Case Studies in Liquidity Risk: Countrywide Financial

This is the third case study in liquidity risk, following earlier blogs on AIG and Bank of America.  Countrywide Financial was at the heart of the 2007-2009 credit crisis and the focus of these earlier blogs on the credit crisis:

van Deventer, Donald R. “Lessons from Failures in Silo Risk Management: Countrywide Financial Corporation,” Kamakura blog, www.kamakuraco.com, May 18, 2009. Redistributed on www.riskcenter.com on May 20, 2009.

van Deventer, Donald R. “Lessons from Failures of Silo Risk Management: Countrywide Financial, Update 3,” Kamakura blog, www.kamakuraco.com, June 8, 2009.  Redistributed on www.riskcenter.com on June 9, 2009

This blog features the funding short-falls at Countrywide Financial Corporation during the credit crisis for a very important reason: firms that have not had a “near death experience” are highly unlikely to be able to measure what would happen to them in a credit crisis using their own data alone.  As we will see in what follows, even with massive borrowings from the Federal Home Loan Bank (see blogs above) and even after announcing a merger with Bank of America, Countrywide would not have survived without borrowings from the Federal Reserve.

Under the Dodd-Frank Act of 2010, the Board of Governors of the Federal Reserve was required to disclose the identities and relevant amounts for borrowers under various credit facilities during the 2007-2010 financial crisis.  These credit facilities provide perhaps the best source of data about liquidity risk and funding shortfalls of the last century.  This data is available for purchase from Kamakura Corporation and is taken from the Kamakura Risk Information Services Credit Crisis Liquidity Risk data base.

The data used in this analysis is described in more detail below. The data consists of every transaction reported by the Federal Reserve as constituting a “primary, secondary, or other extension of credit” by the Fed. Included in this definition are normal borrowings from the Fed, the primary dealer credit facility, and the asset backed commercial paper program. Capital injections under the Troubled Asset Relief Program and purchases of commercial paper under the Commercial Paper Funding Facility are not included.

This blog entry lists the borrowing by Countrywide that was identified by the Fed as “primary, secondary, or other extensions of credit” between February 8, 2008 and March 16, 2009.  Borrowings under this facility clearly represent a funding shortfall and an excellent measure of the degree to which liquidity was insufficient. A number of key events for Countrywide Financial are noted in this recent blog:

van Deventer, Donald R. “A Credit Crisis Chronology Part 1 Through February 2008: This Time Isn’t Different,” Kamakura blog, www.kamakuraco.com, May 13, 2011.

Key dates for Countrywide Financial are excerpted here:

April 30, 2006	In an April memo discussing Countrywide’s issuance of subprime 80/20 loans, which are loans that have no down payment and are comprised of a first loan for 80% of the home’s value and a second loan for the remaining 20% of value, resulting in a loan to value ratio of 100%, Countrywide CEO Angelo Mozilo wrote “In all my years in the business I have never seen a more toxic pr[o]duct.” (Source: Levin report, page 232).
July 24, 2007	Countrywide Financial announces 33% drop in second quarter profits and says subprime mortgage problems were spreading to conventional home loans (Source: www.ft.com).
August 16, 2007	Countrywide taps $11.5 billion commercial paper back up line (source: Bloomberg)
August 24, 2007	Bank of America buys $2 billion in Countrywide Financial preferred stock (Source: www.ft.com)
October 26, 2007	Countrywide reports first loss in 25 years, losing $1.2 billion in the third quarter (Source: www.ft.com)
January 11, 2008	Countrywide announces sale to Bank of America (Source: Levin report, page 47).
July 1, 2008	Countrywide acquisition closes (Source: Wall Street Journal)

The primary, secondary, or other extensions of credit by the Federal Reserve to Countrywide during the period February 8, 2011 to March 16, 2009 can be summarized as follows:

Borrowing dates:	Continuous from March 24, 2008 through July 1, 2008, the closing of the acquisition by Bank of America, with first borrowings from March 17-18, 2008
Average from
2/8/2008 to 3/16/2009	$574.8 million
Average when Drawn	$2.27 billion
Maximum Drawn	$6.3 billion

Kamakura Risk Information Services version 5.0 Jarrow-Chava default risk models showed a cumulative 5 year default probability for Countrywide Financial Corporation of 33.75% on January 14, 2008, the first business day after the announcement of the acquisition of Countrywide Financial Corporation.  Obviously, this is a very high level of distress for Countrywide.

By the beginning of Countrywide’s first borrowings from the Fed, which started on March 17, 2008, Countrywide Financial’s cumulative 5 year default probability was 34.69%, still very highly distressed:

The borrowings from the Federal Reserve start on March17-18 with an initial borrowing of $1.5 billion, indicating that (even after the January 11 announcement of the acquisition by Bank of America) Countrywide could no longer fund itself completely. From March 24 to the closing of the merger on July 1, Countrywide liquidity risk generated very erratic borrowing patterns, averaging $2.27 billion with a peak of $6.3 billion, nearly 50% of the amount that ultimate acquirer Bank of America had to borrow from the Fed during its peak point of distress:

In addition to these borrowings (“primary, secondary or other extensions of credit” from the Federal Reserve), Bank of America sought support from the Commercial Paper Funding Facility run by the Federal Reserve on one occasion, but the Fed reported no borrowings from Countrywide during this time, primarily because Countrywide succeeded in pushing as much business as possible into its banking subsidiary where there was the potential to borrow from the Federal Home Loan Bank system and the Fed.

Implications of Funding Shortfall Data

The sports truism that “it ain’t over until the fat lady sings” was never more true than in the case of Countrywide.  As one member of the Countrywide Board of Directors told me after the acquisition was announced, “We’re just trying to survive to the closing.”  Both the KRIS default probabilities and the need to borrow from the Federal Reserve AFTER the acquisition was announced show that Countrywide barely escaped failure.  In our next study of liquidity risk, we turn to another Bank of America acquisition, Merrill Lynch.

Background on the Federal Reserve Data

A summary of the Federal Reserve programs that were put into place and summary statistics are available from the Federal Reserve at this web page:

http://www.federalreserve.gov/newsevents/reform_transaction.htm

Today’s blog focuses on one set of disclosures by the Federal Reserve: primary, secondary and other extensions of credit by the Fed.  This includes direct, traditional borrowings from the Federal Reserve, the primary dealer credit facilities, and the asset backed commercial paper program described at the link above.  These borrowings do not include commercial paper purchased under the Commercial Paper Funding Facility nor do they include the equity stakes taken by the U.S. government under the Troubled Asset Relief Program.

Kamakura took the following steps to consolidate the primary, secondary and other extensions of credit:

• From www.twitter.com/zerohedge Kamakura downloaded the daily reports, in PDF format, from the Federal Reserve on primary, secondary and other extensions of credit from February 8, 2008 until March 16, 2009, approximately 250 reports in total
• Kamakura converted each report to spreadsheet form
• These spreadsheets were aggregated into a single data base giving the origination date of the borrowing, the name of the borrower, the Federal Reserve District of the borrower, the nature of the borrowing (ABCP, PDCF, or normal), the maturity date of the borrowing, and (in the case of Primary Dealer Credit Facility) the name of the institution holding the collateral.
• Consistency in naming conventions was imposed, i.e. while the Fed listed two firms as “Morgan Stanley” and “M S Co” Kamakura recognized to the maximum extent possible that they are the same institution and used a consistent name
• To the maximum extent possible, the name of the ultimate parent was used in order to best understand the consolidated extension of credit by the Fed to that firm.

For information regarding the Kamakura Credit Crisis Liquidity Risk data base, please contact us at info@kamakuraco.com.  Please use the same e-mail address to contact the risk management experts at Kamakura regarding how to simulate realistic liquidity risk events in the Kamakura Risk Manager enterprise-wide risk management system.

Donald R. van Deventer
Kamakura Corporation
Honolulu, Hawaii
May 18, 2011