The Repo Market Crisis: When Overnight Rates Hit 10% and the Fed Returned

Khan Capital | October 2019

---

Key Takeaways

• Overnight repo rates spiked to 10% on 17 September 2019, four times the Federal Reserve’s target range, as a confluence of corporate tax payments, Treasury settlement pressures, and depleted bank reserves exposed the fragility of the post-crisis monetary plumbing that underpins the entire financial system.
• The Federal Reserve was forced to intervene for the first time since 2008, injecting over $75 billion in overnight repo operations and subsequently launching a programme of Treasury bill purchases that added $60 billion per month to its balance sheet, a de facto return to quantitative easing that the Fed insisted was merely “reserve management.”
• The repo crisis revealed that the Fed had shrunk its balance sheet too far through its 2017 to 2019 quantitative tightening programme, reducing bank reserves from $2.8 trillion to $1.5 trillion and crossing the invisible threshold below which the financial system’s demand for liquidity exceeded the supply.
• The episode foreshadowed the far more severe liquidity crisis of March 2020, demonstrating that post-2008 regulations, while making individual banks safer, had created systemic fragility in the plumbing that connects them, a vulnerability that would be tested catastrophically six months later.

Part of: Market Crises & Crashes — Khan Capital’s hub on market crashes and financial history.

The Repo Market Crisis: When the Plumbing Seized

On the morning of 17 September 2019, something broke in the most fundamental market in global finance. The overnight repurchase agreement (repo) rate, which normally trades within a few basis points of the federal funds rate, surged to 10%, a level not seen since the 2008 financial crisis. Banks and broker-dealers that routinely borrowed cash overnight against Treasury collateral found themselves unable to do so at reasonable rates. The secured overnight financing rate (SOFR), the benchmark intended to replace LIBOR as the reference rate for trillions of dollars in financial contracts, spiked to 5.25%. For a few hours, the machinery that enables the modern financial system to function was grinding to a halt.

The repo market is the circulatory system of global finance. Every day, roughly $1 trillion in overnight and short-term loans are transacted, with institutions lending cash against government securities as collateral. Banks use repo to manage their day-to-day liquidity. Hedge funds use it to finance leveraged positions. Money market funds use it to deploy cash. When repo markets function normally, they are invisible. When they seize, everything downstream is affected, from Treasury market liquidity to corporate borrowing costs to the Fed’s ability to implement monetary policy.

What Caused the Spike

The proximate triggers were mundane: quarterly corporate tax payments (roughly $35 billion withdrawn from the banking system) coincided with the settlement of $54 billion in newly issued Treasury securities, both occurring on 16 September. These predictable cash drains would normally be absorbed by excess reserves in the banking system. But the reserves were no longer there.

The Federal Reserve had been shrinking its balance sheet since October 2017 through a process known as quantitative tightening (QT), allowing up to $50 billion per month in maturing Treasury and mortgage-backed securities to roll off without reinvestment. Bank reserves, which had peaked at approximately $2.8 trillion in 2014, had been drawn down to roughly $1.5 trillion by September 2019. The Fed had assumed, based on surveys and modelling, that the banking system’s minimum comfortable reserve level was somewhere around $1.2 trillion. The September repo spike demonstrated that this estimate was wrong, or at least that the distribution of reserves across institutions mattered as much as the aggregate level.

The concentration problem was critical. JPMorgan Chase alone held over $300 billion in reserves, but post-crisis regulations, including the Liquidity Coverage Ratio (LCR) and internal stress-test requirements, made these reserves effectively unavailable for lending into the repo market even at highly attractive rates. The largest banks, which had traditionally served as shock absorbers in money markets, were constrained by the very regulations designed to make them safer. The system-wide reserve level may have been adequate in aggregate, but the reserves were locked inside institutions that could not, or would not, deploy them.

Date	Overnight Repo Rate	Fed Funds Target	Fed Intervention
16 Sep 2019	2.43% (+25bp above normal)	2.00-2.25%	None
17 Sep 2019	~10% (intraday)	2.00-2.25%	$53bn overnight repo
18 Sep 2019	~3.0%	2.00-2.25%	$75bn overnight repo
19 Sep 2019	~2.1%	1.75-2.00% (cut)	$75bn + 14-day term repo
Oct-Nov 2019	Normalised	1.75-2.00%	$60bn/month T-bill purchases begin

The Fed’s Response: “Not QE”

The Federal Reserve’s response was swift but awkward. On 17 September, the New York Fed conducted its first overnight repo operation since the financial crisis, offering $75 billion in cash against Treasury collateral. The operation was oversubscribed, confirming the severity of the liquidity shortfall. Over the following days, the Fed expanded both overnight and term (14-day) repo operations, and by October had announced a programme of purchasing $60 billion per month in Treasury bills to rebuild the reserve buffer.

Fed Chair Jerome Powell was at pains to distinguish this balance sheet expansion from quantitative easing. “In no sense is this QE,” he insisted at the October press conference. The distinction was technically defensible: QE involves purchasing longer-duration securities to push down long-term rates, while T-bill purchases are designed to increase reserves without directly affecting the yield curve. But the practical reality was that the Fed was expanding its balance sheet by $60 billion per month, just months after completing a quantitative tightening programme that it had described as running on “autopilot.” The credibility cost was significant.

What the Market Was Misunderstanding

The market treated the repo crisis as a technical glitch: a plumbing problem that the Fed quickly fixed with temporary operations. This characterisation was dangerously complacent. The repo spike revealed a structural vulnerability in the post-crisis financial architecture that had not been addressed and, six months later, would contribute to the far more severe liquidity crisis of March 2020.

The fundamental problem was the interaction between two well-intentioned but incompatible policy frameworks. Post-2008 banking regulations required institutions to hold large buffers of liquid assets, including reserves. Simultaneously, the Fed’s QT programme was reducing the aggregate supply of reserves. When the supply of reserves dropped below the level demanded by regulation-constrained banks, the price of reserves (the repo rate) spiked. The system was not broken by a failure of regulation but by the unintended interaction of regulation with monetary policy.

The market also underestimated the forward implications. The repo crisis demonstrated that the Fed could not withdraw from its post-crisis balance sheet expansion without destabilising money markets. This implied that a significant portion of the Fed’s balance sheet was now structurally permanent: reserves could not drop below roughly $1.5 trillion without risking a repeat of September’s dysfunction. The era of genuinely normalised monetary policy, with a small central bank balance sheet and market-determined short-term rates, was effectively over.

Investor Implications

Fixed income: The repo crisis and the Fed’s response confirmed that the Fed will act aggressively to prevent money market dysfunction, effectively capping the upside on short-term rates. This supports carry strategies in short-duration fixed income but raises questions about the long-term trajectory of the Fed’s balance sheet and its implications for inflation expectations.

Financials: Bank stocks were largely unaffected by the repo spike, as the crisis was a system-level liquidity issue rather than a credit event. However, the episode highlighted the unintended consequences of post-crisis regulation on bank market-making capacity, a structural headwind for bank profitability that is unlikely to reverse.

Systemic risk: The repo crisis is a warning about the fragility of financial plumbing that most market participants take for granted. Investors should stress-test portfolios for scenarios in which overnight funding markets seize, including the knock-on effects on leveraged strategies, money market funds, and collateral chains. The March 2020 liquidity crisis demonstrated that these scenarios are not hypothetical.

Conclusion

The repo market crisis of September 2019 was a canary in the coal mine. It revealed that the post-2008 financial system, while safer at the individual institution level, harboured new vulnerabilities at the system level. The interaction of stringent bank liquidity regulations with the Fed’s balance sheet reduction created a reserve scarcity that the central bank did not anticipate and could not prevent without reversing course on quantitative tightening. The Fed’s “not QE” response stabilised money markets but established a precedent that would prove consequential: the central bank’s balance sheet could not be meaningfully reduced. The repo crisis was small in its immediate impact but enormous in what it revealed about the structural dependence of modern financial markets on central bank liquidity provision.

Sources: Federal Reserve Bank of New York repo operation results and SOFR data; Federal Reserve Board H.4.1 statistical releases and FOMC meeting minutes; BIS Quarterly Review analysis of repo market dynamics; JPMorgan Chase reserve holdings from 10-Q filings; Bloomberg terminal data for repo rates and Treasury yields.

Related Reading: The repo crisis was a precursor to the far more severe March 2020 liquidity crisis that exposed the same structural vulnerabilities at catastrophic scale. For the Fed’s broader policy trajectory, see Fed Goes Nuclear and the Great Taper. The “not QE” debate resurfaced when the Fed later began its most aggressive hiking cycle in 40 years. The pivot that ended tightening and set up the QT reversal is covered in The Powell Pivot: How the Fed Blinked and Markets Roared Back.

Update (May 2026). See Khan Capitals’ latest coverage: FOMC 8-4 split and the Q1 stagflation print.