10

How Risky Were the Fed’s Rescues of Other Firms?

We have seen what Fed policymakers did to ensure Lehman’s bankruptcy and how they might instead have rescued the firm. We can get another perspective on policymakers’ treatment of Lehman by comparing it to their treatment of other financial institutions that faced liquidity crises during 2008, specifically, Bear Stearns, AIG, Morgan Stanley, and Goldman Sachs. The Fed helped these firms avoid bankruptcy with a mixture of liquidity support and long-term financing of assets.

Why did the Fed choose not to rescue Lehman but to rescue these other firms? Fed officials say that Lehman did not have adequate collateral for the loan it needed and that these other firms did. Because of this difference, a Fed rescue of Lehman would have been risky and illegal but the assistance given to others was safe enough to be legal.

Once again, the available evidence does not support the Fed’s position. When the Fed lent to Morgan Stanley and Goldman Sachs, the collateral it accepted was similar to the collateral that Lehman could have pledged for the loan it needed (a range of securities accepted by the Primary Dealer Credit Facility). Therefore, the Fed’s loans to Morgan and Goldman were no safer than a loan to Lehman would have been. In lending to AIG and Bear Stearns, it appears, based on the nature of the collateral and other loan terms, that the Fed took on more risk than it would have in rescuing Lehman – although we cannot be sure for AIG, because the Fed has refused to give a detailed accounting of AIG’s collateral.

Another response of the Fed to the financial crisis was the creation of the Commercial Paper Funding Facility (CPFF) in October 2008. The CPFF bought commercial paper from many corporations at a time when the run on money market funds had disrupted the commercial paper market and made it difficult for firms to obtain working capital. It appears that the Fed’s extension of credit under this program, like the Bear Stearns and AIG loans, was riskier than a Lehman rescue would have been. Fed officials have said the CPFF was protected by insurance premiums on the commercial paper it bought, but the argument behind this claim has major flaws.

Liquidity Support for Morgan Stanley and Goldman Sachs

After the Lehman failure, the investment banks Morgan Stanley and Goldman Sachs experienced large losses of liquidity that threatened them with bankruptcy. The Fed kept the two firms in business with loans from the Primary Dealer Credit Facility and the Term Securities Lending Facility; support peaked in late 2008 at $107 billion for Morgan Stanley and $69 billion for Goldman Sachs. Ultimately, these firms survived with the help of their bank-holding-company status, equity injections from both private investors and the Troubled Asset Relief Program, and the general stabilization of the financial system in 2009.

The Fed facilities accepted the same types of collateral from Morgan Stanley and Goldman Sachs that they accepted from LBI after the LBHI bankruptcy. Some of this collateral was low-quality; for example, at the peak of Morgan Stanley’s borrowing, about $2 billion of its PDCF collateral were securities rated CCC or lower.¹ If Lehman had faced the same collateral standards and had been given free access to the PDCF (without the restrictions on funding LBIE in London), it could have borrowed enough to avoid bankruptcy.

When the PDCF lent to LBI after the bankruptcy of LBHI, the haircuts set on LBI’s collateral were larger than the typical haircuts in the tri-party repo market. By contrast, the PDCF haircuts set for Morgan Stanley and Goldman Sachs on September 14 were smaller than the market haircuts for most collateral types. The haircuts for speculative-grade corporate bonds, for example, were 16.7 percent for LBI, 13 percent in the market, and only 6.5 percent for Morgan Stanley and Goldman Sachs.²

The Bear Stearns Rescue

When Bear Stearns experienced its liquidity crisis in March 2008, the Fed first stepped in to help in the early hours of Friday, March 14. The Fed lent Bear $12.9 billion to meet its obligations that day and Bear pledged $13.8 billion of securities as collateral. That arrangement was similar to the Fed’s overnight lending to investment banks through the PDCF after that facility opened on March 17.³

On Sunday, March 16, the Fed provided a different kind of assistance to Bear by creating Maiden Lane LLC and lending it approximately $29 billion to purchase real estate assets from Bear. Because this transaction appears riskier than the Fed’s overnight lending to Bear and others, it deserves closer examination.

The Maiden Lane Deal

The Maiden Lane transaction is described in a June 2008 report prepared for the New York Fed by the accounting firm of Ernst and Young.⁴ Maiden Lane purchased $30 billion of assets from Bear, based on Bear’s valuations. The assets were selected by agreement of the Fed and JPMorgan Chase. The Fed retained BlackRock Financial Management to manage Maiden Lane’s portfolio, instructing BlackRock to sell the Bear assets over time to maximize returns. It took BlackRock about four years to sell all the assets.

The $30 billion paid by Maiden Lane came from a non-recourse loan of $28.8 billion from the Fed and a subordinated loan of $1.2 billion from JPMorgan Chase. Under this arrangement, JPMorgan would absorb the first $1.2 billion of any losses on Maiden Lane’s assets, and the Fed would absorb the rest. The interest rate on the Fed’s loan was 2.5 percent, which was the discount rate at the time. In June 2012, the Fed reported that its loan had been repaid in full.

According to the Ernst and Young report, the $30 billion of assets purchased by Maiden Lane included $16.8 billion of mortgage-backed securities [MBSs], $8.2 billion of commercial mortgages, $1.6 billion of residential mortgages, and $3.3 billion of credit and interest-rate derivatives. Virtually all the MBSs were investment-grade, and all the mortgages were classified as performing.

In 2010, the New York Fed released a financial statement for Maiden Lane as of December 31, 2008.⁵ This statement broke down Maiden Lane’s assets by the method used to value them. A total of $11.4 billion of assets were valued based on Level 3 inputs, the most subjective method in the hierarchy of the Financial Standards Accounting Board. Maiden Lane’s Level 3 assets included all its holdings of commercial and residential mortgages, and about $2 billion of its mortgage-backed securities.

An Adequate Haircut?

How well-secured was the Fed’s loan to Maiden Lane? Under the deal with JPMorgan, the Fed lent $28.8 billion and received a senior claim on $30 billion in assets. Essentially, the Fed lent to Maiden Lane against $30 billion of collateral with a haircut of $1.2 billion, or 4 percent.

This haircut was significantly lower than those for other Fed loans during the financial crisis. The PDCF, which did most of the Fed’s lending against mortgage assets, imposed haircuts of 6.5 percent on investment grade MBSs and 8.3 percent on whole mortgages.⁶

Moreover, the indefinite term of the Maiden Lane loan made it substantially riskier than PDCF loans, which were overnight. Haircuts for overnight repos are chosen to cover the costs to cash lenders of liquidating collateral and to protect them from one-day changes in collateral values. If collateral depreciates over time, a lender reduces the daily cash it provides to maintain the haircut and avoid losses. In the Maiden Lane deal, by contrast, the Fed’s commitment of $28.8 billion left it exposed to the risk of falling collateral values.

In sum, because of its indefinite term, the Maiden Lane loan was riskier than PDCF lending to Lehman would have been even if the haircuts had been the same. The riskiness of the Maiden Lane loan was increased further by the small haircut set by the Fed.

The Performance of the Maiden Lane Portfolio

The previous section assesses the riskiness of the Maiden Lane loan when the Fed made it, based on the terms of the loan. Another perspective on the risk to the Fed comes from examining what actually happened after the loan was made.

As the financial crisis worsened, default rates on mortgages rose and Maiden Lane’s assets became much riskier. In 2010, the Financial Times produced evidence on this point by analyzing a large sample of MBSs from Maiden Lane’s portfolio.⁷ In April 2008, just after Maiden Lane was created, 93 percent of the MBSs by value were rated AAA, and 99.9 percent were investment grade (BBB or higher). Two years later, only 19 percent were AAA and 28 percent were investment grade. Forty-eight percent were CCC or lower.

As Maiden Lane’s assets became riskier, their values were marked down, and the Fed’s loan became undercollateralized. The Board of Governors website gives weekly figures for Maiden Lane’s assets and for the balance on its loan from the Fed.⁸ Total assets first fell below the loan balance on October 22, 2008. The gap between assets and the debt to the Fed peaked at $3.1 billion on May 6, 2009, when assets were $25.7 billion and the debt was $28.8 billion.

Eventually the financial crisis eased, Maiden Lane’s assets appreciated, and it fully repaid the Fed’s loan in 2012. This outcome does not mean, however, that the loan was safe. The Fed had lost $3.1 billion as of May 2009, and its final loss could have been that much or more if the financial crisis had worsened.

The AIG Rescue

At the insurance conglomerate AIG, as at the big investment banks, losses related to real estate produced a collapse of confidence and a liquidity crisis. The Fed first assisted AIG with an $85 billion line of credit on September 16, just one day after the Lehman bankruptcy. In the following months, the Fed took several other actions to stabilize AIG, including additional loans and the creation of Maiden Lane II and Maiden Lane III.⁹

Fed officials have asserted that their lending to AIG, like all their lending under Section 13(3), was well-secured. However, much of AIG’s collateral consisted of equity shares in the firm’s insurance subsidiaries. These subsidiaries were private companies, which are difficult to value because share prices cannot be observed in financial markets. Fed officials have never presented any evidence on the value of AIG’s collateral relative to the credit extended to the firm. Indeed, they explicitly refused to provide estimates of AIG’s collateral values to the Fed’s Congressional oversight committees, despite a legal mandate to do so.

Under these circumstances, it is difficult to determine the adequacy of AIG’s collateral, but there are pieces of evidence that allow an analysis of this issue.

The Details of the September 16 Line of Credit

In the AIG crisis, as with Lehman, the Fed and the Treasury department initially tried to broker a private-sector rescue. A consortium led by JPMorgan Chase and Goldman Sachs met at the New York Fed on September 15 and prepared a term sheet for a $75 billion line of credit for AIG. The fact that the consortium produced such a document shows that the private sector was close to rescuing the firm. On the morning of September 16, however, the consortium members decided not to go through with the deal. By that point, AIG had run out of cash and faced imminent bankruptcy.

On the evening of September 16, the Board of Governors authorized the $85 billion line of credit from the New York Fed to AIG. The terms were similar to the deal considered by the private-sector consortium, with $10 billion added “as a cushion.”¹⁰ The line of credit had a term of two years and an interest rate of LIBOR plus 850 basis points, a rate much higher than the rates the Fed was charging other financial institutions at the time. In addition, AIG was required to give the Treasury department a 79.9 percent equity interest in the firm. (In the private-sector deal, this equity interest would have gone to the consortium of lenders.)

The collateral pledged by AIG was quite different from the collateral pledged to the Fed for other loans it made. Loans from the PDCF and loans to the Maiden Lane facilities, for example, were collateralized by specific securities and whole loans, and the cash provided by the Fed was based on the value of the collateral and on haircuts it deemed prudent. By contrast, when the Fed announced the September 16 loan to AIG, it said simply:

The loan is collateralized by all the assets of AIG, and of its primary non-regulated subsidiaries. These assets include the stock of substantially all of the regulated subsidiaries.

The “regulated subsidiaries” were primarily insurance companies, such as American Life Insurance Company. Under state insurance laws, AIG could not pledge the assets of these companies, but it could pledge AIG’s equity in the companies themselves. On September 16, New York Fed security personnel went to AIG headquarters and took possession of paper stock certificates for the companies.

The exact structure of the AIG deal was complex. Some of AIG’s subsidiaries (not the ones whose stock was pledged as collateral) guaranteed its obligation to the Fed. Some assets of these subsidiaries were designated as collateral for the $85 billion loan, and other assets were pledged as collateral for the guarantees.

The Safety of the AIG Loan: Policymakers’ 2010 Testimony

In the years following the AIG loan, Fed officials often asserted that the loan was well secured, so there was little risk to the Fed. On this point, as on others, the most detailed discussions appear in the 2010 FCIC testimony of Chairman Bernanke and New York Fed General Counsel Baxter. Both of them emphasized that the loan was collateralized by stock in the insurance companies owned by AIG.

According to Bernanke:¹¹

Unlike Lehman, which was a financial firm whose entire going-concern value was in its financial operations, AIG was the largest insurance company in America. And the Financial Products Division, which got into the trouble, was just one outpost of this very large and valuable insurance company … . So unlike Lehman, which didn’t have any going-concern value, or not very much, AIG had a very substantial business, a huge business, more than a trillion dollars in assets and a large insurance business that could be used as collateral to borrow the cash needed to meet Financial Products’ liquidity demands. So that’s a very big difference. And indeed, the Federal Reserve will absolutely be paid back by AIG.

Later in his testimony, Bernanke reiterates:¹²

[I]t was our assessment that they had plenty of collateral to repay our loan … . [T]he problems with AIG didn’t relate to weaknesses in their insurance businesses, it related very specifically to the losses of the Financial Products Division. The rest of the company was, as far as we could tell, an effective, sound company with a lot of value, and that was the basis on which we made the loan.

In his written FCIC testimony, Baxter supports Bernanke’s view:¹³

Unlike the naked guarantee needed to facilitate the merger of Barclays and Lehman, our committed credit to AIG on September 16, 2008 was fully secured by good collateral, namely, AIG’s sound retail insurance businesses. In fact, before any money was disbursed to AIG on September 16, AIG delivered share certificates to the New York Fed that we continue to hold as collateral in our vaults. These shares fully secured every penny we lent to AIG on September 16, 2008.

The Safety of the AIG Loan: Policymakers’ Statements in 2014–2015

More recently, both Timothy Geithner (who did not testify before the FCIC) and Ben Bernanke have discussed the AIG rescue. Both of them discuss AIG in their memoirs, and in their 2014 testimony in Starr International Co. v. U.S. The Starr case was a lawsuit brought by AIG stockholders claiming that the Fed’s terms for the AIG loan were unduly harsh. (In 2017, a federal Appeals Court ruled against the plaintiffs.)

These assessments of the AIG loan are more equivocal than the FCIC testimony of Bernanke and Baxter. Geithner says in his 2014 memoir that AIG’s insurance companies were “reasonably solid collateral.”¹⁴ In elaborating on this point, he says:

Those insurance businesses would have a good chance of retaining their value if their parent company didn’t go down [emphasis added].

This statement suggests implicitly that the AIG loan would not have been well secured if the company had entered bankruptcy.

At one point Geithner says, “I believed we had gotten taxpayers a reasonable deal,” but elsewhere he says:¹⁵

We would be exposing the Fed to the risk of an imploding insurance company, and there was a real possibility that our loan would simply buy the world time to prepare for a horrific default.

In the Starr lawsuit, the plaintiff’s counsel challenged Geithner to justify the harsh terms of the AIG loan: the high interest rate and the demand for 80 percent equity in the company. In response, Geithner said:¹⁶

I thought we were taking enormous, unprecedented risks, and that there was substantial risk that we would lose billions of dollars, if not tens of billions of dollars.

Geithner also testified, however:

I also believed that there was a reasonable prospect that over time, over a longer period of time, if we were successful in preventing AIG’s failure and if we were successful in averting another global depression, that we had a reasonable chance of recovering our assistance.

In his 2015 memoir, Bernanke restates his 2010 view that AIG had adequate collateral:¹⁷

Unlike Lehman, AIG appeared to have sufficiently valuable assets – namely, its domestic and foreign insurance subsidiaries, plus other financial services companies – to serve as collateral and to meet the legal requirement that the loan be “secured to the satisfaction” of the lending Reserve Bank.

At the same time, Bernanke echoes Geithner in citing risk to justify the terms of AIG’s loan:

Tough terms were appropriate. Given our relative unfamiliarity with the company, the difficulty of valuing AIG FP’s [Financial Products’] complex derivatives positions, and the extreme conditions we were seeing in financial markets, lending such a large amount inevitably entailed significant risk. Evidently, it was risk that no private-sector firm had been willing to undertake. Taxpayers deserved adequate compensation for bearing that risk. In particular, the requirement that AIG cede a substantial part of its ownership was intended to ensure that taxpayers shared in the gains if the company recovered.

Bernanke also says:¹⁸

If the loan to AIG helped stabilize financial markets, then AIG’s companies and assets would likely retain enough value to help repay the loan over time. But if financial conditions went from bad to worse, driving the economy deeper into recession, then the value of AIG’s assets would suffer as well. And, in that case, all bets on being repaid would be off. We had to count on achieving the better outcome.

The tone of these comments differs from Bernanke’s FCIC testimony (AIG “had plenty of collateral”) and Baxter’s (the collateral “fully secured every penny we lent”). Why the difference? One possible factor is the contexts in which statements were made. In their FCIC testimony, Bernanke and Baxter were defending the Fed against charges that the AIG loan was excessively risky, so they had an incentive to minimize the risk. By contrast, in his Starr testimony, Geithner was replying to the plaintiff’s claim that the Fed had imposed “extortionary” terms on AIG. He needed to emphasize the risks to the Fed to justify the terms. Starr may also have influenced the discussions of AIG in Geithner’s and Bernanke’s memoirs.

The Fed’s Opacity about Collateral Values

How well secured was the AIG loan? We could better answer that question if we knew the value of the firm’s collateral. Unfortunately, we know little about AIG’s collateral, in part because the Fed has resisted requests for information about it.

Under Section 129 of the Emergency Economic Stabilization Act of 2008, the Fed’s Board of Governors must report to its Congressional oversight committees each time it invokes its Section 13(3) authority to make a loan. Each report must include a justification for the loan, a description of its terms, and “available information concerning the value of any collateral held with respect to such a loan.”

The report for the September 16 loan to AIG says, “the Board does not believe the authorization of the Credit Facility will result in any net cost to taxpayers.”¹⁹ The report does not, however, provide any information about the value of AIG’s collateral, despite the law saying it must do so. Indeed, the report explicitly declines to estimate the collateral’s value, based on an unusual argument that doing so could impede repayment of the loan:²⁰

In light of the complexities involved in valuing the extremely broad range of collateral and guarantees securing all advances under the Credit Facility [AIG’s line of credit], the Board believes any estimate at this time of the aggregate value that ultimately will or may be received from the sale of collateral or the enforcement of the guarantees in the future would be speculative and could interfere with the goal of maximizing value through the company’s global divestiture program and, consequently, the proceeds available to repay the Credit Facility.

In 2012, I submitted a Freedom of Information Act (FOIA) request to the Board of Governors for several documents, including “a list of the specific assets pledged as collateral for the September 2008 loan to AIG and the value of each asset as determined by the Federal Reserve.” The Board searched its records and found a document “responsive to the request,” but declined to release it. The Board cited Exemption 8 of FOIA, which covers documents related to “the regulation or supervision of financial institutions.”

I appealed this decision to federal district court, where I was represented by the Public Citizen Litigation Group.²¹ In the course of this litigation, the Board described the document responsive to my FOIA request as follows:

Spreadsheet listing specific certificated and uncertificated securities and instruments, in particular, stock, promissory notes, and membership interests issued by certain AIG subsidiaries, delivered by AIG to the FRBNY as a portion of the collateral for the FRBNY’s extension of up to $85 billion in credit to AIG (the AIG Revolving Credit Facility).

Unfortunately, the judge in the case upheld the Board’s decision to withhold this document under Exemption 8.²²

Evidence on the Security of the September 16 Loan

I have not found an estimate of the value of AIG’s collateral in any public source. I have, however, found fragments of evidence concerning the adequacy of the collateral for securing the Fed’s $85 billion loan. This evidence is far from conclusive, but it casts doubt on the security of the loan.

The Views of Financial Institutions The consortium led by JPMorgan Chase and Goldman Sachs decided not to lend $75 billion to AIG, even with the high interest rate and equity stake that the Fed received. One possible reason is that consortium members thought AIG’s collateral was inadequate.

There is some evidence to support this conjecture. Two official documents report the views of consortium members: a 2009 report of the Special Inspector General for TARP (SIGTARP), and a 2010 report of the Congressional Oversight Panel (COP) for TARP. The SIGTARP report cites “a JPMorgan vice chairman” as reporting:²³

The group developed a loan term sheet, but an analysis of AIG’s financial condition revealed that liquidity needs exceeded the valuation of the company’s assets, thus making the private participants unwilling to fund the transaction.

The COP report says:²⁴

One bank that participated in the private-sector rescue effort told the Panel that the banks also concluded that AIG did not have adequate collateral to support the necessary loan.

Fed officials have disputed these accounts. The SIGTARP report, after relating the view of the JP Morgan executive, continues:

FRBNY officials told SIGTARP that, in their view, the private participants declined to provide funding not because AIG’s assets were insufficient to meet its needs, but because AIG’s liquidity needs quickly mounted in the wake of the Lehman bankruptcy and the other major banks decided they needed to conserve capital to deal with adverse market conditions.

The COP report includes a similar claim by Fed staff.²⁵ It is difficult to know who is right without more information.

Analysis at the New York Fed It appears that New York Fed staff analyzed the value of AIG’s collateral in the days before the loan. On September 14, Assistant Vice President Alejandro LaTorre circulated a memo titled “Pros and Cons of Lending to AIG” to colleagues including President Geithner.²⁶ Attached to the memo was a presentation about AIG’s insurance companies prepared on September 13 by the New York Fed’s Bank Supervision Group. In these documents, discussions of AIG’s collateral are cryptic, but the tone is negative.

LaTorre’s memo lists “pros” of lending to AIG, which include various financial disruptions that would occur if the firm failed, and “cons” including moral hazard and concern that “lending to AIG could be perceived as inconsistent with treatment of Lehman.” Point 5 in the list of cons concerns AIG’s insurance companies:

5.Assets available from Ins. Co. subs [subsidiaries] may not be sufficient to cover potential liquidity shortfalls as many of the subs do not appear to be sources of strength.

→Life Ins. Co. subs have significant unrealized losses on investments.

→P&C [property and casualty] could be source of strength; paid $1.4B dividends, but amounts small relative to size of hole.

The presentation from the Bank Supervision group is titled, “AIG Subsidiaries: Are they a Source of Strength?” The first slide after the title is headed “Ability to support has weakened,” and appears to be the source of LaTorre’s comments. The slide mentions unrealized losses on investments and says “dividends to parent down sharply overall,” giving figures of $4.9 billion for 2007 and $1.4 billion for 2008 year-to-date.

Another slide in the presentation, titled “What happens in a sale?,” mentions risks to the solvency of AIG’s insurance companies:

Using a weighted average, all the subsidiaries shocked would wipe out their capital in a liquidation if assets are sold at a 12 percent loss or greater.

Maiden Lanes II and III

Following the September 16 $85 billion loan, the Fed took several other actions to assist AIG. On October 6, it lent $38 billion to the firm through a Securities Borrowing Facility. The Commercial Paper Funding Facility created on October 7 (discussed below) bought $16 billion of AIG’s commercial paper. On November 10, the Fed aided AIG by creating Maiden Lane II and Maiden Lane III and lending them $20 billion and $24 billion respectively. The total of all these commitments was $183 billion.²⁷

This section focuses on the loans to the two new Maiden Lanes, which helped AIG increase its liquidity and decrease risk. Maiden Lane II bought $21 billion of AIG’s illiquid mortgage-backed securities, using the $20 billion loan from the Fed and a $1 billion subordinated loan from AIG. Maiden Lane III bought $29 billion of collateralized debt obligations (CDOs) from counterparties of AIG, using the $24 billion Fed loan and $5 billion from AIG. The purchases by Maiden Lane III allowed AIG to terminate credit default swaps tied to the CDOs, which were risky derivatives positions.

We can interpret these two arrangements, like the first Maiden Lane that assisted Bear Stearns, as long-term loans from the Fed against illiquid collateral. The subordinated loans from AIG were equivalent to haircuts on the collateral: about 5 percent for Maiden Lane II and 17 percent for Maiden Lane III. Despite these haircuts, falling asset prices meant the Fed’s loans eventually became undercollateralized when the assets of Maiden Lanes II and III fell below their debts to the Fed. These shortfalls peaked in mid-2009 at $2.1 billion for Maiden Lane II and $2.7 billion for Maiden Lane III.²⁸ Once again, however, the assets’ prices recovered as the financial crisis eased and the Fed’s loans were repaid in full in 2012.

The Commercial Paper Funding Facility

After the Lehman bankruptcy caused the run on money market funds on September 17–18, US corporations found it difficult to issue commercial paper. To address this problem, the Fed established the Commercial Paper Funding Facility (CPFF) on October 7. In this case (unlike any other credit extension under Section 13(3)), the public record includes a memo from the Board’s General Counsel, Scott Alvarez, on the legal justification for the action.²⁹ The memo argues that lending through the CPFF met the legal requirement for satisfactory security, but once again the reasoning of Fed officials is unsound.

The CPFF borrowed money from the New York Fed and used it to purchase commercial paper from corporations around the country. The purchased commercial paper served as the CPFF’s collateral for its loan from the Fed. This arrangement was economically equivalent to direct purchases of commercial paper by the Fed, as Alvarez’s memo acknowledges.³⁰ At the end of 2008, the CPFF owned $335 billion of commercial paper, about 20 percent of all commercial paper issued by US corporations.³¹

The CPFF purchased only commercial paper with the highest rating, A1/P1/F1. That rule was not very restrictive, however, because 90 percent of commercial paper had that rating. In fact, LBHI’s commercial paper had that highest rating until its bankruptcy. About a third of the commercial paper bought by the CPFF was asset-backed and the rest was unsecured, and most of the unsecured commercial paper was issued by financial institutions.³² I focus on the Fed’s financing of unsecured commercial paper, which is relatively easy to analyze.

When the CPFF purchased unsecured commercial paper, it required the issuer to pay an “insurance fee,” in addition to the interest on the commercial paper. The fee was fixed at 100 basis points per year. For ninety-day commercial paper, the fee was approximately 0.25 percent of the security’s face value. General Counsel Alvarez’s memo argues that an insurance fee is one acceptable form of security under Section 13(3). It also argues that a 100 basis point fee was adequate to protect the CPFF from losses. I will not question the general point about insurance fees, but the justification for the level of the fee is weak.

The memo says the 100 basis point fee “is designed to be an insurance premium based on historical loss rates for A1/P1/F1 CP.” It argues:³³

Like an insurance company or fund, these premiums (along with any earnings on the CP in the CPFF SPV [special purpose vehicle]) would serve as a source of funds to repay the CPFF SPV’s extension of credit from the Reserve Bank if losses result from the CP. Also, like an insurance company or fund, the pool of premiums would be available to offset any losses, that is, the premium paid by one issuer would be available to offset losses of other issuers, not just losses from the issuer paying the premium. Mutualization of losses is an important and defining characteristic of an insurance company or fund. Moreover, the aggregate premiums retained in the CPFF SPV were computed to cover those expected losses over the expected life of the facility. Historically, the default rate of A1/P1/F1-rated CP is very low.

The flaws in this reasoning are egregious. In October 2008, when the CPFF was established, the level of distress in financial markets was far greater than at any other time since the 1930s. Historical loss rates surely understated the risk on commercial paper, especially commercial paper issued by financial institutions.

In addition, the analogy to mutualization in insurance is inappropriate because of the correlation of risk across different commercial paper issuers. In their 2010 article on commercial paper during the financial crisis, Marcin Kacperczyk and Philipp Schnabl state the obvious: “diversification reduces exposure to idiosyncratic risk but cannot reduce exposure to systematic risk which affects all commercial paper issuers at the same time.”³⁴ A worsening of the financial crisis might have produced defaults on a significant fraction of commercial paper, with losses to the CPFF that greatly exceeded the insurance fees it received.

In justifying the CPFF and the rescues of Bear Stearns and AIG, Fed officials stretched to portray their lending as well secured. They made claims that are not supported by the available evidence, and their economic reasoning was unsound. The loans that officials chose to make carried significant risk, contrary to their denials.

In important respects, officials’ actions and arguments in these cases are the opposite of their treatment of Lehman Brothers. The top Fed officials have stretched to make the case that a loan to Lehman could not have been well secured, despite clear evidence that this position is wrong. The next two chapters ask why Lehman was treated so differently from other financial institutions, including one (Bear Stearns) that needed a rescue before Lehman did and others (such as AIG) that needed rescues after Lehman did.