Primary Dealer Credit Facility (US)

Specifically, it was established under the Fed’s emergency authority under Section 13(3) of the Federal Reserve Act as a lender in “unusual and exigent circumstances.”

Under the PDCF, the Federal Reserve became a potential lender to primary dealers. To manage the accompanying risks, the Federal Reserve positioned analysts at each of the major independent primary dealers, thus creating a monitoring program to improve lines of communication between the Federal Reserve and the dealers. In July, the Federal Reserve entered a Memorandum of Understanding (MOU) with the Securities and Exchange Commission (SEC) that affirmed the SEC as the ultimate supervisor of the investment banks. This was intended to show concerned individuals and government agencies that the Federal Reserve had not unduly expanded its oversight powers through the PDCF (Geithner 2010; Adrian, Burke, and McAndrews 2009).

At the creation of the PDCF, the Federal Reserve announced its intention to maintain the program for six months. The hope was that by then, fear in the markets would have abated to a sufficient level and that primary dealers would have been able to arrange other methods of financing. However, the Fed cautiously kept an open-ended timeline, noting that the program “may be extended as conditions warrant to foster the functioning of financial markets.” In the end, the program was extended a total of four times before being terminated in February 2010.

The Bear Stearns crisis highlighted the importance of investment banks to the liquidity of securitization markets. Even though the Fed had previously taken steps through its OMOs program and the TSLF to address the stresses in the short-term wholesale funding markets, there was concern that these vehicles might not fill any gap in liquidity that might occur. The TSLF was announced first but was not yet operational. More information on TSLF can be found in Leon Hoyos’s Term Securities Lending Facility Case (Leon Hoyos 2019). Establishment of the PDCF reflected a determination that in the brewing situation, and given the critical role of the primary dealers, lender-of-last-resort financing needed to be extended beyond commercial banks, allowing the Fed to provide short-term funding directly to investment banks as well.

The PDCF utilized its existing operational relationships with the primary dealers and the tri-party repo system that it used for its OMO repo operations and in which JP Morgan Chase & Co. and Bank of New York Mellon were the two tri-party repo clearing banks (GAO 2011). Affiliates of foreign banks, such as Barclays, BNP Paribas Securities, Daiwa Securities America, Deutsche Bank Securities, and UBS Securities LLC, that operated in the US as primary dealers were eligible to utilize the PDCF and did (Reuters 2010). The disclosure of just how much these foreign-related entities borrowed from the facility became a point of contention in public perception (Reuters 2010). However, on an aggregate basis, dealers with European parents were often heavy borrowers from the TSLF but very light borrowers from the PDCF. RBS, for example, borrowed from the TSLF on 57 occasions but never borrowed from the PDCF (Achayra et al. 2014).

The Fed included several design features that were intended to mitigate the risk inherent in the PDCF. These included having collateral valued by the clearing banks at the least available value and applying haircuts.

For collateral eligible in OMOs, haircuts assigned were equivalent to haircuts under the open-market operations. For collateral not eligible in OMOs, haircuts were determined by the asset’s risk and were generally higher than those under OMOs. However, the Fed’s haircuts were less than market haircuts would have been during the crisis, thus providing to borrowers more funding against a particular collateral than they might have received elsewhere (Adrian, Burke, and McAndrews 2009).

The lending rate was equivalent to the discount window’s primary credit rate at the FRBNY, a standard that reinforced the Fed’s intent to have the PDCF operate similarly to the discount window and to make the funding financially accessible. The primary credit rate is the discount window’s most favorable rate. 

However, under normal conditions, the discount rate exceeds the overnight repo rate for most eligible securities, with the result that the PDCF would not be an especially attractive means of financing an inventory of securities in normal market conditions. This meant that the eligible borrowers would be incentivized to use the PDCF only as a backstop, not as a primary funding source, as markets returned to normal levels. (Adrian, Burke, and McAndrews 2009). Additionally, the Fed actively counseled borrowers to seek funding in the markets before utilizing the PDCF (Adrian, Burke, and McAndrews 2009).
 

The FRBNY incorporated a frequency-based penalty fee to primary dealers who accessed the facility more than 45 days out of the preceding 180 business days, so as to discourage usage when other means of funding were available. The fee would increase as the primary dealer continued to access the facility past the 45-business-day mark:

(i)        First 45 days: no fees

(ii)       46-90 days: 10 basis points, annualized rate

(iii)      91-135 days: 20 basis points, annualized rate

(iv)      136-180 days: 40 basis points, annualized rate

We have not located any information that suggests that any borrower paid the frequency-based fee.

Initially, the PDCF was restricted to collateral eligible for OMOs, investment-grade securities, municipal securities, and asset-backed securities (ABS), including mortgage-backed securities. In September 2008, accepted collateral was expanded to include all types of collateral eligible in the tri-party repurchase agreement system, which included noninvestment-grade securities. This potentially provided access to a greater amount of funding, since a primary dealer could now bring a wider range of collateral to the PDCF if it could not finance it in the market.

The PDCF terms did not impose a borrowing limit on the size of loans that primary dealers could take but allowed the borrowing dealer to choose the size of their loan request. The amount of money available to a primary dealer under the PDCF was thus limited by the amount of haircut-adjusted eligible collateral that each primary dealer could present to its clearing bank. However, although not limited as to amount, loans were subject to a frequency fee. To avoid stigma and encourage participation, few details regarding PDCF participants were published at time of use.

In creating the PDCF, the Fed was worried that primary dealers would not take advantage of this opportunity due to stigma, a serious concern that threatened to inhibit the facility’s effectiveness and one that the Fed had previously encountered with respect to its Discount Window (The Wall Street Journal 2008). If a bank was particularly weak and in need of a loan but was the only bank known to be in that situation, the bank might be reluctant to borrow, despite need, for fear of identification. Even if the Fed didn’t publish the names of its borrowers, creditors might surmise which bank the user was, creating a negative reputation around that particular bank. Thus, PDCF activity was publicly reported only in aggregate on a weekly and monthly basis. It was not until December 2010 that the Fed released transaction details, including participants and their individual loan amounts, after Congress mandated such disclosures.

In September and November 2008, the Federal Reserve extended credit to London-based affiliates of three investment banks that were becoming bank holding companies and to Citigroup, respectively. Although not formally PDCF loans, the loans to these affiliates were under terms similar to those of the PDCF, with a few differences. With respect to these “London loans,” however, the Fed accepted collateral denominated in foreign currencies from the London-based affiliates, and it applied higher haircuts to this collateral (GAO 2011). See the Appendix for more discussion of the London loans.