The 401(k) Brokerage Window: Evaluating the Impact of Expanded Investment Choice on Retirement Outcomes

Why unlimited investment freedom may be exactly what a retirement plan should not provide

The self-directed brokerage account, or SDBA, sounds almost irresistible. Instead of being confined to the twenty or thirty investment alternatives selected by an employer's 401(k) committee, an employee can open a brokerage window and choose among thousands of mutual funds, exchange-traded funds, individual stocks, bonds, and, depending on the platform and plan restrictions, other investments. The apparent advantage is freedom, but retirement-plan design should not be evaluated by the number of choices it provides. It should be evaluated by the probability that participants will actually accumulate sufficient retirement wealth.

Those are not the same objective. A substantial body of retirement, behavioral finance, and portfolio research suggests that the central problem confronting the ordinary 401(k) participant is not insufficient investment choice. It is the difficulty of making consistently rational investment decisions over several decades. "An SDBA may introduce additional decision complexity and analytical responsibilities that many retirement plan participants may not be prepared to manage

What the Brokerage Window Actually Changes

A conventional participant-directed 401(k) normally offers a curated menu of designated investment alternatives. The plan fiduciaries select the menu, monitor it, negotiate service arrangements, review expenses, and commonly provide diversified target-date funds, index funds, bond funds, stable-value or money-market alternatives, and perhaps managed-account services. An SDBA opens a door outside that menu by allowing participants to select investments beyond those designated by the plan.

The Department of Labor distinguishes brokerage-window investments from designated investment alternatives. Investments selected through a brokerage window generally are not treated as designated investment alternatives merely because they are available through the plan. That distinction matters because the plan's standardized disclosure obligations apply differently to a limited menu of designated alternatives than to the thousands of securities that may be accessible through a brokerage account.

The fiduciary has not disappeared simply because a brokerage window exists. Plan fiduciaries remain responsible for matters such as prudently selecting and monitoring service providers and considering the nature and quality of the brokerage arrangement. What has changed is that the individual investment selected through the window ordinarily has not been screened and affirmatively designated by the plan's investment committee. The participant therefore assumes substantially more responsibility for investigating and monitoring the securities purchased through the window.

This is one of the less visible consequences of an SDBA. What appears to be expanded investment opportunity is simultaneously a transfer of analytical responsibility from institutional processes to the employee. That transfer may be appropriate for a sophisticated investor, but it is not automatically beneficial merely because it enlarges the range of available investments.

Most Participants Do Not Suffer From Too Little Choice

The premise underlying an SDBA is that a participant may need more investment alternatives than the employer has provided. For a financially sophisticated participant with unusual portfolio requirements, that may be true. It is much more difficult, however, to argue that lack of choice is the dominant problem in American 401(k) plans.

Research by Ning Tang, Olivia Mitchell, Gary Mottola, and Stephen Utkus examined nearly one million participants in approximately 1,000 defined-contribution plans. The researchers found that the overwhelming majority of plan sponsors already provided efficient investment menus. The more serious inefficiency arose after the menu had been created: participants frequently constructed inefficient portfolios from otherwise adequate investment choices. Their analysis estimated that poor participant portfolio decisions could reduce potential retirement wealth materially over time.

That finding changes the way the SDBA question should be framed. If the core investment menu were itself the primary failure, expanding the menu would be an obvious solution. If the principal problem is instead the participant's ability to construct an efficient portfolio, providing thousands of additional choices may intensify rather than correct the problem. The relevant issue is therefore not simply access to investments, but the quality of the decisions participants make once access is expanded.

Choice Overload Is Not Merely a Theory

Behavioral economics has repeatedly challenged the conventional assumption that more choice necessarily produces better outcomes. Sheena Iyengar, Wei Jiang, and Gur Huberman examined 401(k) participation using records involving nearly 800,000 employees and found that participation tended to decline as the number of investment alternatives increased. Their work estimated that every additional ten funds in the menu was associated with roughly a two-percentage-point reduction in participation.

Participants confronting larger menus also showed evidence of using simplifying heuristics. In some cases they shifted more heavily toward money-market and bond funds and away from equities. The exact magnitude of these results should not be mechanically applied to every modern retirement plan because plan design, automatic enrollment, target-date funds, and default structures have evolved. The behavioral principle, however, remains highly relevant: as the complexity of a decision increases, participants do not necessarily respond by conducting more sophisticated analysis. They often respond by simplifying, delaying, or avoiding the decision.

A menu of twenty carefully selected alternatives presents a manageable allocation decision. A brokerage platform containing thousands of securities presents an ongoing research obligation. Most employees did not enroll in a retirement plan because they wished to become security analysts, yet the SDBA structure can place precisely that burden upon them.

Participant Portfolio Construction Is the More Serious Problem

One of the most consequential findings in the retirement literature is that professional plan design and individual investment behavior often move in opposite directions. Tang and his colleagues found that approximately 94% of the plan menus they examined were efficient relative to the benchmark methodology they employed, yet many participants subsequently assembled inefficient portfolios from those same menus.

The implication is important. In many cases, the employer had largely solved the diversification problem through the construction of the core menu, and the participant recreated the problem through portfolio selection. An SDBA gives that same participant a vastly larger universe in which to repeat the error.

This does not mean participants are unintelligent. Portfolio construction is genuinely difficult. Appropriate diversification requires consideration of expected return, volatility, correlation, time horizon, sequence risk, liquidity, employment risk, outside assets, taxes, and changing retirement objectives. Investment selection also requires distinguishing between compensated and uncompensated risk. An employee who places a substantial percentage of a retirement account in a favored technology stock may believe that he has identified an opportunity, while from a portfolio-theory perspective he may simply have added uncompensated idiosyncratic risk to an asset pool whose purpose is to finance decades of retirement income.

Vanguard has itself cautioned plan sponsors that participants with insufficient expertise can jeopardize retirement savings by concentrating in a single stock, industry, or fund and that brokerage access creates the possibility of overtrading and day trading. Its analysis concludes that a brokerage option is not appropriate for every plan or participant population. That warning is significant because it comes from a major retirement-plan recordkeeper rather than from an institution philosophically opposed to participant choice.

The Brokerage Interface Changes Participant Behavior

The brokerage account also changes the psychological character of the 401(k). The traditional retirement account is designed primarily as an accumulation vehicle, whereas the brokerage interface is designed to facilitate transactions. Those purposes are not identical, and the distinction matters because increased ease of trading can alter investor behavior.

Research using more than one million 401(k) participants found that high-turnover trading was costly, while periodic rebalancing could be beneficial. Participants holding only balanced or lifecycle funds produced among the strongest risk-adjusted outcomes in the study. Another large study of approximately 1.2 million participants found extraordinary inertia among ordinary 401(k) investors: roughly 80% made no trades during the two-year observation period, while another 11% made only one trade. The relatively small group that traded more actively was disproportionately older, higher-income, male, and more financially engaged.

An SDBA therefore does not merely permit better diversification. It also permits participants to convert long-term retirement investing into active securities selection. Market timing, performance chasing, sector rotation, concentrated stock positions, emotional selling during market declines, and buying securities after conspicuous price increases all become easier. The fact that a transaction can be executed effortlessly does not mean that the decision itself is economically sound.

Fees Quietly Compound Alongside Everything Else

A second underappreciated issue is cost. A large employer-sponsored 401(k) may have access to institutionally priced funds, collective investment trusts, low-cost index vehicles, or other negotiated arrangements. Moving through the brokerage window can expose participants to separate account charges, trading expenses, retail fund share classes, investment-management expenses, transaction costs, and other charges depending upon the provider and investment selected.

Fee structures vary by plan sponsor, brokerage platform, and selected investment vehicle. Participants should carefully review all applicable platform, transaction, and management fees before utilizing a brokerage window. The Department of Labor's examination of brokerage windows noted examples involving annual brokerage maintenance fees and emphasized that mutual-fund share classes available through brokerage windows may differ from those available within the retirement-plan menu. Participants therefore need to compare not merely performance but total cost between investments available through the core plan and those purchased through the brokerage account.

The significance of those costs is often obscured because fees are quoted annually. Their economic effect, however, is cumulative. The Department of Labor has illustrated the mathematics with a hypothetical participant beginning with $25,000 and earning 7% for 35 years. If fees reduce returns by 0.5 percentage point annually, the account would grow to approximately $227,000. If expenses instead reduce returns by 1.5 percentage points, the ending value falls to approximately $163,000, a difference of roughly 28% attributable to one additional percentage point of annual cost over the accumulation period. Mathematical Disclaimer: "This hypothetical mathematical illustration is for educational purposes only and does not project or guarantee future investment performance or fee impacts. Actual account values will fluctuate based on market conditions, actual returns, and specific plan expenses.

That is the central problem with seemingly modest investment expenses. Every unnecessary charge must be overcome by additional investment performance merely to restore the participant to the position that would otherwise have existed.

More Funds Also Means More Opportunities to Buy Inferior Funds

The assumption that a brokerage window improves a plan because it contains thousands of investments ignores a basic statistical fact: expanding the universe of choices expands the number of poor choices as well as good ones. Ian Ayres and Quinn Curtis, examining more than 3,500 401(k) plans with more than $120 billion in assets, documented the problem of high-cost and economically dominated investment alternatives.

Their research estimated that the combination of fees and portfolio inefficiencies could impose meaningful costs relative to low-cost index alternatives. They also observed that brokerage windows potentially expose participants to hundreds or thousands of funds, including expensive alternatives that could leave participants worse off than they would have been under a constrained menu.

A platform containing 5,000 funds therefore does not give an investor 5,000 good investments. It gives the investor thousands of possibilities that must be distinguished, compared, and rejected or selected on rational economic grounds. The additional task of identifying the appropriate investment is itself part of the cost of expanded choice.

SDBA Users Are Not Typical Participants

There is an important counterargument. The people who actually use brokerage windows are not representative of the average 401(k) participant. Vanguard reports that brokerage participation is extremely low. In 2023, approximately 21% of its full-service plans offered a brokerage option, yet only around 1% of participants with access used it. Vanguard also reports that participants with larger balances disproportionately use brokerage windows.

Current Schwab data show a similar selection effect. Its SDBA participants have relatively large average balances, indicating that many are substantial retirement savers rather than novice investors with very small accounts. Many also obtain professional advice. These characteristics prevent any defensible conclusion that SDBAs necessarily produce inferior outcomes.

A high-balance participant using an SDBA to obtain a particular low-cost ETF, specialized bond strategy, or asset class missing from the core lineup may be behaving entirely rationally. Nor is there persuasive evidence that every SDBA participant underperforms the core plan menu. Much of the strongest empirical literature examines participant behavior, choice complexity, trading, portfolio efficiency, and fees more broadly rather than conducting a randomized comparison of SDBA users against otherwise identical nonusers.

That limitation is important because the case against unrestricted brokerage access should not depend on an empirical claim that has not been established. The better argument is narrower and stronger: SDBAs expose participants to a larger set of risks and behavioral errors that retirement-plan research has already documented, while the benefits accrue most clearly to a relatively small group of sophisticated participants.

The Proper Question Is Not Whether an Expert Can Use an SDBA Successfully

An expert plainly can use an SDBA successfully. The relevant question for retirement-plan design is whether expanding unrestricted investment choice improves expected outcomes across the participant population for whom the plan exists. That inquiry is materially different from asking whether a sophisticated investor can make intelligent use of additional securities.

Suppose that only a small percentage of participants are sufficiently sophisticated to benefit from several thousand additional investment alternatives while the remaining participants already possess an efficient, diversified core menu. The existence of that sophisticated minority does not establish that unlimited choice represents sound retirement-plan architecture for the population as a whole.

The Department of Labor's ERISA Advisory Council has itself stopped short of concluding that brokerage windows are inherently harmful. After examining the issue, the Council did not recommend sweeping new regulation of brokerage windows generally, although it recommended further examination of plans relying exclusively on brokerage windows. That position is sensible because the evidence does not support prohibition, but it does support careful skepticism regarding the assumption that expanded choice is inherently beneficial.

Freedom Is Not the Same as Professional Design

The deeper issue is philosophical. A retirement plan is not merely a brokerage account with favorable tax treatment. It is an institution designed to transform current earnings into future financial independence. The employer and its advisers therefore face a design question: should the plan maximize the participant's range of possible choices, or should it maximize the probability of an adequate retirement outcome?

Those goals can conflict. Professional judgment often consists partly in reducing complexity by screening alternatives, eliminating inferior choices, creating rational defaults, and structuring the decision environment so that reasonable behavior becomes easier. That principle is familiar across professions and applies with particular force to retirement planning, where errors may not become visible for twenty or thirty years.

A carefully designed 401(k) menu can provide genuine autonomy without abandoning professional judgment. A participant can choose among diversified equity, fixed-income, target-date, balanced, and specialized alternatives without being asked to evaluate several thousand securities. Such a structure preserves meaningful choice while recognizing that the participant's objective is not to exercise investment freedom for its own sake, but to accumulate sufficient assets to finance retirement.

This is where Rational Paternalism provides the better framework. The purpose is not to eliminate choice but to structure it around the participant's stated long-term objective. That distinction matters because retirement-plan participants may discover the consequences of investment errors only after decades have passed, when the ability to repair the damage has diminished substantially.

The Brokerage Window Should Be an Exception, Not the Architecture

There are legitimate uses for SDBAs. Sophisticated investors may require exposure unavailable through the core menu. A participant working with a qualified investment adviser may use the window as part of a broader household portfolio. Plans may reasonably accommodate such participants without forcing specialized investments into the menu available to everyone.

That argument supports the SDBA as a limited escape valve rather than as the organizing principle of retirement-plan design. The research repeatedly points toward the advantages of curated menus, low-cost investment alternatives, rational defaults, diversified portfolio construction, periodic rebalancing, and mechanisms that reduce the impact of behavioral error.

Employers are generally capable of building efficient investment menus. Participants frequently undermine those menus through inefficient portfolio construction. Greater numbers of choices can create decision complexity. High-turnover trading can reduce investment efficiency. Additional investment universes introduce additional costs and additional opportunities for concentration, speculation, and poor selection.

The most important advances in modern 401(k) design have therefore not come from the creation of infinite choice. They have come from better defaults, automatic enrollment, automatic escalation, target-date funds, professionally managed allocations, institutional pricing, and diversified investment menus. These mechanisms are designed to make rational long-term behavior easier.

The proper role of the SDBA is therefore not to replace the architecture of a well-designed retirement plan but to supplement it selectively for participants who have both a legitimate need for broader investment access and the competence, professional advice, or investment discipline necessary to use that access intelligently. The relevant question is not whether participants should be permitted to choose anything. It is whether giving them access to everything materially improves the probability that they will achieve the purpose for which the 401(k) exists: financial security in retirement.

The third-party research and academic studies cited herein are believed to be reliable, but their accuracy and completeness cannot be guaranteed. Views expressed reflect the author's analysis as of the date published.