[] [] [prev-tail] [tail] [up]

The Structural Obsolescence of Mutual Funds in Taxable Accounts

For taxable portfolios exceeding $100,000, the classic open-end mutual fund — one without an ETF share class, a qualification the end of this section explains — is an obsolete technology. The obsolescence is driven by two structural forces:

1.
The Bessembinder Empirical Distribution: As section “Picking Stocks” details, all of the net wealth created by the U.S. equity market since 1926 came from roughly the top 4% of listed stocks; the other 96% collectively matched one-month Treasury bills.42 Active mutual fund managers do not reliably capture that asymmetry — their fees and closet-indexing constraints work against it — so you are paying an active fee for a distribution the manager is structurally unlikely to catch.
2.
Redemption-Induced Capital Gains Pass-Through: When a mutual fund faces heavy redemptions (such as during a market panic), the manager must sell appreciated securities to raise cash. The resulting capital gains are distributed and taxed to the remaining, loyal shareholders. You are effectively paying a tax penalty for the panic selling of other investors.

Direct Indexing as the Optimal Wrapper

The modern alternative to the mutual fund or standard ETF in a taxable account is direct indexing. Instead of purchasing shares of a fund that owns stocks, you open a separately managed account (SMA) where a quantitative trading algorithm directly purchases the underlying constituent stocks of an index (such as the S&P 500 or Russell 3000) in your name.

Direct indexing delivers two structural advantages:

Continuous Tax-Loss Harvesting

On any trading day a constituent stock declines below your cost basis, the algorithm automatically sells it to book a capital loss, immediately rotating the proceeds into a highly correlated substitute (e.g., selling PepsiCo to buy Coca-Cola) to maintain index tracking. To comply with the wash-sale rule under IRC §1091, the algorithm holds the substitute for at least 31 days before rotating back to the original name. The accumulated capital losses are passed directly to you, providing a “loss bank” to shelter capital gains realized from business sales, private equity exits, or real estate transactions. The literature here is reasonably settled: Chaudhuri, Burnham and Lo’s long-horizon study of systematic loss harvesting found average annual tax alpha near 1% for high-bracket taxpayers, with wide dispersion driven by entry timing and market path.133 Treat 1.0% to 1.5% as the plausible range in the early years, and read the decay curve below before capitalizing it into a decision.

Elimination of Phantom Gains

Because the account is yours, you are insulated from the transaction flows of other investors. You never receive a capital gains distribution that you did not personally authorize.

Managing the Harvest Decay Curve The high tax-loss harvesting yield of direct indexing is non-linear. As the portfolio appreciates, the cost basis of the rungs remains static while stock prices rise, reducing the volume of positions trading below basis. By year ten of a bull market, the annual harvest yield typically decays to 0.3% to 0.5%. To combat this decay, you should implement cash-funded top-ups (which introduce high-basis shares) and configure the account to systematically reinvest dividends.

Furthermore, the accumulated unrealized gains inside a direct indexing account are manageable:

The Onboarding Transition Transitioning an existing appreciated portfolio into direct indexing requires caution. Selling a legacy mutual fund or ETF to fund the SMA will trigger the exact capital gains tax you seek to avoid. To mitigate this tax shock:

The Offboarding Problem The transition above runs in one direction. Getting out has no equivalent machinery, and the asymmetry is the strategy’s least-discussed cost. Terminating the mandate does not liquidate anything — the manager simply stops managing, and you keep what is in the account. Where you previously held one line item you now hold three to five hundred individual positions carrying a decade of embedded gain, and no way to consolidate them that does not realize it.

The residue is not only a tax problem. You inherit the operational load the manager was absorbing: corporate actions, hundreds of dividend streams, and wash-sale tracking across every account you and your spouse control, including IRAs, where a repurchase permanently destroys the loss rather than deferring it (section “Wash Sales”). Whoever administers your affairs later inherits it too — five hundred lots with a decade of basis history is a materially worse estate to settle than two ETFs (section “Late-Life Vulnerability: The Plan for Diminished Judgment”).

The exits are the ones already described, ranked by what they actually cost: hold to death and take the step-up; donate the lowest-basis lots to a DAF; gift appreciated shares to family sitting in the 0% capital-gains bracket and let them sell; harvest whatever still trades below basis and apply carryforwards against the winners; or contribute the whole basket to an exchange fund (section “Concentrated Stock Positions”), which converts the positions into a diversified partnership interest at the price of a seven-year lockup. Failing all of those, pay the tax across several years against an annual gains budget. None of them is fast.

The decision rule follows directly, and it is stricter than the harvest math alone implies. The yield decays to 0.3–0.5% by year ten while the fee does not decay at all, so the crossover computed in section “Tax-loss harvesting” tells you when to stop funding the account — but stopping is not leaving. Adopt direct indexing only if you are genuinely committed to holding it until death, or that you have a specific large realization event the losses are being manufactured to shelter. If neither is true, you are buying a decaying benefit and accepting a permanent obligation to obtain it.

Where the ETF Wrapper Wins Below the $100,000 direct indexing threshold, or inside tax-deferred retirement accounts (where capital gains and losses carry no tax consequences), the broad-market ETF remains the optimal vehicle. ETFs use an in-kind creation and redemption mechanism with authorized participants, allowing them to adjust portfolio constituents without triggering fund-level capital gains.

The ETF Share Class, and Why the Mutual-Fund Verdict Is Now Conditional Everything above assumes the mutual fund and the ETF are separate vehicles holding separate portfolios. That assumption began dissolving in 2025, and it changes the calculus enough that you should check before writing off a legacy fund position.

Vanguard ran a patented structure from 2001 in which an ETF was simply another share class of an existing mutual fund, sharing one portfolio. Because the ETF class can perform in-kind redemptions with authorized participants, it purges low-basis lots from the shared portfolio — and the benefit accrues to every share class, including the mutual fund holders. That patent expired in May 2023. Roughly eighty asset managers filed for the same exemptive relief, and on 17 November 2025 the SEC granted the first of those applications to Dimensional Fund Advisors, establishing the template the rest are being approved against.

Two practical consequences:

Do not reflexively liquidate a legacy mutual fund position to buy the ETF.

If your fund’s sponsor adds an ETF share class, the tax-efficiency gap that justified the switch largely closes without you realizing a gain. Selling an appreciated fund to escape distributions you may no longer receive is the sort of self-inflicted tax bill this chapter exists to prevent. Ask the sponsor whether an ETF share class is filed or approved before you act.

Where a share class exists, converting is usually better than selling.

Converting mutual fund shares into the ETF share class of the same portfolio is generally not a taxable event, because you have not disposed of your interest in the fund — the same mechanic that made Vanguard’s Investor-to-Admiral conversions tax-free. Selling Fund A to buy ETF B is a realization event; converting Class 1 to Class 2 of Fund A is not. Confirm the treatment with the sponsor in writing, since it depends on the specific relief granted.

None of this rescues active mutual funds, whose problem is fees and the Bessembinder distribution, not wrapper mechanics, and none of it changes the direct-indexing case above the $100,000 threshold — a shared portfolio still cannot harvest losses at the individual-lot level on your behalf. What it does is retire the blanket claim that any mutual fund in a taxable account is obsolete technology. The correct question is now narrower: does this specific fund have, or plan, an ETF share class?

[] [] [prev-tail] [front] [up]