By Andrew Welsch
Charles Schwab spent a good chunk of its six-hour-long investor day on May 14 explaining to analysts and shareholders how the company is using artificial intelligence to boost its business. Investors, however, are far more focused on whether AI poses a threat to the substantial profits Schwab derives from so-called sweep cash, the money that clients hold in brokerage accounts that earn almost no interest. AI could change that equation by powering tools that allow investors to automatically move idle cash from their brokerage account to money-market funds or other higher-yielding accounts that are far less profitable for Schwab, as well as other brokerages that derive profits from customers' cash.
The AI threat more broadly -- that customers would migrate from traditional financial services to AI start-ups -- has been weighing on shares of Schwab, LPL Financial, Raymond James Financial, and Ameriprise Financial this year despite generally strong earnings growth. But the cash sweep issue is particularly nettlesome.
JPMorgan Chase, the nation's largest bank, kicked off the latest round of hand-wringing in April when it disclosed that it is developing an AI-powered tool that would help customers automatically move money from their checking or savings accounts to higher-yielding options. Schwab currently pays customers just 0.01% in interest on cash held in sweep accounts. Large money-market funds pay an average of 3.44%, according to Crane Data.
The announcement, which was light on details, sparked a selloff in shares of Schwab, Raymond James, and LPL. It's easy to imagine the popularity of a tool that allows customers to automatically move from a yield of 0.01% to 3.5%, and why investors are worried.
Sweep Stakes
Analysts estimate that profits from sweep accounts contribute to a sizable chunk of profits, although amounts vary, totaling from 40% to 100% of firmwide preprovision net revenue -- that is, earnings before setting aside funds for loan defaults or credit losses, according to Wolfe Research analyst Steven Chubak.
At Schwab, which operates a bank, sweep cash contributes to bank deposit account fees and net interest income, which is the difference between the interest paid to customers on their deposits and what the company earns on interest-bearing assets such as fixed-income securities and loans. Almost half of Schwab's $6.5 billion in total revenue for the first quarter came from net interest income. Put another way, that's more than 100% of Schwab's reported first-quarter net income of $2.5 billion.
Then there is LPL. It doesn't operate a bank; instead, it places customer deposits with partner banks that pay LPL a fee. Revenue generated from client cash -- including sweep accounts and money-market accounts -- came to $1.66 billion for 2025, according to the company's annual report. LPL generated nearly $17 billion in total revenue last year and $863 million in net income. Due to the importance of this source of income, the company's annual report warns that significant interest-rate changes could hurt the fees it earns from banks, which would eat into profits.
Details about JPMorgan's forthcoming cash optimization tool are hazy, and a spokesman declined to provide more information. But other firms could easily develop their own, creating competitive pressure to add one. During his investor day presentation, Schwab CEO Rick Wurster said the company doesn't plan to offer its own cash optimization tool, noting that clients can already move money from sweep accounts to a variety of higher-paying options such as money-market funds. The company also said strong lending activity this year can support net interest margin growth.
On recent first-quarter earnings calls, other CEOs have also played down the AI threat to cash profits. LPL CEO Rich Steinmeier said he didn't see "an imminent risk." Raymond James CEO Paul Shoukry said it wasn't much more than "an incremental threat," and more a concern for online brokerage firms than for companies like his, where the client has a relationship with a financial advisor.
Wealth management executives' recent reassurances don't seem to have resonated with investors, perhaps because AI is developing at a rapid pace and increasingly focusing on pocketbook issues. In February, financial-technology company Altruist introduced a tax-planning capability for its AI platform called Hazel, which sparked a selloff in wealth-management-related stocks. In May, OpenAI said it would enable customers to link their financial accounts to ChatGPT and receive personalized guidance on how to better manage their finances.
"Though Schwab doesn't see much risk here, we wonder what the company can really say at this point, given the longer tail and open-ended nature of some investors' fears here," wrote Truist Securities analyst David Smith in a May 15 research note.
New Models
Wolfe analyst Chubak wrote in an April 22 report that investor worries about cash sweep profits will probably persist as agentic AI tools improve over time. Chubak's proposed solution is for companies to change their business model, moving toward more fee-based revenue, such as charging clients or advisors a so-called platform fee based on total assets.
Chubak believes that if they do so, they can weather the AI storm. He calculates that a platform fee as low as 0.07%, or seven basis points, on average, would probably suffice. "This would subsidize lost revenue from offering higher payouts on customer cash, and improve the durability of earnings, supporting a higher target P/E [price/earnings] multiple," he wrote.
Chubak thinks LPL's stock is the likely biggest beneficiary. But he acknowledges that transitioning to the new model wouldn't be risk-free for brokerages.
Small Sums
Some analysts believe that investors are overreacting to potential AI threats to profits earned from client cash. Competitors such as Vanguard and Interactive Brokers already offer investors higher-yielding default options for idle cash. Yet "this does not seem to be a factor in client assets flows," wrote Goldman Sachs analyst Alex Blostein in a May 15 note.
Brokerage executives have also taken pains on earnings calls to point out that sweep account balances are currently small. "So, just let me emphasize again, our average balance is now $6,000," Ameriprise Financial Chief Financial Officer Walter Berman told analysts during his company's April 23 earnings call.
Smith, the Truist analyst, wrote on May 12 that Schwab's transactional sweep cash on a per account basis has already shrunk from about $17,000 in 2022 to $12,000 today. Clients earning an extra 250 basis points of yield on $12,000 of sweep balances would make an extra $300 annually, he wrote. "We question if this extra income would be worth it to most customers, particularly if it required turning over the keys to their finances to an AI agent to optimize the yield."
Schwab, like its peers, has been touting the ways it is deploying AI to improve productivity and drum up more business with customers. For example, the company recently unveiled an AI agent that provides individual investors with portfolio insights and research. It's just the start, as Schwab plans to launch more AI agents and tools this year. Its competitors are planning similar launches.
"We do not expect the AI/cash overhang to disappear immediately, but Schwab's message was that it can grow through the debate rather than structurally change the business model," wrote Citizens analyst Devin Ryan on May 15. He suggests that investors focus not just on the movement of client cash over time but also on how Schwab and other brokerages can innovate to offset AI pressures as they emerge.
Write to Andrew Welsch at andrew.welsch@barrons.com
This content was created by Barron's, which is operated by Dow Jones & Co. Barron's is published independently from Dow Jones Newswires and The Wall Street Journal.
(END) Dow Jones Newswires
May 23, 2026 04:00 ET (08:00 GMT)
Copyright (c) 2026 Dow Jones & Company, Inc.
Comments