The Fed Funds Arbitrage Complex: How Institutional Plumbing Shapes the Yield Curve
The plumbing of money markets is not neutral. Every structural feature is a policy choice, made by someone, at some moment, with a theory behind it that has since been forgotten.
The overnight federal funds rate is, on paper, the simplest instrument in the American financial system. One bank lends its excess reserves to another, unsecured, for a single night. The rate at which this transaction clears is the federal funds rate — the benchmark upon which trillions of dollars of credit are priced. And yet the apparent simplicity of this rate masks an extraordinary complexity in the plumbing that produces it.
What most market participants understand as “the fed funds rate” is actually the effective federal funds rate (EFFR): a volume-weighted median of transactions reported to the Federal Reserve Bank of New York. But the transactions that compose this rate are not uniformly distributed. They cluster around specific institutional actors, specific times of day, and specific structural incentives that have little to do with the textbook model of interbank lending.
The Arbitrage That Defines the Floor
The most important structural feature of the modern fed funds market is that the majority of lending comes not from commercial banks but from Federal Home Loan Banks (FHLBs). The FHLBs, as government-sponsored enterprises, are ineligible to earn interest on reserve balances (IORB) at the Federal Reserve. This creates a permanent arbitrage: FHLBs lend their excess cash in the fed funds market to banks that can earn IORB, at a rate below IORB. The borrowing banks pocket the spread.
The federal funds market is not a market in the ordinary sense. It is an institutional artifact — the product of regulatory asymmetries that were never designed to interact as they do.
This is not a marginal phenomenon. FHLB lending accounts for the vast majority — often over 90% — of federal funds market volume. The “interbank” market, in the textbook sense, is largely extinct. What remains is a structural arbitrage channel whose pricing is determined by the gap between IORB and the rate at which FHLBs are willing to lend.
Why the Spread Matters
The IORB-EFFR spread is one of the most closely watched indicators in money markets, and for good reason. When the spread widens, it signals an abundance of reserves in the system — the FHLBs have more cash than banks are willing to borrow at tight spreads. When it narrows, it can signal reserve scarcity, a shift in regulatory incentives, or a change in the composition of the borrower base.
But the spread is also a policy variable, whether the Fed acknowledges it or not. By adjusting the IORB rate, the standing repo facility (SRF) rate, and the overnight reverse repo (ON RRP) offering rate, the Federal Reserve implicitly sets the bounds within which the EFFR can trade. The EFFR is not a “market rate” in any meaningful sense. It is an administered rate that emerges from the interaction of policy tools and institutional plumbing.
The Temporal Illusion
Most analyses of the fed funds market operate on a single time horizon: the overnight. This is understandable — the instrument is, after all, an overnight loan. But the structural forces that shape the overnight rate operate on much longer horizons.
Consider the FHLBs’ lending behavior. Their excess cash balances are a function of advance repayments, mortgage prepayment flows, and the maturity profile of their consolidated obligation bonds. These are quarterly and annual phenomena, not overnight ones. When a large volume of FHLB advances mature simultaneously, the resulting cash surge floods the fed funds market and compresses the IORB-EFFR spread — not because of any change in monetary policy, but because of the structural rhythm of the housing finance system.
To understand the overnight rate, you must understand the quarterly. To understand the quarterly, you must understand the institutional. Most analysts stop at the overnight. The consequences are systematic.
Similarly, reserve distribution matters. The fed funds market is not a single market — it is a network of bilateral relationships, each shaped by balance sheet constraints, regulatory capital requirements, and the internal transfer pricing models of large bank holding companies. A bank’s willingness to borrow fed funds is not simply a function of the rate offered; it is a function of how that borrowing interacts with its leverage ratio, its liquidity coverage ratio, and its internal cost-of-funds model.
Implications for the Yield Curve
The yield curve is conventionally understood as a sequence of forward rates, each reflecting the market’s expectation of future short-term rates plus a term premium. But if the short-term rate itself is an institutional artifact — shaped by FHLB behavior, reserve distribution, and regulatory constraints — then the foundation of the yield curve is less “expectational” than it appears.
This has profound implications for term premium estimation. If the overnight rate is persistently influenced by structural factors that are orthogonal to monetary policy expectations, then models that decompose the yield curve into “expectations” and “term premium” components may systematically misattribute structural plumbing effects to one category or the other.
The practical consequence is that investors who rely on term premium estimates to make duration decisions may be acting on a signal that is partly noise — noise generated not by changes in risk preferences or inflation expectations, but by the plumbing of the money market itself.
A Design Problem, Not a Market Problem
The fed funds market, as it exists today, is not the product of deliberate design. It is the accumulated result of decades of institutional evolution, regulatory layering, and crisis-era improvisation. The FHLBs were created in 1932 to support housing finance; their role as the dominant fed funds lender was never intended. The IORB regime was introduced in 2008 as an emergency measure; its role as the primary tool of rate control was never the original plan.
This is not necessarily a problem. Institutional systems often function well despite — or because of — their historical contingency. But it does mean that the fed funds market cannot be understood through the lens of a simple supply-and-demand model. It requires an institutional lens: one that takes seriously the specific actors, incentives, and constraints that produce the rate we observe.
The plumbing of money markets is not neutral. Every structural feature is a policy choice, made by someone, at some moment, with a theory behind it that has since been forgotten. Understanding the fed funds market requires recovering those choices — and recognizing that the rate we observe each morning is not the verdict of “the market” but the output of a machine whose design we have collectively forgotten.