Single-Tranche Term Repurchase Agreements (US)

As mortgage defaults and foreclosures continued to climb, the severe strains that started to plague credit markets in the middle of 2007 worsened further. Losses on housing-related securities and derivative instruments continued to climb, causing substantial damage to the balance sheets of large financial institutions that had levered up on these same securities. As their positions worsened, banks found it increasingly difficult to attract funding that wasn’t priced at exorbitantly high rates or for very short terms. Term funding markets, specifically those that centered on agency mortgage-backed securities (MBS), quickly dried up as fears of illiquidity and even insolvency spread. To remedy these concerns, the Federal Reserve  announced a program called the Single-Tranche Term Repurchase Agreements, which auctioned off repurchase agreements (repos) to primary Dealers every week. This provided a critical source of funding to these institutions, which, at the time, could not access other avenues of funding, such as the discount window. The repos were short term, priced at market rates, and matured 28 days after the settlement date. Of the 20 institutions categorized as primary dealers at the beginning of 2008, 19 participated in the program, which had auctions running from March 7, 2008, to December 31, 2008. Usage peaked at, but never exceeded, $80 billion per month, though the Fed said in its initial press release that the program’s size could have gone up to $100 billion. While the program was smaller compared to other market liquidity initiatives, ST OMO operated at capacity for most of its duration, and spreads between agency MBS repo and Treasury repo rates fell dramatically toward the end of the issuance window.

On March 7, 2008, the Federal Reserve  announced the Single-Tranche Term Repurchase Agreements facility, a set of single-tranche open market operations (ST OMO) that were designed to alleviate severe liquidity-related stresses on the interbank lending market, particularly with respect to agency mortgage-backed securities (MBS). The program, which used Section 14 of the Federal Reserve Act as its legal basis, had primary dealers participate in a series of auctions of 28-day repurchase agreements (repos) to provide an additional mechanism for short-term funding. These institutions did not have access to facilities like the discount window, and thus their access to lender-of-last resort facilities was limited. In order to obtain repo funding, dealers also had to pledge collateral in the event they would be unable to pay the agreement back.  

Eligible collateral included (1) Treasury securities, (2) federal agency debt, and (3) mortgage-backed securities that were issued or guaranteed by federal agencies. ST OMO was primarily just a series of modified conventional open market operations that the Federal Reserve Bank of New York (FRBNY) routinely conducted, with the primary difference being that normal repos had several tranches that were priced differently based on the riskiness of the underlying collateral. ST OMO, however, had only one tranche, thus allowing riskier collateral to be utilized at a lower effective rate. This was most notably seen with agency MBS, which was seen as the riskiest type of OMO collateral and was used heavily in the auctions.

The auctions began on March 7, 2008, with the first repo agreement settling on March 10. Of the 20 institutions categorized as primary dealers, 19 participated in the auctions. The auctions were conducted at market rates, which ranged from an average of about 280 bps for the set of auctions on March 7, 2008, to an average of 8 bps for those on December 31 2008. A total of 375 auctions were conducted over the program’s issuance window, totaling $855 billion in trade value. From April 30 to the expiration of the program’s issuance window on December 31, 2008, there were $80 billion in repurchase agreements outstanding, indicating considerable primary dealer participation.

In general, there was minimal formal evaluation done for ST OMO. It is possible that the program was used as a signaling device in an attempt to destigmatize participation in some of the traditional facilities, especially since most of the collateral pledged was mortgage backed securities. The program was fairly small and was announced around the same time as other programs like the Term Auction Facility (TAF), Term Securities Lending Facility (TSLF), and Primary Dealer Credit Facility (PDCF). Some suggest that, based on the consistently capped usage and falling repo rates as ST OMO wound down, the program was a success for the institutions that used it most. However, others suggest a limited US impact, since the biggest borrowers were foreign institutions such as Credit Suisse, Deutsche Bank, and BNP Paribas.

ST OMO was not given much attention nor formal evaluation. It was simply a set of open market operations done under a lender-of-last-resort pretense rather than as part of conventional monetary policy. This led to two important distinctions. First, that the funding term for ST OMO repos was longer than for conventional repurchase agreements. Repo and reverse repo agreements conducted in normal times most commonly had overnight terms, with the longest term allowed normally being 65 days. Second, the single-tranche aspect of the program allowed agency MBS to be put up as collateral at a lower effective rate despite their underlying risk, thus providing much-needed funding to considerably strained term funding markets, which had all but frozen. However, since ST OMO was also introduced around the same time as many other programs, such as TSLF and the PDCF, it was more difficult to disentangle the stand-alone impact it had on term funding markets.

Michael J. Fleming explained that, while ST OMO had market-determined pricing, which appeared to keep auction rates high, the usage of the program and the subsequent decline in stop-out rates suggests that the auctions were priced well (Fleming 2012). Certain institutions, such as Credit Suisse, Goldman Sachs, BNP Paribas, Countrywide, and Cantor Fitzgerald, had more than half their outstanding credit at the Federal Reserve as single-tranche repos (Eisenbeis and Herring 2014). Robert Eisenbeis and Richard Herring suggested that these five primary dealers, as well as Morgan Stanley, RBS, UBS, Deutsche Bank, and Barclays, had a non-negligible amount of ST OMOs outstanding, and thus benefited quite a bit. However, for this group of dealers, the authors of the paper stated that ST OMO “had negligible impact” (Eisenbeis and Herring 2014). Generally, however, there appears to be very limited evaluation on the impact of ST OMO itself, especially compared to TAF, TSLF, and PDCF. In spite of its use as a lender-of-last-resort facility, the small size of the program ($80 billion outstanding at any given time) and standard legal basis may have contributed to a relative lack of evaluation, as well.

Bill English and Trish Mosser discussed some aspects of ST OMO, explaining that, “it was simple to announce and implement and was well-understood by the primary dealers, with no stigma attached to its use” (English and Mosser 2018). Additionally, since it was an FRA Section 14 program and used conventional authority, its implementation was immediate. The authors cited an immediate, verifiable positive impact on one-month repo spreads for agency MBS, which quickly fell after ST OMO and the other primary dealer programs were put in place. Initial demand was quite high for the program but tapered off once other, broader facilities were put in place.See Figure 6 for more information.  Finally, they explained that, due to the comparatively narrow range of usable collateral, the program was not as effective as it could have been in easing term funding market strains (English and Mosser 2018).

Despite a lack of academic attention, in 2011, the program garnered a fair bit of press in the public sphere, most notably in an article written by Bob Ivry. He lambasted the program for being “secretive” and containing transaction-level details such as bank-specific participation, collateral used, and rates paid, that had not been revealed to anyone in the public sphere (Ivry 2011).

David Altig, executive vice president and director of research at the Federal Reserve Bank of Atlanta, challenged these criticisms, stating that he believed the transactions were not secretive, as the FRBNY had issued a press release on March 7, 2008, that specified some of the details of ST OMO. Additionally, Altig argued that the press releases, combined with the overall auction results (which were published) and the fact that the list of primary dealers was readily available, suggested that the program was far from secretive (Altig 2011).

Well-known financial journalist Felix Salmon mostly sided with Altig in the discussion about ST OMO, but he pointed out that Altig’s remarks “made the Fed seem a lot more transparent than it actually is.” Additionally, he criticized the Fed for not bothering to enhance public understanding of the program, explaining that Ivry’s article was the first time that many in the public sphere, including then–Representative Barney Frank, found out about the program (Salmon 2011).

While it is true that, in the summer of 2011, the Federal Reserve released transaction-level data for many of these programs, including ST OMO, it was forced to do so after being sued by Bloomberg reporters in November 2008 under the Freedom of Information Act (FOIA). Only after a 16-month legal battle, combined with an appeal that ended with a ruling against the Fed, did it release transaction-level data. The reluctance could have been due to a lack of desire on the part of the Fed to disclose the information of individual banks that were most vulnerable, to avoid stigma against them (Feuer 2010; Ivry and Keoun 2011).