A 15% stock-lending share is not always 15% of the same thing

Four US investing platforms advertise customer participation in stock-lending income with percentages clustered between 10% and 15%. The similarity is misleading. Robinhood says customers receive up to 15% of the gross revenue it generates by lending securities. Webull publishes a formula that multiplies loaned value by the annualised interest rate and then by 15%. Public says its customers receive 10% of the total net proceeds earned and received by Apex. SoFi’s published programme announcement describes 15% of total net proceeds. The percentages look ready for a league table, but the denominators are not the same.[1][2][3][4][5]

The difference between gross and net is economically decisive. Suppose a borrower pays $100 in lending revenue. A programme paying 15% of gross would allocate up to $15 to the customer before considering how the broker funds its own costs from the remainder. A programme paying 15% of net could produce the same $15 only if net proceeds equal gross proceeds. If $20 of expenses, counterparty allocations or other deductions are taken before “net proceeds” are calculated, 15% of the remaining $80 is $12. A 10% net share would be $8. The example is illustrative, not an estimate of any firm’s deductions, but it demonstrates why the headline rates cannot be compared without the base.[1][2][3][4]

Four disclosures, three different formulations

Robinhood’s support page uses the clearest gross formulation. It says a participant receives up to 15% of the gross revenue Robinhood generates when it lends securities to the market. The daily calculation applies up to 15% to the weighted average rebate rate earned by Robinhood for the stock and multiplies it by the number of shares borrowed. The words “up to” matter: 15% is a ceiling, not a guaranteed fixed share for every security and every day.[1]

Webull’s page provides an algebraic formula: loaned share quantity multiplied by market price, multiplied by the annualised interest rate divided by 360, multiplied by 15%. That is closer to a gross-style calculation because the published rate is applied directly to the value of the loan rather than to an explicitly defined “net proceeds” pool. But the page also says Apex, Webull’s clearing firm, identifies the stocks to lend. The formula does not, by itself, disclose every commercial allocation among the borrower, Apex, Webull and any other intermediary.[2]

Public’s language is different. Apex operates the programme and shares a portion of the interest it receives with the customer and Public Investing. The customer receives 10% of the total net proceeds earned and received by Apex for lending the shares, and Public says the rate may change but will not fall below 10%. “Net proceeds” is the operative denominator. The page does not state the gross borrower fee, each deduction used to reach net proceeds or Public’s own share. It would therefore be wrong to infer that Public simply keeps 90% of the borrower’s gross payment.[3]

SoFi’s launch announcement likewise described a payment equal to 15% of total net proceeds, with Apex facilitating the programme. Its current support material confirms the programme and its risks but, in the sources examined for this article, does not restate the 15% rate. The precise formulation should therefore be presented as SoFi’s published launch term, not as an independently verified November 2026 contractual rate. A current comparison should check the customer agreement or live in-app disclosure before treating it as unchanged.[4][5]

A percentage cannot reveal the actual yield

Even where two programmes use the same denominator, the customer’s realised return depends on securities-specific demand. Hard-to-borrow shares can command high loan rates; widely available securities may generate almost nothing and may never be borrowed. Robinhood says income varies and eligible shares are not guaranteed to be lent. Public makes the same point. Webull says the interest rate is variable and often changes daily. A platform paying a lower percentage of a much higher lending rate can produce more customer income than one paying a higher percentage of a low rate.[1][2][3]

The programme operator also controls important parts of the process. Customers generally cannot select the borrower, negotiate the rate or guarantee that their shares will be used. Robinhood’s programme considers all eligible positions once enabled and states that Robinhood Securities may use borrowed stock for onward lending, settlement or collateral. Webull says Apex selects shares according to market demand. Public says Apex is the customer’s counterparty and may lend whole shares to investors and institutions. Those structures affect both the revenue chain and the credit exposure.[1][2][3]

A rigorous broker table would therefore report at least five variables: the customer-share percentage; whether it applies to gross revenue, gross rebate, net proceeds or another base; whether the rate is fixed or “up to”; who is the direct borrower from the customer; and how the realised dollar payment is shown. Without those fields, ranking 15% above 10% is not analysis. It is a comparison of numerators stripped from their denominators.[1][2][3][4]

The non-cash terms may matter more than the rate

Securities lending changes the legal and tax characteristics of the position while the shares are on loan. Robinhood says the investor loses shareholder voting rights during the loan and may receive a manufactured payment or cash in lieu rather than a qualified dividend. Public likewise warns that customers waive voting rights and may receive payments in lieu with different tax treatment. Webull identifies cash-in-lieu treatment for distributions. Those consequences can outweigh a small monthly lending payment, particularly around a contested corporate vote or a dividend record date.[1][2][3]

The protection framework changes too. Webull says shares loaned under its fully paid lending programme are not covered by SIPC, while collateral is posted for customer protection. Public says loaned shares lack SIPC protection and describes collateral posted by Apex in a custody account, marked to market. The risk is not that the investor loses market exposure—the customer still bears gains and losses in the share price—but that the lending counterparty must return equivalent securities and the collateral process must work if it defaults.[2][3]

This does not make stock lending inherently improper. It is a longstanding market function that supports settlement and short selling, and the examined brokers disclose meaningful risks and payment mechanics. The investigative issue is comparability. A consumer sees four percentages in the same numerical neighbourhood, yet one is “up to” a share of gross revenue, another is embedded in a loan-value formula, and two are shares of net proceeds. The commercial chain is different even before security demand, taxes, voting and collateral are considered.[1][2][3][4][5]

What a fair comparison should say

The defensible conclusion is not that one programme necessarily pays more. Publicly available pages do not supply the same security, loan period, borrower rate and gross-to-net reconciliation across all four firms. What the evidence establishes is that the advertised percentages are not interchangeable. Robinhood’s 15% ceiling and Public’s 10% floor operate on differently described bases. Webull’s formula is more directly tied to loan value and annualised rate. SoFi’s 15% figure is located in a published launch announcement and needs current-contract confirmation.[1][2][3][4][5]

A better disclosure standard would show a worked example from borrower payment to customer payment: gross fee received, each permitted deduction, net proceeds, the broker or clearing firm’s retained share and the customer’s amount. It would also state the average realised customer share across all loans, not only the contractual maximum or minimum. Until brokers publish that bridge consistently, a stock-lending percentage should be read as one component of an opaque revenue waterfall—not as a self-sufficient measure of customer value.[1][2][3]

A worked example shows why the denominator dominates

Assume a borrower pays a 10% annualised fee on securities worth $10,000 for a full year, producing $1,000 of gross lending revenue before intermediary costs. Under a programme paying 15% of gross revenue, the customer would receive $150. Under a programme paying 15% of net proceeds, the result depends on deductions. If the lending agent retained $300 of the $1,000 before the split, net proceeds would be $700 and the customer’s 15% would be $105. At 10% of those same net proceeds, the customer would receive $70. These are illustrations, not the actual economics of any reviewed loan.[1][2][3][4][5]

The example demonstrates why a higher advertised percentage can produce a lower payment when applied to a narrower base. It also shows why “the broker keeps 85%” can be inaccurate for net-proceeds programmes. Part of the difference between the borrower’s fee and the customer’s payment may be retained by a lending agent, clearing broker, custodian or other intermediary before the broker’s stated split is applied. The current pages do not publish one standardised waterfall covering every loan.[1][2][3][4][5]

Tax, voting and recall rights are separate from the split

Lending can replace an ordinary dividend with a payment in lieu, potentially changing tax treatment depending on the customer and jurisdiction. The lender ordinarily loses voting rights while shares are on loan and may need to recall them before a record date. A programme can have an attractive revenue share yet still be unsuitable for a customer who prioritises voting, tax characteristics or immediate control.[1][2][3][5]

Collateral and counterparty protections also matter. Fully paid securities lending typically requires collateral and daily valuation, but the form, amount, custodian and insolvency treatment should be read in the governing agreement. A customer’s exposure is not identical to an uninsured unsecured loan merely because shares leave the account, but neither is it eliminated by quoting a percentage of lending revenue.[1][2][3][5]

A comparable disclosure would show the whole waterfall

A useful standard would require firms to disclose, for a representative loan, the borrower’s gross fee, agent or clearing deductions, the broker’s share, the customer’s share, the duration and the annualised yield on the customer’s security value. It should also distinguish a contractual minimum from a promotional example and state how often rates can change.[1][2][3][4][5]

Until those fields are standardised, comparison sites should not rank programmes by the headline percentage alone. Robinhood’s and Webull’s 15% language is not automatically equivalent to SoFi’s 15% of net proceeds, and Public’s minimum 10% of Apex net proceeds is not automatically less generous on every loan. The realised dollar payment is the only outcome that incorporates the loan rate, utilisation, deductions, split and time.[1][2][3][4][5]

Utilisation can matter more than the split

A generous revenue share produces nothing while shares are not borrowed. Loan demand concentrates in securities that are scarce, heavily shorted or needed for settlement. A broad portfolio of highly liquid large-cap shares may generate little income even under a high split, while one hard-to-borrow security can generate a much larger return for a period. Programmes generally do not promise that any eligible holding will be lent.[1][2][3][5]

Customers also need a denominator for advertised yield examples. An annualised percentage based on a brief high-fee loan can look substantial even if the shares were borrowed for only a few days. A comparable statement should show realised dollars, average value on loan, days on loan and the customer’s annualised yield over the same period. That would expose the combined effects of utilisation and split instead of promoting one favourable rate.[1][2][3][4][5]

Report an error or submit an update