What repo actually is
A repurchase agreement, or repo, is a collateralized short-term loan. One side delivers Treasury securities and receives cash. The other side delivers cash and receives Treasuries as collateral. The next morning — or in a week, or in a month — the trade unwinds at a slightly higher cash amount. The difference is the interest rate, the repo rate.
That sounds plumbing-level boring, and it is. It also happens to be one of the largest markets in the world, the dominant source of overnight financing for U.S. Treasury inventories held by primary dealers, and the rate that anchors the entire short-end curve.
The two main venues
U.S. repo splits into two structurally different markets:
- Tri-party repo, cleared through the Bank of New York. Money-market funds lend cash; primary dealers borrow against Treasury collateral; BNY handles the collateral management. This is the largest, simplest, and most standardized venue.
- Bilateral repo, where two counterparties trade directly — including sponsored repo, where a dealer sponsors a hedge fund into the FICC clearinghouse so the cash side is netted on the dealer’s balance sheet. This is where leveraged-fund Treasury positioning gets financed.
The two venues serve different counterparties and respond to different stress signals. Tri-party tells you about the relationship between cash investors and dealer balance sheets. Bilateral — particularly sponsored DVP — tells you about hedge-fund leverage in the Treasury basis trade and other relative-value positions.
SOFR, IORB, and the corridor
The benchmark overnight rate in the U.S. is now SOFR (Secured Overnight Financing Rate), which is calculated from a broad sample of repo transactions across tri-party and cleared bilateral venues. SOFR is what replaced LIBOR.
The Fed surrounds SOFR with two administered rates:
- IORB (Interest on Reserve Balances): the rate the Fed pays banks on reserves. This is the upper bound of the corridor — in theory, banks should not lend cash overnight at less than what they earn at the Fed.
- RRP rate (Reverse Repo): the rate the Fed pays counterparties (mostly money-market funds) to park cash overnight. This is the lower bound — eligible counterparties should not lend cash overnight for less than they can earn at the Fed.
In a comfortable funding regime, SOFR sits modestly above the RRP rate and modestly below IORB. Money-market funds prefer to lend in the private repo market because they earn a few basis points more than RRP. The system has slack.
The single most important spread to watch is SOFR − IORB. When this spread is comfortably negative, the system is well-supplied with reserves and funding is easy. When it pushes toward zero, balance sheets are getting full. When it goes positive — SOFR above IORB — you are in a funding stress regime.
The reverse repo facility, and why it matters as a buffer
The Fed’s overnight RRP facility is a kind of cash absorbent. When the system has more cash than it knows what to do with — for example, after a long period of QE or after a TGA drawdown — money-market funds can park that cash at the Fed and earn the RRP rate.
RRP’s role in the broader plumbing is twofold:
- It puts a floor under short-end rates by giving cash investors an outside option.
- It functions as a buffer that the Treasury can drain when it issues bills to rebuild the TGA. As long as RRP balances are large, bill issuance is largely absorbed by money-market funds shifting cash from RRP into bills, and bank reserves are barely affected.
This is why the level of RRP balances is one of the cleanest reads on how much slack the system has. When RRP is large, the funding market can absorb a lot of supply without flinching. When RRP is depleted, the next dollar of bill supply has to come out of bank reserves — and that is the regime in which funding can suddenly tighten.
How dealers actually use repo
Primary dealers run large inventories of Treasuries to make markets. They cannot fund those inventories out of equity capital — the size is far too large. They fund them in repo, rolling overnight against tri-party cash.
This has a few consequences:
- Dealer balance sheet is the binding constraint. When dealers are full — quarter-end, year-end, or after large auction settlements — the marginal repo rate has to clear higher to compensate for the balance-sheet cost. SOFR pops, the GC-OIS spread widens, and bills cheapen relative to OIS.
- Hedge-fund leverage rides on top. Relative-value funds finance Treasury basis trades and swap-spread trades through bilateral repo. When financing tightens, those trades have to deleverage — and that selling can land in the cash Treasury market.
- Equity prime brokers need GC repo to work. Cash equities don’t finance in Treasury repo, but the broader securities-financing complex — including margin loans and securities lending — runs on the same dealer balance sheet. When repo gets tight, prime-brokerage financing tends to follow with a lag.
The connection to risk assets
The link between repo and equities is not a single equation. It runs through several channels at once:
- Cost of carry. Higher SOFR — absent a Fed move — raises the cost of running long inventory in any asset, from Treasuries to equities. Multiples are sensitive to the level and stability of this cost.
- Dealer hedging capacity. When dealer balance sheets are full, they have less capacity to warehouse risk. Bid-ask widens, gamma hedging is less efficient, and equity volatility tends to be higher for the same flow.
- Margin and prime-brokerage financing. Persistent funding tightness raises the cost of leverage for hedge funds and other levered participants. When that cost rises faster than expected returns, position-cutting follows.
- Reflexivity. Funding stress that shows up in repo, in cross-currency basis, and in commercial-paper spreads tends to be a leading indicator of credit-spread widening, which then leads equities. By the time the equity market is the headline, the plumbing has usually been signaling for weeks.
What to actually watch
The short list of repo-and-funding indicators worth keeping on a daily dashboard:
- SOFR vs. IORB: the corridor position. The single best summary measure.
- SOFR vs. RRP rate: tells you whether money-market funds are happy in private repo or being pushed into the Fed.
- RRP balances: the funding-market slack indicator. The trajectory matters more than the level.
- Tri-party GC vs. bilateral DVP repo rates: divergence here often signals hedge-fund-specific stress (the basis trade) rather than system-wide stress.
- Quarter-end and year-end prints: predictable but informative. The size of the SOFR pop and the speed of the normalization are the cleanest read on dealer balance-sheet capacity.
Where the framework breaks
Two failure modes to avoid:
- Treating high RRP as automatically bullish. Large RRP balances mean the system is well-supplied with cash, but they can also mean money-market funds don’t want to take any private credit risk — a defensive posture. Direction of change matters more than level alone.
- Reading every SOFR pop as funding stress. A great deal of the short-end action is calendar-driven (quarter-end, tax dates, settlement bunching) and reverses on its own within days. Sustained moves are what matter; one-day spikes usually don’t.
Putting it together
Repo is the engine room. SOFR is the dial. RRP is the buffer tank. And risk assets ride on top of all of it — usually invisibly, until the engine starts running hot. For a strategist or a publisher trying to make sense of cross-asset moves, the worst position to be in is staring at equities and trying to back out a story when the story is already visible in the plumbing data.
That’s the angle I bring to daily research notes for institutional clients: read the plumbing first, narrate the tape second.
Need this kind of analysis for your platform?
Daily macro and cross-asset commentary for broker-dealers, trading platforms, and financial publishers.
Send an Inquiry →