By Chris Tobe, CFA, CAIA

Four years ago I warned that crypto was exposing a major weakness in 401(k) regulation: the self-directed brokerage window. https://commonsense401kproject.com/2022/06/18/brokerage-windows-exposed-by-crypto/

Now a new federal court ruling involving AT&T makes that warning even more important.

The issue is bigger than Bitcoin.

If courts and regulators treat brokerage windows too loosely, they could become the back door through which crypto, private equity, private credit and other high-fee alternative investments enter ERISA plans—with substantially less fiduciary scrutiny than they would receive if they appeared directly on the plan’s investment menu.

And there is another issue that should concern every 401(k) fiduciary:

Who is paying the brokerage-window provider to make these investments available?

That question goes directly to conflicts of interest and potentially to ERISA’s prohibited-transaction rules.

The AT&T Decision Should Not Become a Brokerage-Window Safe Harbor

In Alas v. AT&T, a federal district court recently reconsidered an earlier ruling concerning disclosures of indirect compensation received by Fidelity in connection with AT&T’s brokerage window.  https://www.aol.com/articles/judge-just-changed-major-401-110500000.html

The ERISA Industry Committee celebrated the decision, explaining that Fidelity may receive payments from mutual funds available through the brokerage window and that the dispute concerned whether disclosure of that compensation was sufficient for fiduciaries to determine whether the arrangement was reasonable.

That may sound like a technical disclosure dispute.

It isn’t.

The underlying structure is exactly what fiduciaries should be examining:

401(k) money → brokerage window → investment provider → payments to recordkeeper/platform.

Once money starts moving in that circle, ERISA fiduciaries should be asking much more than whether a compensation range was adequately disclosed.

They should be asking:

Who pays Fidelity or another brokerage provider?

How much?

What are they paying for?

Does compensation affect which investments get access to the platform?

Are there revenue-sharing, shelf-space, servicing, data, custody, trading or other payments?

Are affiliates receiving compensation?

And most importantly:

Does any of this constitute a prohibited transaction involving plan assets and a party in interest?

ERISA’s prohibited-transaction rules do not disappear because an investment is located behind a brokerage-window button.

Crypto Has Exposed the Problem

I first wrote about this in 2022 in “Brokerage Windows Exposed by Crypto.”

The basic problem hasn’t changed.

A traditional 401(k) menu might contain 10, 15 or 20 investments. Fiduciaries select those investments, monitor them, evaluate their fees and hopefully understand who is getting paid.

Then the same fiduciary opens a brokerage window containing hundreds or thousands of investments.

Suddenly everyone starts pretending that fiduciary responsibility somehow became radically different.

Why?

The participant clicked the mouse.

That strikes me as an extraordinarily weak foundation upon which to build ERISA policy.

The plan fiduciary still selected the brokerage-window provider.

The plan fiduciary still approved the arrangement.

The plan fiduciary still negotiated—or failed to negotiate—the provider’s compensation.

And the brokerage-window provider can still make money from transactions involving retirement assets.

Participant choice should not become a magic ERISA eraser.

Professor Hilary Allen Identified Who Really Needs Whom

My August article, “Professor Hilary Allen Is Right: Crypto Has No Place in a 401(k),” addressed an even more fundamental problem.  https://commonsense401kproject.com/2026/08/11/professor-hilary-allen-is-right-crypto-has-no-place-in-a-401k/

Professor Allen turns Wall Street’s “democratization” argument upside down.

The question isn’t whether 401(k) participants desperately need access to crypto.

It is whether crypto desperately needs access to 401(k) participants.

That is an enormously important distinction.

America’s defined-contribution retirement system contains trillions of dollars of assets plus something extraordinarily valuable to an asset manager: a continuing stream of payroll contributions.

Money comes in every two weeks.

Month after month.

Year after year.

For decades.

For an industry needing new buyers, 401(k)s aren’t simply another distribution channel.

They may be the ultimate source of permanent demand.

That is why Professor Allen’s “bagholder” argument is so important. Large existing crypto holders eventually need someone willing to buy from them.

Putting crypto into retirement accounts potentially creates millions of new buyers contributing automatically.

The fiduciary question therefore shouldn’t be:

How can we give workers access to crypto?

It should be:

Why does the crypto industry want access to workers’ retirement money?

And who gets paid when that access is provided?

PwC Accidentally Explained Why Crypto Doesn’t Belong Here

My June article, “Crypto in 401(k)s: PwC Accidentally Says the Quiet Part Out Loud Again,” approached the problem from another direction.  https://commonsense401kproject.com/2026/06/10/crypto-in-401ks-pwc-accidentally-says-the-quiet-part-out-loud-again/

PwC’s own global crypto-regulation analysis describes an extraordinary infrastructure necessary to make crypto markets function safely:

special custody regimes;

special liquidity requirements;

special disclosure standards;

special operational-resilience systems;

special market-conduct rules;

special stablecoin reserve requirements;

special governance structures;

special supervisory regimes;

and cross-border regulatory coordination.

Traditional diversified stock and bond funds don’t require an entirely new global financial-regulatory architecture just to make them suitable investments.

Crypto does.

PwC also acknowledges continuing problems involving regulatory arbitrage, fragmented supervision and inconsistent implementation across jurisdictions.

That doesn’t sound like an asset class crying out to become part of America’s retirement system.

It sounds like a warning label.

And it makes the brokerage-window issue even more important.

If an investment is too complicated, opaque and conflicted to survive normal fiduciary scrutiny, the solution shouldn’t be:

Put it in the brokerage window.

Then There Is ERISA §406

This is where I think the brokerage-window debate has been much too timid.

ERISA doesn’t merely impose a general prudence requirement.

It contains prohibited-transaction rules.

The IRS summarizes prohibited transactions as including the use of plan assets for the benefit of a disqualified person, fiduciary self-dealing, a fiduciary’s receipt of consideration in connection with transactions involving plan assets, and certain transactions involving services or facilities between a plan and a disqualified person. Exemptions exist, but they have conditions.

That matters enormously when a brokerage window involves:

Recordkeepers.
Broker-dealers.
Custodians.
Crypto exchanges.
Fund companies.
Asset managers.
Trust companies.
Affiliated investment products.
Revenue sharing.
Indirect compensation.

My 2025 article, “Crypto as a Prohibited Transaction in 401(k) Plans—Target Date and Brokerage Windows,” argued that this is where crypto creates an entirely different ERISA problem.  https://commonsense401kproject.com/2025/11/03/crypto-as-a-prohibited-transaction-in-401k-plans-target-date-and-brokerage-windows/

Imagine a plan recordkeeper or affiliate receives direct or indirect economic benefits because participants buy particular products through its brokerage window.

Calling the transaction “participant directed” doesn’t answer the prohibited-transaction question.

The fiduciary created the window.

The fiduciary selected the service provider.

The service provider is a party in interest.

And somebody is getting paid.

The question is whether the transaction falls within ERISA §406 and, if so, whether all the requirements of an applicable exemption are actually satisfied.

That analysis should happen before participants’ retirement money starts flowing—not after somebody files a lawsuit.

Cunningham v. Cornell Makes This More Important, Not Less

The Supreme Court’s unanimous Cunningham v. Cornell University decision makes the distinction particularly important.

The industry would understandably like fiduciaries and courts to jump immediately to whether a service-provider arrangement is reasonable.

But §406 and §408 are not the same thing.

A transaction can first fall within the prohibited-transaction provisions and then require an exemption.

That means the existence of a brokerage window shouldn’t end the inquiry.

It should begin it.

Who is the party in interest?

What transaction occurred?

What direct or indirect compensation was received?

Who received it?

Was it reasonable?

Was the arrangement properly disclosed?

What exemption is being relied upon?

Those questions become especially important when the investment providers themselves have enormous financial incentives to obtain access to retirement assets.

Private Equity Is Watching

This is why the crypto debate matters even to people who couldn’t care less about Bitcoin.

Crypto may merely be the test case.

Private equity and private credit managers have the same fundamental business problem:

They need assets.

And America’s 401(k) system contains trillions of dollars of them.

The sales pitch is remarkably similar.

First call the investment “alternative.”

Then call it “institutional.”

Then say wealthy people have access and ordinary workers unfairly don’t.

Call expanded distribution “democratization.”

Package complicated assets inside simpler-looking wrappers.

Put them inside target-date funds, CITs—or potentially brokerage windows.

And collect fees that become progressively harder for participants to identify.

The brokerage window could become the perfect delivery mechanism because it creates the appearance that the employer didn’t select the investment.

The participant did.

But that ignores the architecture that made the transaction possible.

Someone selected the window.

Someone selected the brokerage provider.

Someone negotiated the compensation arrangement.

Someone decided what investments could be sold through it.

And someone is making money.

ERISA fiduciaries should follow that money.

The Brokerage Window Could Become Wall Street’s ERISA Back Door

Imagine where this could lead.

A fiduciary committee might reject putting Bitcoin directly on the core menu.

Too risky.

Reject putting a private-equity fund directly on the menu.

Too illiquid.

Reject a private-credit fund.

Too difficult to value.

Then someone says:

Don’t worry. Just put them in the brokerage window. Participants can decide for themselves.

That could turn brokerage windows into an enormous regulatory arbitrage machine.

The investments carrying the greatest valuation problems, liquidity risks, fees and conflicts could end up receiving less fiduciary scrutiny precisely because they are more complicated.

That is backwards.

The more complicated an investment becomes, the more fiduciary scrutiny should be required.

The more illiquid it becomes, the more scrutiny should be required.

The more difficult its valuation becomes, the more scrutiny should be required.

And the more compensation flowing among service providers and investment managers, the more important ERISA’s prohibited-transaction protections become.

Crypto Is the Canary in the 401(k) Coal Mine

Crypto didn’t create the brokerage-window problem.

Crypto exposed it.

It exposed the fiction that moving an investment from a core menu into a brokerage window somehow transforms the fiduciary economics of the transaction.

It exposed the danger of participant choice being used as an excuse for weaker oversight.

And most importantly, it exposed the enormous commercial value of gaining access to America’s retirement savings infrastructure.

Today the product is crypto.

Tomorrow it could be private equity.

Private credit may be right behind it.

The industry’s argument will always sound attractive:

More choice. More access. More democratization.

ERISA asks a different set of questions:

Is it prudent?

Is it loyal?

Who is getting paid?

Is there a conflict?

Is there a prohibited transaction?

Those questions shouldn’t disappear when a participant enters a brokerage window.

They should become even more important.

CommonSense Bottom Line

A brokerage window should never become a prohibited-transaction window.

Professor Hilary Allen has correctly identified the economic danger of turning workers into the next generation of crypto buyers. PwC has inadvertently documented the extraordinary regulatory machinery required to support crypto. And the emerging brokerage-window litigation demonstrates how complicated the indirect-compensation relationships already are.

Put those three issues together and the warning is pretty simple:

Wall Street wants access to trillions of dollars of retirement savings.

Crypto is showing us how the plumbing could work.

Private equity and private credit could use the same plumbing on a much larger scale.

Before opening that door, every ERISA fiduciary should ask the oldest and most useful question in finance:

Who gets paid?

Then ask the question ERISA adds:

Are they allowed to get paid that way?