial value by year 10. Furthermore, TLH creates a deferred tax liability: the replacement securities have lower cost bases, meaning larger capital gains upon eventual sale. The net present value of TLH depends on the investor's time horizon, expected future tax rates, and whether the portfolio will be donated or stepped up at death.

For investors in lower tax brackets, or those with limited capital gains to offset, the TLH advantage of direct indexing narrows substantially, and the simplicity and lower minimum investment of ETFs may dominate.

Capacity and Scalability Constraints

Factor ETFs benefit from enormous economies of scale. A $20 billion value ETF can offer an expense ratio of 15 basis points because the fixed costs of index licensing, fund administration, and compliance are spread across a massive asset base. Trading costs per share decline with fund size due to crossing networks and in-kind creation/redemption mechanisms that minimize market impact.

Direct indexing, by contrast, faces capacity constraints at both the small and large end. At the small end, investors with portfolios below $100,000 to $250,000 may not hold enough positions to achieve adequate diversification while maintaining meaningful factor tilts. The minimum viable portfolio size for effective direct indexing with tax-loss harvesting is typically $250,000, though some platforms have lowered this through fractional shares.

At the large end, direct indexing portfolios exceeding $50 million to $100 million begin to encounter market impact in smaller-cap names, particularly for size and momentum factors that require trading in less liquid securities. Frazzini, Israel, and Moskowitz (2018) estimated that market impact costs for a momentum strategy scale roughly as the square root of portfolio size, meaning a $100 million direct indexing portfolio faces roughly three times the market impact of a $10 million portfolio.

Net-of-Cost Factor Return Comparison

Combining the cost, exposure, and tax dimensions, we can estimate the net factor premium delivered by each approach under different scenarios.

The pattern is clear. For taxable investors with portfolios above $250,000 and marginal tax rates above 35%, direct indexing delivers meaningfully higher net factor premiums, primarily through tax-loss harvesting and reduced tax drag. For tax-exempt investors (endowments, IRAs, foundations) and smaller portfolios, factor ETFs remain the more efficient vehicle due to lower all-in costs and simpler implementation.

The momentum factor shows the largest gap between ETF and direct indexing net premiums because momentum has the highest turnover (creating the most tax drag in ETFs) and the widest factor dilution in long-only implementations.

Practical Decision Framework

The choice between factor ETFs and direct indexing reduces to four variables:

Portfolio size: Below $250,000, ETFs are generally more practical. Above $500,000, direct indexing becomes increasingly attractive.

Tax situation: Taxable accounts with high marginal rates benefit most from direct indexing. Tax-advantaged accounts (401k, IRA) should generally use ETFs.

Factor intensity desired: Investors seeking strong factor tilts benefit from the higher loadings achievable through direct indexing. Investors who want modest factor exposure alongside broad market returns will find ETFs sufficient.

Complexity tolerance: Direct indexing requires monitoring wash sale rules, managing hundreds of tax lots, and working with a platform that integrates with the investor's broader tax picture. ETFs require none of this.

Neither approach is universally superior. The empirical evidence suggests that taxable investors with sufficient scale should consider direct indexing for their factor allocations, while recognizing that the tax-loss harvesting benefit is front-loaded and decays over time. For tax-exempt portfolios, factor ETFs remain the default choice, with cost competitiveness that direct indexing cannot match at smaller scales.

References