Morgan Stanley and E*Trade: wealth scale is a client-data problem

The week

On February 20, 2020, Morgan Stanley announced an agreement to acquire the online brokerage E*Trade in an all-stock transaction valued at about $13 billion. “All-stock” means Morgan Stanley paid with its own shares rather than cash, so E*Trade’s owners would become Morgan Stanley shareholders.

The terms were set out in Morgan Stanley’s announcement. E*Trade shareholders would receive 1.0432 Morgan Stanley shares for each E*Trade share, worth $58.74 at Morgan Stanley’s closing price the day before. E*Trade brought more than 5.2 million client accounts holding over $360 billion of retail client assets, plus about $56 billion of low-cost bank deposits.

Morgan Stanley described its own wealth business as about 3 million client relationships with $2.7 trillion of client assets. Combined, the firm counted $3.1 trillion of client assets, 8.2 million retail client relationships and accounts, and about 4.6 million stock plan participants. The release also pointed to roughly $7.3 trillion of assets that those customers held at other firms. That last figure was the prize, and capturing any of it depends on knowing which clients hold it.

The two firms served investors in very different ways. Morgan Stanley is best known for its adviser-led wealth business: clients, typically well off, who pay a professional to help manage their money. E*Trade was built for self-directed investors, people who open an account online and place their own trades with no adviser involved.

The deal did not arrive out of nowhere. It followed Charles Schwab’s agreement in November 2019 to acquire TD Ameritrade, another large online broker. And it came a few months after the major online brokers cut commissions on online US stock trades to zero in October 2019. A commission is the fee a broker charges each time a customer buys or sells a stock. When that fee fell to zero, one of the main ways online brokers earned money disappeared overnight, and the pressure to find scale and other sources of revenue grew sharply.

Morgan Stanley presented the combination as joining its adviser-led wealth business with three things E*Trade brought: its self-directed brokerage, its stock plan administration and its digital banking. Stock plan administration is the less familiar of these. Many companies pay employees partly in shares or share options, and someone has to run the accounts where those shares sit. That someone often becomes the first financial firm an employee deals with, years before they think about hiring an adviser.

Morgan Stanley was already in that business. In May 2019 it had completed its purchase of Solium, a Canadian stock plan administration firm, for about C$1.1 billion, and rebranded it Shareworks by Morgan Stanley. The E*Trade announcement said E*Trade’s US stock plan business would be combined with Shareworks. The firm would therefore be running several stock plan platforms at once, each with its own participant records and its own login.

The financial case leaned on integration. Morgan Stanley expected about $400 million in cost savings, which it attributed to technology infrastructure, shared corporate services and combining the two firms’ bank entities. It also expected about $150 million of funding benefits from putting E*Trade’s deposits to work. E*Trade’s chief executive, Mike Pizzi, was to keep running the business and lead the integration, reporting to Morgan Stanley’s chief executive, James Gorman.

The strategic picture was easy to draw on a slide: meet people early through their employer or a do-it-yourself account, then serve them with advice as their wealth and needs grow. Less easy to draw was whether the systems underneath could actually be joined so that a client felt like one person to the firm rather than three.

The timing added a complication nobody could have planned for. The announcement came on a Thursday. The following week, US stock markets had their worst week since 2008 as the coronavirus spread, and within a month most of both firms’ staff were working from home. The deal was priced in one market and integrated in another.

The transaction still needed regulatory approvals and a vote of E*Trade’s shareholders, and Morgan Stanley expected it to close in the fourth quarter of 2020. Schwab’s purchase of TD Ameritrade, valued at about $26 billion when announced, was working through its own approvals on a similar timetable.

The strategic logic of the deal was widely discussed: combine a large adviser-led wealth business with a self-directed platform and a workplace stock plan business, and serve clients across their financial lives. Less attention went to the condition the logic depends on, which is that the platforms can actually be joined.

For wealth managers and broker-dealers of any size, the deal made three architectural layers visible. “Architecture” here means how a firm’s systems are arranged and connected. The three layers are the client layer, the account and money movement layer, and the channel layer through which new clients arrive. Each is a separate integration problem, and they must be tackled in that order.

That ordering claim runs against how most growth programs are funded, so it needs a defense. Channels are visible, can launch in a quarter and produce sign-up numbers a board can see. Client data work takes years and produces mostly negative evidence: fewer duplicates, fewer reconciliation breaks, fewer complaints. The sections below take each layer in turn, show how the problem differs between kinds of firm, and test the sequence against the strongest case for doing it the other way round.

A house with three floors

A wealth firm’s systems can be pictured as a house with three floors.

The client layer is the top floor: who the client is. Name, address, family, risk appetite, goals, the people they have authorized to speak for them. It is the firm’s memory of the person.

The account and money movement layer is the ground floor: where the client’s money and investments actually sit and how they move. This includes custody, which means safekeeping the client’s securities, and clearing, which is the back-office process that settles a trade so that the shares and the cash change hands correctly.

The channel layer is the front door and the hallways: the ways clients arrive and move around, such as a mobile app, a website, an adviser’s office or an employer’s stock plan.

A merger can put a beautiful new front door on the house. If the floors above and below are not connected, clients walk in and find that nobody upstairs knows who they are.

Why the floors depend on each other

The floors are not independent. Account opening needs a verified client before it can create an account. Money movement needs each account to know its owner, its authorized signers and its restrictions before a dollar moves. Every channel, from an app screen to an adviser dashboard, is a view onto client and account data held somewhere else.

The dependency also runs one way. A channel can be replaced without touching custody. A custody conversion can proceed on a messy client layer, but every duplicate client record becomes a duplicate account mapping that someone has to check by hand. A clean client layer makes both of the other jobs cheaper, while a clean channel makes neither of them cheaper. That asymmetry is the whole case for the sequence.

The client layer: knowing it is the same person

A client who holds a self-directed account, an advised account and employer stock plan shares often exists as three records in three systems. Each system was built for its own business, with its own idea of what a “customer” is.

Moving that client from one channel to another, or offering advice when their circumstances change, requires recognizing that the three records are one person, with consent to use the data across contexts. That is a data and identity problem before it is a product problem.

Much of the record is already required

Broker-dealers already collect much of the client layer because rules require it. The customer identification program rule under the Bank Secrecy Act requires a broker-dealer to verify each customer’s identity when an account is opened. FINRA Rule 2090, the know-your-customer rule, requires reasonable diligence to know the essential facts about every customer.

FINRA Rule 4512 sets out the account information firms must keep. Since 2018 it has also required a reasonable effort to obtain the name of a trusted contact person for non-institutional accounts. For companies and similar legal entity customers, FinCEN’s customer due diligence rule, in force since May 2018, adds the identification of beneficial owners.

Regulation Best Interest took effect for broker-dealers on June 30, 2020, a few months before the deal closed. A recommendation to a retail customer must be in that customer’s best interest, based on a reasonably developed investment profile. Registered investment advisers owe a fiduciary duty under the Investment Advisers Act and need a similar picture of each client.

A firm with three systems therefore often holds the same regulated facts three times, collected under three procedures and possibly in conflict. When an examiner asks for a customer’s investment profile, the first question becomes which version counts as the record. A client master with clear ownership rules answers that once, for every channel.

The everyday version for mid-sized firms

For mid-sized firms, the equivalent challenge is joining up a custodian’s data, a CRM, a financial planning tool and a client portal. A CRM, or customer relationship management system, is the software where advisers and staff keep notes, contacts and tasks for each client. A custodian is the outside firm that holds the client’s assets. Each tool tends to hold its own copy of basic client facts.

Firms that do this well maintain a client master record and clear rules about which system owns which attribute: address and contact details, risk profile, investment objectives, beneficiaries, trusted contacts. When two systems disagree, the rule says which one wins and who fixes the other.

Why matching is harder than it looks

Matching is harder than it looks. Clients use different name forms, joint and entity accounts complicate ownership, and households need their own structure for planning and fee billing.

Take a client who appears in the self-directed system under her first initial and surname, and in the stock plan system under the full double-barreled name on her payroll. In the advised system she appears as a joint owner with her husband and as a trustee of a family trust. Whether that is one client, two or four depends on the purpose: she is one person for identity and consent, part of a household for planning and fees, and a representative of a separate legal entity for the trust. A good client layer can hold all of those relationships at once.

How matching works, and how it fails

Matching usually runs in two passes. Deterministic matching links records that share a strong identifier, such as a taxpayer identification number together with a date of birth. Probabilistic matching scores weaker evidence, such as similar names, a shared address or a common phone number, and sends borderline pairs to a person for review.

Each pass has a characteristic failure. Deterministic rules miss clients whose identifiers were keyed differently or never collected, which is common in older records. Probabilistic rules can merge two different people, such as a parent and an adult child with the same name at the same address.

A false merge is worse than a missed match, because it can show one client another client’s holdings. Careful programs set the automatic-merge threshold conservatively, log every merge with the evidence behind it and keep a tested way to split a merged record.

The review queue is where the real cost lies. Someone with access to sensitive data has to work through it, and the queue grows with every acquisition. Programs that budget for that team, rather than for matching software alone, are the ones that finish.

A joined record is also a bigger target

Joining records has a security cost that growth plans rarely price. A single client profile that combines identity documents, holdings, bank details and stock plan data is more valuable to an attacker than any of its parts. It is also more damaging to lose.

Regulation for that risk has tightened. In May 2024 the SEC amended Regulation S-P, which covers broker-dealers, investment advisers, investment companies and transfer agents. Covered firms must now run a written incident response program and notify affected individuals no later than 30 days after becoming aware of unauthorized access to sensitive customer information. Larger firms had to comply by December 2025 and smaller ones by June 2026.

The practical consequence is that the client master needs stronger access controls than the systems it replaced. Role-based access, logging of who viewed which profile, and fast revocation when staff move roles all become part of the integration plan rather than afterthoughts.

Two numbers that show progress

A firm can measure progress with two numbers: the share of accounts linked to a single client record, and the number of client attributes maintained in more than one system. The first should rise over time, the second should fall.

Firms that skip this work end up reconciling spreadsheets before every review meeting, and advisers see different versions of the same client depending on which screen they open. Besides being inefficient, that inconsistency makes clients doubt whether the firm really knows them.

Households drive fees as well as planning

Householding has a direct revenue consequence. Many advisory fee schedules apply breakpoints based on the combined assets of a household. If a household’s accounts are not linked, the client may be charged a higher rate than the firm’s own schedule promises. Mutual fund breakpoints, the sales charge discounts available on larger purchases, raise the same issue for brokerage accounts, and FINRA has pursued firms that failed to apply them.

Household definitions should therefore be written down, applied consistently and reviewed whenever accounts are merged or split. A billing error on a household is usually discovered by the client, and refunds with interest are the cheaper part of the cost.

How the client layer differs by type of firm

An independent registered investment adviser typically relies on one or more custodians for account records and owns the client relationship in its CRM. Its client layer is mostly a CRM discipline, with the custodian’s data feed as the source for positions and transactions.

A dual registrant, acting as both broker-dealer and investment adviser, has a harder job. The same client may hold brokerage and advisory accounts governed by different standards of conduct. The record has to show which capacity applied to each recommendation, and the client has to be able to tell the difference.

A bank trust department adds fiduciary accounts governed by trust documents and state law, where the client may be a trust with several beneficiaries and a co-trustee. A credit union that offers investments often does so through a third-party broker-dealer. The brokerage account then lives on the broker-dealer’s books while the member relationship lives on the credit union’s core system. Linking the two depends on contracts and privacy notices as much as on technology.

The account and money movement layer: where errors are visible

Beneath the client experience sit custody, clearing, cash sweep programs, transfers between accounts and the statements and tax reporting clients expect. A cash sweep program automatically moves uninvested cash in a brokerage account into a bank deposit or money market fund, where it may earn interest.

Combining these after an acquisition is a long program, because every account conversion touches books and records, cost basis, tax reporting and client communications. Cost basis is what the client originally paid for an investment, which determines how much tax they owe when they sell. Get it wrong and the client’s tax form is wrong.

Errors here are visible to clients immediately and can raise recordkeeping questions for broker-dealers under SEC rules. A client may forgive a clunky app. They are much less forgiving of a statement that shows the wrong balance.

How accounts move between firms

When a client moves an account from another firm, most US brokerage transfers run through ACATS, the Automated Customer Account Transfer Service operated by the National Securities Clearing Corporation. FINRA Rule 11870 sets the timetable. The firm giving up the account must validate the instruction or take exception within one business day, then complete the transfer within three business days of validation.

The fast path depends on clean data. Transfers are rejected for mismatched names, account titles or taxpayer identification numbers. Each rejection restarts the clock and usually means a phone call to a client who had expected the move to be simple.

Cost basis travels separately. Under IRS broker reporting rules for covered securities, the transferring broker must send a transfer statement with basis information, generally within 15 days of the transfer. If that information arrives late or incomplete, the receiving firm’s year-end tax forms will be wrong for those positions. Inside a merger the same discipline applies at far larger scale, with no outside deadline to enforce it.

Shorter settlement leaves less room for error

In May 2024 standard US securities settlement moved from two business days after the trade to one. A shorter cycle leaves less time to fix a mis-mapped account, a missing standing instruction or a wrong allocation before money and shares must change hands. Conversion plans written under the old cycle had a little slack in them. Under T+1, a bad instruction loaded on a conversion weekend can produce a failed settlement within a business day.

What a conversion really involves

Smaller firms face a version of this when they change custodian or clearing firm. Conversion planning, data mapping and reconciliation are the critical path, and they should be staffed and tested as such rather than left to the vendor.

Data mapping means working out, field by field, how information in the old system translates into the new one. Reconciliation means checking afterwards that every position, every share and every dollar arrived where it should.

Mock conversions with real data, reconciliation of positions and cost basis to the penny, and a rollback plan separate an uneventful weekend from months of client service problems. A rollback plan is the agreed way to undo the move and return to the old system if something goes badly wrong.

The evidence a conversion should leave behind

Broker-dealers must make and keep books and records under SEC Rules 17a-3 and 17a-4, and a conversion changes where those records live. Examiners and internal auditors who review a conversion afterwards tend to ask for the same artifacts:

• the data mapping specification, signed off by the business owner of each field;
• the results of each mock conversion, with the breaks found and how each was resolved;
• the go or no-go record for the live weekend, with the criteria used;
• the post-conversion reconciliation of positions, cash and cost basis, and how residual breaks were cleared;
• the client communications sent, and complaint volumes before and after;
• confirmation that historical records stay retrievable for their required retention periods after the old system is switched off.

The last item is the one most often forgotten. Decommissioning an old platform without a plan for its records can leave a firm unable to produce a statement it is required to keep. Archiving those records in a searchable, compliant form is a project of its own, with its own budget.

Cash deserves its own plan

Cash is a special case. Sweep programs, margin, and transfers between brokerage and banking products affect client income and regulatory disclosures. Margin is borrowing against investments, and it carries its own rules and risks.

Any change to where client cash sits needs its own communication plan and a check that disclosures still match practice. If clients were told their cash would sit in one place and it is now held in another, the gap between the paperwork and reality is a regulatory risk as well as a trust problem.

The economics explain why cash draws scrutiny. Sweep programs that place client cash in an affiliated bank turn those balances into deposits that fund the bank. Morgan Stanley’s own announcement counted E*Trade’s deposits, and the funding benefit they would bring, among the deal’s attractions.

When client cash is that valuable to the firm, the rate paid to clients and the way it is set become a conflict-of-interest question. For an adviser it touches fiduciary duty and disclosure. For a broker it touches best interest. Any integration that moves cash between programs should document the rationale, the rate comparison and the disclosure changes before the move, not after.

The channel layer: how new clients arrive

E*Trade’s stock plan administration business gave Morgan Stanley a route to employees of corporate clients, many of whom would later need advice. That is a distribution architecture: a low-cost digital entry point connected to a higher-touch service.

Building that connection requires the client and account layers to work, plus a way to hand a client from one experience to another without making them start again.

Workplace data comes with an employer attached

A stock plan business has two clients. The employer contracts for plan administration, and the employee is a participant. Administration agreements and employer expectations often limit how participant data can be used, and employers are wary of their staff being marketed to. A firm that wants to offer advice to participants has to work within those terms, typically through opt-in education, employer-approved financial wellness programs and a clear line between administration and solicitation.

The data itself is unusual. It includes grant schedules, vesting dates and, for executives, trading windows and insider status. That information is valuable for advice, for example in planning sales around vesting dates and tax. Some of it is also commercially sensitive to the employer. Access controls have to show an adviser what the participant has authorized, and nothing about the employer that the adviser should not see.

Choices open to independent firms

Independent wealth managers cannot buy a workplace platform. They can choose their digital entry points, such as a self-service onboarding flow, a planning tool or a low-minimum managed portfolio. Each should connect cleanly to advisers instead of sitting as a separate product with separate data.

Consent shapes what is possible

Consent and privacy shape what is possible. Data gathered for a workplace plan or a self-directed account may carry notices and permissions that limit how it can be used for advice or marketing.

Firms that design consent capture into each entry point, and record it against the client master record, can move clients between channels without re-papering them. Firms that do not may find their most promising leads cannot lawfully be contacted.

Affiliate sharing has its own rules

After an acquisition, the channels are usually housed in different legal entities, and privacy rules follow the entity lines. For SEC-registered broker-dealers and investment advisers, Regulation S-P governs privacy notices and opt-outs for sharing with nonaffiliated parties. Regulation S-AM limits using certain information received from an affiliate to make marketing solicitations, unless the consumer has had notice and a chance to opt out. Banks have parallel rules under their own regulators.

Whether a self-directed client can be offered advice can therefore depend on which entity collected the data, what notice the client received and whether they opted out. Counsel should confirm how these rules apply to a given structure. The technical consequence is clearer: notice and opt-out status belong on the client master, recorded by entity and by date.

The handoff is where clients are lost

The handoff is where most firms lose clients. A prospect who completes a digital questionnaire and is then asked to fill in the same information on paper for an adviser has been told, in effect, that the digital channel did not count.

The test is whether information captured once follows the client everywhere: into the CRM, the planning tool, the account opening workflow and the suitability record. The suitability record documents why a recommendation fits a client’s circumstances, and regulators expect firms to keep it. That test is cheap to run and shows exactly where the client layer is broken.

The handoff carries regulatory weight

Moving a client from self-directed to advised is often a recommendation in its own right. Regulation Best Interest covers a broker-dealer’s recommendation of an account type, and Form CRS, the short relationship summary given to retail investors, has to describe the services and fees of each. A handoff that captures the client’s profile once and carries it forward keeps that documentation consistent. A handoff that starts again produces two profiles that may disagree, and an examiner will ask which one supported the recommendation.

One firm, done in the right order

To make the sequence concrete, take an independent wealth manager with a few dozen advisers, two custodians inherited from past acquisitions, a CRM, a planning tool and a client portal. Its leadership wants to launch a low-minimum digital portfolio to attract younger clients.

Step one: the client layer. Before building anything new, the firm counts. It finds that far fewer of its accounts are linked to a single client record than anyone assumed, and that client addresses are kept in four places. It names the CRM as the master for contact details and the planning tool as the master for goals and risk profile, and assigns an operations manager to fix disagreements each week. Nothing about this is glamorous. It mostly involves careful matching, phone calls to clients and a great deal of patience.

Step two: fewer moving parts. With clients properly identified, the firm consolidates onto one custodian. It runs two rehearsal conversions with copies of real data, reconciles every position and cost basis figure, writes a plan for reversing the move, and prepares letters explaining any change to where client cash will sit. The live conversion happens over a weekend, and on Monday the service desk has little to do, which is exactly the outcome a good conversion aims for.

Step three: the new channel. Only now does the firm launch its digital portfolio. Because consent is captured at sign-up and stored against the client master record, and because the onboarding questionnaire feeds the CRM and planning tool directly, an adviser can call a digital client who has just inherited money and pick up the conversation without starting over.

Run the same plan in the opposite order and the picture changes. The new digital product launches first, creating a fifth place where addresses live. Prospects sign up, but nobody is sure which of them already have advised accounts. Advisers receive leads with no consent to contact them for advice. The conversion, when it finally happens, has to untangle all of this at once. The firm has added a channel and multiplied its reconciliation work.

The case for launching the channel first

The strongest counter-argument deserves a fair hearing. Markets do not wait. After commissions fell to zero, a firm that spent two years on client data while competitors launched apps risked losing a generation of new investors. A new channel can also fund the cleanup, and a launch date can force decisions that committees would otherwise defer.

There are cases where channel-first is right: a firm with almost no digital presence, a narrow product that needs few client attributes, or a pilot deliberately kept apart from the main client base. The argument weakens when the new channel is expected to feed advisers, because that is exactly the flow that breaks on bad client data.

A middle path works for many firms. Build the new channel on the target client model from its first day, even before the rest of the firm has migrated, and make it the first system that follows the new ownership rules. That keeps the speed without adding another copy of the address book.

What the sequence costs

Step one is mostly people: a named owner for each client attribute, analysts to work the match review queue, and operations staff to contact clients. Step two is project cost, much of it in testing and in running two platforms side by side for a period. Step three is the visible investment. Firms that budget only for the third step tend to discover the first two later, as overruns. Naming all three costs at the outset gives the board an honest view of the trade-off.

What separates those two paths is sequence, not budget or talent, and it is the most transferable insight from the large deals of 2019 and 2020: the attractive part of the strategy sits on top of the least attractive work, and the order cannot be skipped.

What the board should see in the plumbing

This is more than plumbing for the technology team: each layer connects directly to something a board cares about.

• Revenue. The whole case for joining self-directed, workplace and advised channels rests on moving clients between them as their needs change. If the firm cannot recognize a client across channels, that movement does not happen and the promised growth stays on the slide.
• Cost. Every attribute held in more than one system has to be kept in step by someone. That work grows with every new product and every acquisition.
• Client experience. Clients judge a firm on whether it remembers them. Being asked the same questions repeatedly signals the opposite.
• Regulatory risk. Wrong statements, wrong tax forms, outdated disclosures and uncontacted leads that should not have been contacted are all compliance problems as well as service problems.

A board pack can carry the layers in a handful of indicators. The client layer shows in the share of accounts linked to one client, the number of attributes held in more than one system and the size of the match review queue. The account layer shows in open reconciliation breaks by age and in conversion defects still unresolved. The channel layer shows in the share of digital sign-ups that reached an adviser with consent intact. These indicators need no new systems to produce, only an owner who reports them every quarter.

The competitive pressure from these mergers was real, but smaller firms did not need to imitate them. A sounder sequence was to fix the client layer, then reduce the number of custodians and integration points, and only then add channels. Firms that bolt new client experiences onto fragmented client data multiply their reconciliation work rather than their growth.

Signs a firm has its layers in order

A firm in good shape tends to share a few features:

• One client master record, with written rules about which system owns each client attribute.
• A rising share of accounts linked to a single client, reported regularly to leadership.
• Conversions treated as major projects, with rehearsals using real data and a tested way back.
• Consent captured at every entry point and stored where every channel can see it.
• A digital entry point that hands clients to advisers with their information intact.

The ways firms go wrong are just as consistent:

• Starting with the shiny channel. Launching a new app or onboarding flow before the client layer can support it.
• Leaving conversions to the vendor. Assuming the custodian will handle data mapping and reconciliation, and discovering problems when clients do.
• Treating cash as an afterthought. Moving sweep arrangements without updating disclosures and telling clients.
• Collecting data without permission to use it. Building a pipeline of prospects that cannot lawfully be contacted for advice.
• Measuring activity rather than integration. Counting new sign-ups while the number of duplicate client records keeps growing.

Signs that accounts are linked to one client record

1. If a client has three kinds of account with us, do we know it is the same person, and how?
2. What share of our accounts are linked to a single client record today, and what was it a year ago?
3. For a client’s address, risk profile and beneficiaries, which system is the master, and who fixes the others when they disagree?
4. Do we rehearse conversions with real data, and do we have a tested plan to roll back?
5. When a prospect moves from a digital questionnaire to an adviser, how much do they have to tell us again?

Big mergers make headlines because of the price. The value, for firms of every size, depends on something far less visible: whether the firm can recognize one client as one person, wherever that client chooses to walk in.

With hindsight

The deal closed on October 2, 2020, after Federal Reserve approval, and Morgan Stanley did not stop there. Days later it agreed to acquire the asset manager Eaton Vance, a deal that closed in March 2021 and added another business and another set of systems to join.

The E*Trade brand survived as “E*TRADE from Morgan Stanley”, but the account layer moved underneath it. Client notices in 2023 said Morgan Stanley Smith Barney would take over custody and clearing from E*Trade Securities. Accounts migrated in waves from March, and the final group moved around the start of September 2023. That is roughly three years from closing to having the self-directed accounts on the acquirer’s books.

The workplace business followed the same slow path. In December 2023, Morgan Stanley at Work said redesigned stock plan participant accounts would carry its name as their primary brand. It described the change as a step toward a single user account structure. Its participant accounts had previously gone by the names Shareworks, StockPlan Connect or E*Trade. Several participant platforms converging on one account structure is exactly the client-layer work this article describes, and it was still under way nearly four years after the announcement.

Schwab completed its TD Ameritrade acquisition the same month and then spent several years converting client accounts onto its own platform in waves. The final group converted in May 2024. Moving clients in batches, rather than all at once, is a sensible way to limit risk, but the length of the program stands out. It is a reminder of how long the account and money movement layer takes to integrate at scale, even for a firm with deep resources and every incentive to move quickly.

The new owners also faced an early stress test. The surge in retail trading in 2020 and early 2021 tested self-directed platforms’ capacity and customer service within months of the deals.

The ambition behind the deal grew rather than shrank. In April 2021 Gorman set a target of $10 trillion of client assets across Morgan Stanley’s wealth and investment management businesses, and his successor, Ted Pick, kept it. A target of that kind depends on moving clients between channels, which is the flow that depends on the client layer.

Cash turned out to be the most exposed part of the account layer. Morgan Stanley disclosed that from April 2024 it had been responding to requests from the SEC’s Enforcement Division about advisory account cash swept to affiliate bank deposit programs. Around the same time it raised the rates paid on advisory sweep cash. In March 2025 the SEC told the firm it had concluded the investigation and did not intend to recommend enforcement action.

Others did not escape as lightly. Wells Fargo and Merrill Lynch each settled SEC charges over cash sweep practices, and class actions filed in 2024 alleged that Morgan Stanley or its E*Trade unit paid unreasonable interest on sweep cash. Where client cash is held, what it earns and how that is disclosed proved to be a question about client trust as much as about funding.

The client layer turned out to carry compliance weight as well as growth potential. In April 2024, the Wall Street Journal reported that several federal regulators were examining how Morgan Stanley’s wealth business verified client identities and screened for money-laundering risk. FINRA, the industry’s self-regulatory body for brokers, was reported in 2025 to be reviewing client vetting too. The firm has said it invested heavily in its anti-money-laundering and client-vetting programs.

We are not aware of published findings from those reviews, and nothing here should be read as a conclusion about them. But the episode shows something important about a client record that joins many channels. Besides serving as a sales tool for moving clients from self-directed to advised accounts, it must hold the due diligence that follows that client into each one: who they are, how their identity was checked, and what was learned about the source of their money.

For firms of any size, a single view of the client serves two masters at once. It makes growth across channels possible, and it is where regulators will look to see whether the firm truly knows who it is serving. Building it well pays off on both counts.