The Core Idea
In the third quarter of 2026, BlackRock reported about $250 million in securities lending revenue. That money did not come from managing portfolios or charging expense ratios. It came from lending out shares held inside its funds to outside borrowers, collecting a fee, and splitting the proceeds. The fund investors who owned those shares were not asked.
That fee is securities lending income. A fund that holds stock on behalf of millions of people quietly lends some of it to short sellers and market makers, takes collateral in return, and earns a small spread. The transaction is buried in a footnote, not a headline.
The machine we are drawing is the lending program that runs inside nearly every large index fund and ETF. Roughly $4.7 trillion in securities sat on loan globally as of the third quarter of 2026. The reader's retirement fund, his S&P 500 index holding, his bond ETF, each one likely lends a slice of what it holds. The revenue is real. So is the risk, and the split.
What Happened
Securities lending has been growing steadily, but 2026 pushed it to new levels. Global lending revenue hit about $5.6 billion in the third quarter alone, up more than 20 percent from a year earlier. Loan balances climbed to $4.7 trillion. The business is no longer a side operation. It is a major revenue line for the firms that run the largest funds.
The scale at a single firm makes the point. BlackRock's lending arm, an affiliate called BTC, acts as lending agent for its own funds. The funds keep about 81 percent of the income. BlackRock's affiliate keeps the rest. In a quarter where lending revenue reached $250 million, that split sent roughly $47 million to the manager's side of the table, earned on assets the fund investors supplied.
Why this reaches the reader: his index fund almost certainly participates. If he holds a Vanguard equity, BlackRock, or State Street fund, his shares are part of the pool available for loan. He never signed a separate lending agreement. The fund's prospectus authorized it. The income shows up in a line item most investors skip, and the risk sits one level deeper than that.
They Proved It Worked in '76. Then Buried It.
In 1976, Chevron tapped an energy source powerful enough to run a city. No fuel costs. No supply chain. Then they killed it. Unocal and Texaco did the same.
Now one company has spent sixty years perfecting what Big Oil refused to touch. Google locked in a 15-year contract. Bill Gates wrote a $100 million check.
On November 17th, Washington hands it an edge no other energy source gets.
Structural Lens: Why This Can Happen to a Giant
A securities loan begins with an asset that would otherwise sit idle. The fund holds thousands of stocks. Most days, those shares do nothing but track the index. A lending program puts them to work. A borrower, typically a broker or a short seller who needs to deliver shares, requests them. The fund ships the shares out and receives collateral in return, almost always cash equal to at least 102 percent of the shares' value.
That collateral is the pivot. The fund does not leave it sitting in a vault. It reinvests the cash in short-term instruments, usually money market funds or Treasury bills, and earns a return. The borrower pays a fee. The gap between the fee earned, the return on the collateral, and the costs is the lending profit. The whole structure works because the collateral exceeds the loan, and both sides mark to market every day.
The load-bearing part is the collateral reinvestment. The shares are out the door, and what backs them is not the borrower's promise alone but a pool of cash invested in something else. If that something else loses value, the fund has shipped out real shares and holds collateral that is no longer worth enough to buy them back. Every dollar of lending profit traces back to this reinvestment choice.
The lending agent is usually an affiliate of the fund manager. BlackRock's funds use BlackRock's own subsidiary. State Street's funds use State Street. That agent decides which borrowers qualify, what collateral to accept, and where to reinvest. The agent earns a cut of the income for running the program. The fund's board approves the arrangement, but day to day, the same company that manages the fund also runs the lending desk.
Here is the twist that catches people. The fund bears the investment risk on the reinvested collateral, but the lending agent earns its fee regardless of how that reinvestment performs. If the collateral pool loses value, the loss comes out of the fund, not the agent. The agent's income is a percentage of gross lending revenue, collected before any reinvestment loss is subtracted. The interests are not perfectly aligned, and the footnote does not make that plain.
Risk Transfer: Where the Pressure Builds
The first joint is the reinvestment pool inside the fund. In 2008, AIG's lending operation held roughly $75 billion in borrower collateral and reinvested it in mortgage-linked securities. When those securities collapsed, the collateral shrank to about $60 billion. AIG owed borrowers $75 billion back and had $60 billion to pay it with. The gap nearly destroyed the company. Fund lending programs today face tighter rules, but the structure is the same: the fund's collateral pool is only as safe as the paper it buys.
The second joint is the affiliate relationship. When the lending agent is a subsidiary of the fund manager, the manager earns revenue twice: once from the expense ratio and once from the lending program. Regulators require an SEC exemptive order for this arrangement, and BlackRock operates under one. The fund keeps the larger share of lending income, but the agent sets the terms. Conflicts do not require bad faith. They only require different incentives on opposite sides of the same table.
The third joint is opacity. The SEC adopted Rule 10c-1a in 2023 to force public reporting of covered securities loans. The rule was supposed to take effect in 2026. It has been delayed twice. Reporting will not begin until September 2028 at the earliest, and public data will not appear until 2029. Until then, the reader cannot see how much of his fund is on loan, to whom, at what rate, or what the collateral is invested in. He owns the shares. He bears the risk. He cannot see the terms.
What Can Persist (And What Can Break)
What persists: the basic bargain. Securities lending genuinely reduces costs for index fund investors. The income, small as it is per share, offsets a portion of the fund's expenses and can make a low-cost fund effectively even cheaper. Collateral rules are stricter than they were before 2008. Most large lending programs reinvest conservatively. For a long-term holder of a broad index fund, the lending program is a net positive in normal markets.
What can break: the assumption that the collateral is always money-good. Lending programs failed in 2008 not because borrowers defaulted but because the reinvested collateral lost value. The stress does not come from the loan. It comes from what the fund did with the cash it received. In a crisis where short-term credit markets freeze, the safest-looking reinvestment pool can gap down overnight, and the fund must still return full collateral to the borrower.
Bottom Line
This machine sits on the other side of the locate system we have traced, the check a broker runs before every short sale. The borrower who passes that check gets his shares from a lending program like this one, inside a fund the reader owns. It connects, too, to the brokerage cash sweep we have followed, another place where idle assets earn revenue for the manager before the investor. The pattern is the same: assets the reader thinks are sitting still are in motion, and someone else is earning a fee on the movement.
The next reading arrives with the SEC's decision on whether Rule 10c-1a will survive to its 2028 deadline or face another delay. That will decide whether the reader ever gets to see the lending terms on his own shares, or whether the machine keeps running in the quiet where it was built.

