In September 2019, repurchase agreement (repo) rates spiked, rising as high as 10% intraday. The spike was significant as rates were more than 300 basis points above the federal funds target range–thirty times larger than the preceding week. The United States Department of Treasury’s Office of Financial Research (OFR) conducted a study on the spike, which analyzed a combination of intraday timing data and transaction-level data from September 2019. The OFR published a Working Paper in April 2023 identifying three key findings from its study: (1) a confluence of factors contributed to the dramatic spike in repo rates; (2) the repo markets’ segmentation and lack of price transparency exacerbated the spike; and (3) the Federal Reserve’s daily repo operations alleviated the spike. We summarize each of these findings below.
1. Factors that Caused the Dramatic Spike in Repo Rates
The OFR’s study found that a confluence of factors–large Treasury issuances, corporate tax deadlines, and an overall lower level of reserves–caused the 2019 spike in repo rates. While each factor alone would not have been sufficient to cause the spike, the combination of factors occurring at the same time led to the unusual spike in rates.
The OFR found that the disruption of the repo market began on September 16, 2019, when the Treasury settlement coincided with corporate tax deadlines. In the week leading up to the spike in repo rates, the Treasury Department issued $78 billion of government debt that was due to settle on September 16. As a result, the supply of cash available in the financial system declined while the demand for repo increased in order to finance the purchases of Treasury securities. The same day the Treasury settlements were due, corporate tax payments for the third quarter of 2019 were due, which caused a decrease in the money market fund total assets by about $35 billion from the week prior. Together, these events resulted in a large transfer of reserves from the financial market to the government and triggered a liquidity squeeze in repo markets.
In addition to the Treasury settlement and corporate tax deadlines, by mid-September 2019, aggregate bank reserves had declined to a multiyear low of less than $1.4 trillion, while net Treasury positions held by primary dealers had reached an all-time high. The OFR’s study cites the reserve constraints on banks and bank-affiliated dealers as a possible contributing factor in the repo spike because rates in the repo market are highly dependent on the supply of Treasuries and reserves. The confluence of these factors caused an increased demand for repo and decreased liquidity in the repo market, ultimately temporarily driving repo rates higher.
2. Market Segmentation and Lack of Price Transparency Exacerbated the Spike
The OFR’s study suggests the U.S. repo market segmentation and lack of price transparency among market segments exacerbated the spike. In support of this finding, the OFR points to the dispersion of repo rates across different segments in the repo market. For example, on September 16, when cash was scarce in the market, rates began to rise in the Delivery-to-Payment (DVP) interdealer brokered market segment while the rates in the tri-party market segment remained relatively flat throughout the day. The dispersion of rates across segments suggests some market participants knew cash supplies were scarce while others were either unaware or not able to lend in response to increased tightness. When cash became less scarce following the Federal Reserve’s intervention, rates in the DVP market segment substantially decreased.
3. Federal Reserve
The Federal Reserve responded to the September 2019 spike in repo rates with an announcement that it would introduce cash into the market through a repo facility. The OFR’s study suggests the announcement alleviated the spike as repo rates declined substantially in the DVP market segment. Additionally, the liquidity provided by the Federal Reserve eased constraints on dealers, allowing them to move an additional $10 billion into the repo market.
The OFR’s identification of factors that contributed to the 2019 spike in repo rates aids in understanding the sources of volatility in repo markets. The finding that the lack of price transparency contributed to price spikes could be used by regulators to impose transaction reporting in certain areas of the repo market, such as their recent proposal to require price transparency in the bilateral repo market.
The author wishes to acknowledge the contributions of summer associate Stephanie Flynn.