Robinhood moved more than $6bn out of bank sweeps to fund margin lending

Robinhood moved more than $6 billion of customers’ uninvested cash out of its off-balance-sheet bank-sweep programme and into broker-held free credit balances in February 2026. The company’s own filing says the purpose was to fund growth in margin lending. To an enrolled customer, the balance could still look like cash earning the same stated rate inside the same app. Economically and legally, however, the location of the money changed: cash previously deposited at programme banks became a liability of the broker that Robinhood could use in its business, subject to securities rules.[1][2][3]

The shift was not a customer withdrawal and it did not mean Robinhood simply took ownership of customer money. A free credit balance remains payable to the customer on demand, and broker-dealers operate under reserve, possession-and-control, net-capital and customer-protection requirements. But it is not the same instrument as a deposit placed at an FDIC-insured programme bank. That distinction matters because the programme’s protection, balance-sheet treatment, liquidity function and revenue mechanism are different even when the customer-facing interest rate is unchanged.[1][2][3][4]

A financing decision embedded in a cash programme

Robinhood defines Cash Sweep as participating users’ uninvested brokerage and banking cash automatically moved to deposits at a network of banks. Because those deposits are for customers’ benefit at third-party banks, the company describes the amount as off balance sheet and earns a net spread between what the banks pay and what users receive. Its revised High-Yield Cash programme introduced a second route. Under the brokerage-held component, the first $10,000 of enrolled balances per eligible customer can remain at Robinhood as a free credit balance rather than being swept onward.[1][2]

The company did not leave the reason implicit. Its September-quarter filing says the February update was made “to fund growth in margin lending” and caused more than $6 billion to move from Cash Sweep to Cash and Deposits. Robinhood’s customer agreement says cash in the brokerage-held programme remains in the account awaiting investment, is payable on demand, and may be used by Robinhood in connection with its business, subject to applicable law. Read together, the documents establish a funding chain: a portion of customer cash became broker funding while the company’s margin-receivable book expanded.[1][3]

By September 30, Robinhood reported a period-end margin book of $22.145 billion, Cash and Deposits of $20.439 billion and Cash Sweep of $32.296 billion. Those figures are not three versions of the same balance. The margin book is money owed by customers who borrowed against securities. Cash and Deposits is a company-defined group of on-balance-sheet cash, segregated cash, qualifying securities and deposits. Cash Sweep is an off-balance-sheet customer deposit measure. The February transfer did not alone create the entire margin book, but it changed the funding mix available to support it.[1]

The three balances are not additive measures of customer cash and should not be treated as interchangeable.[1]

The customer-protection boundary moved with the money

The bank-sweep programme says eligible cash deposited at programme banks can receive pass-through FDIC insurance, generally up to $250,000 at each participating bank when all deposits in the same ownership category are aggregated. Robinhood advertises a larger potential total by distributing cash among multiple banks, but the customer must account for deposits held directly at the same institutions and for any banks excluded from the network. Robinhood itself is not a bank and is not FDIC insured.[2][4]

Broker-held cash instead sits within the securities-account framework. Robinhood states that it is protected by SIPC while held in the brokerage account. SIPC protection is not a guarantee of the cash’s market value and is not the same as deposit insurance. It is designed for the return of missing cash and securities when a member brokerage fails, subject to an overall customer limit of $500,000 including a $250,000 limit for cash claims. The practical outcome in an ordinary functioning account may feel identical; the failure-resolution route is not.[3][4]

That does not establish that customers were made worse off. The first $10,000 threshold means the shifted amount was distributed across a large number of eligible accounts rather than necessarily concentrated in a small number of large balances. Securities regulation requires broker-dealers to maintain customer reserves and capital, and a free credit balance remains a customer claim. The company also says the customer’s interest rate did not change. The finding is narrower: the same product label and yield can conceal a material change in where cash sits and what the broker is using it for.[1][2][3]

The cash-flow statement shows the scale of the liability shift

Robinhood’s nine-month cash-flow discussion provides a second view. It attributes a $2.693 billion year-on-year increase in operating cash flow to a higher payable to users, primarily driven by customer free credit balances. At the same time, it shows a $3.167 billion operating use from increased securities segregated under federal and other regulations. Those movements should not be netted into a claim that all free credit balances were immediately available for unrestricted lending. They show both sides of the broker model: customer liabilities can provide funding, while customer-protection rules require substantial assets to be segregated.[1]

The revenue mechanics also diverge. Cash Sweep produces a spread paid through programme banks. Broker-held balances can support interest-earning assets such as margin receivables, but Robinhood also pays interest to enrolled customers and incurs the costs of reserves, liquidity and capital. The filing notes that higher interest expense on free credit balances reduced net interest revenue associated with segregated cash. A simple comparison of the customer rate with the margin rate would therefore overstate the economic gain because it would omit funding, reserve, credit, operational and capital costs.[1]

The funding match is useful—but not exact

Margin loans and free credit balances can look naturally matched because both sit inside brokerage accounts and both can reprice as interest rates change. Their liquidity behaviour is different. A customer can demand a free credit balance on short notice, while a margin receivable is repaid when the borrower adds cash, sells assets or the broker liquidates collateral. Robinhood therefore still needs cash, segregated assets and contingent liquidity to bridge withdrawals or market stress. The more important test is not whether the two balances are equal at quarter-end, but whether assets remain available when many customers move cash at once.[1][3]

Credit quality also matters. Margin loans are secured by securities and subject to maintenance requirements, but collateral can gap lower before liquidation, particularly in concentrated or volatile portfolios. Customer free credit balances are not secured funding from Robinhood’s perspective; they are liabilities payable on demand under the account agreement. The funding channel can be economically attractive without being costless or risk-free. Robinhood’s filing establishes the scale and purpose of the transfer, while its ordinary risk disclosures—not the programme label—describe the liquidity and credit controls needed around it.[1][3]

What the records prove—and what they do not

The records prove that Robinhood intentionally changed the programme, that more than $6 billion moved from an off-balance-sheet bank sweep to on-balance-sheet free credit balances, and that funding margin growth was the stated purpose. They also establish that the customer agreement permits use of those balances in Robinhood’s business and that the applicable protection framework differs between the two locations. They do not show a breach of customer-protection rules, a shortage in reserves, a loss to customers, or that the transferred cash can be traced dollar for dollar to particular margin loans.[1][2][3]

An alternative explanation is not really in conflict with the finding: Robinhood may view this as an efficient, regulated way to match a growing asset—margin loans—with a stable liability—customer cash—while preserving the customer yield. That can diversify the company away from relying entirely on third-party bank economics. The risk is that customers may reasonably focus on the displayed annual percentage yield and aggregate cash balance without appreciating that the first portion of their enrolled cash is no longer a bank deposit.[1][2][4]

The next verifiable indicators are the size of the margin book, the payable-to-users balance, the amount of segregated assets, the threshold kept as brokerage-held cash, and any change in the programme’s rate or insurance disclosures. A further increase in the brokerage-held threshold would expand the funding channel. A reduction would return more money to programme banks. The essential disclosure test is whether customers can readily distinguish those outcomes before a failure scenario forces the distinction to matter.[1][2][3][4]

Report an error or submit an update