Extending Safe Harbor to Swaps #1-4

Through the 1980s the market in a new type of derivative called a swap was growing.  The first swap was a currency swap:  IBM was receiving regular payments in Swiss francs which it needed to convert to dollars, while the World Bank was receiving dollar payments and had obligations in Switzerland.  The two parties contracted to exchange their Swiss francs for dollars and vice versa.

The most common swap contract both in the 80s and today is the interest rate swap, where one party pays a floating interest rate (which can change and is based on a money market rate) on a regular basis over a period of years and receives regular payments from the other party based on a fixed rate that is set at the start of the contract.  Such contracts are very useful for banks that have sold fixed rate loans and wish to protect themselves against the possibility that the money market rate rises.

The defining characteristic of a swap contract is the trade of a fixed stream of payments for a stream of payments that will be determined by future prices or events in financial markets.  Swaps were considered a new category of financial contract that did not fall directly under the jurisdiction of any of the existing self regulatory organizations supervised by the SEC.

Instead of extending the purview of one of the existing self regulatory organizations to cover swaps, a consortium of dealer banks chose in 1985 to create a new trade group to coordinate swap policy, the International Swap Dealers Association or ISDA.  Thus, through the 1980s the swaps market was unregulated:  they were not, for the most part, considered securities by the SEC, and the CFTC allowed them to fall through the forward contracts exemption of the Commodities Exchange Act.  Their unregulated status was confirmed in 1993 by formal CFTC exemption from regulation. [1]

The repo amendment to the Bankruptcy Code had established the principle that the imprimatur of a self regulatory organization was not necessary for a financial contract to receive safe harbor; exemption from bankruptcy was possible even for unregulated financial contracts.  Thus when the financial industry sought safe harbor privileges for swaps contracts they did not seek the status of a national securities association for the ISDA – which would have granted swaps the same privileges as other securities under the 1982 amendment.[2] Instead the ISDA lobbied Congress to pass an exemption from the Bankruptcy Code specifically for swaps.  The ISDA and other lobbyists argued that the law favored exchange traded derivative contracts by granting them privileges that were not available to over the counter derivatives like swaps.[3] In short, they called for unregulated financial derivatives to receive the same treatment as regulated financial derivatives.

In 1990 the swaps amendment to the Bankruptcy Code was passed.  As with the repo amendment the safe harbor protection for swaps employed extremely broad language:  the amendment extended to any contractual right arising “under common law, under law merchant, or by reason of normal business practice, whether or not evidenced in writing.”[4] From 1990 on swap market participants were, thus, permitted to seize the margin posted by a bankrupt counterparty, to sell off the counterparty’s positions and to set off the value of the bankrupt’s margin against the obligations owing to the swap participant.  The broad terms of the amendment granted a large class of financial contracts special status in bankruptcy.

After the 1990 amendment was passed, the swaps market grew steadily for the next decade and a half.  In 2008 the dollar amount of interest rate and currency swaps had grown to more than 400 times what it had been in 1987. [5] The swaps market also grew as new products were developed that simulated the returns of tradition financial contracts.  Credit default swaps are synthetic bonds:  one party receives regular interest payments in exchange for paying out the value of an underlying bond if it goes into default.  These swaps can function as bond default insurance, since the other side of this trade makes regular payments in exchange for a guarantee that the face value of the bond will be paid even after the bond goes into default.  Equity swaps are synthetic stocks:  one party receives a payment (or periodic payments) based on the performance of a stock or stock index and makes payment(s) based on the money market interest rate.  The other equity swap counterparty then gets the return of a short seller, he will lose the value of any increase in the value of the stock and gets the proceeds from investing the cash gained by selling borrowed stock in a money market account.  Total return swaps are similar but can be used to simulate the returns of any asset.

While these swaps are often called synthetic assets, there is an important difference between a synthetic asset and a real asset:  real assets require an upfront payment to purchase the asset, whereas synthetic assets have no upfront payment and involve instead a promise to make payment in the future.  Thus while an investor in a real bond must come up with, say $10,000, to purchase the bond, the investor in a synthetic bond pays nothing unless the underlying bond defaults – at which point the investor will have to pay the $10,000.  In theory, then, synthetic assets are infinitely leveraged investments (that is, the investor puts no money down at all).  In practice every investor – with the exception of the dealer banks – is required to post margin when entering into a synthetic investment and this reduces, but by no means eliminates, the degree of leverage involved in the transaction.  It remains unclear to what degree the dealer banks post margin on their trades – which leaves open the possibility that their positions in the swaps market are extremely leveraged.

Another concern raised by swap market developments is that when one considers the variety of swaps that have been created to date – currency swaps, interest rate swaps, credit default swaps, total return swaps, equity swaps – one realizes that there are few financial contracts that cannot fit under the umbrella term “swap”.  It is true that traditional investments like stocks, bonds and mortgages require a large upfront payment, rather than the large tail end payment that is common in the swaps market;  however, given that extremely asymmetric payments are an integral part of the swaps market, it’s far from clear why the time sequence of asymmetric payments should be sufficient to preclude the reclassification of traditional investment vehicles as swaps.  Furthermore, if the distinction is that with stocks and bonds one party has no future obligation to make payments, then the simple expedient of either allowing settlement to take place one week after the trade date or of offering to sell the stock or bond on an installment basis (adjusted, of course, by the money market rate) will turn the traditional asset into a swap.[6]

Given the fact that the overarching term “swap” can apparently encompass most financial contracts, the legal definition of a “swap” is a matter of some importance.  This definition was amended by the Bankruptcy Abuse Prevention and Consumer Protection Act of 2005, discussed below.


[1] However in a regulatory action in 1994 against Bankers Trust Securities Corp. the SEC did claim that some swaps were securities.  This position was very much the exception rather than the rule.  (Carolyn Jackson, 1999, “Have you hedged today?” 67 Fordham Law Review, p. 3223)

[2] The definition of a security in the Securities Acts of 1933 and 1934 is broad enough to include swaps, if the swaps dealers had wanted the designation.  (Olander and Spell, “Interest rate swaps,” 45 Maryland Law Review, 1986, p. 53)  However, since that would subject them to regulation by the SEC, they argued that swaps were not securities and the Gramm-Leach-Bliley Act  (2000) explicitly amended the Securities Acts to exclude swaps from the definition of securities.

[3] “Over the counter” refers to trade that takes place in a decentralized manner and is coordinated by dealers.  Karen Ramdhanie, “Derivatives contracts of insolvent companies,” 18 New York Law School Journal of International and Comparative Law, 1999, p.299

[4] Hance, p. 746.

[5] Data from ISDA http://www.isda.org/statistics/pdf/ISDA-Market-Survey-results1987-present.xls.

[6] Note that even spot transactions have a settlement several days in the future, so all financial transactions are in some sense forward transactions.  Furthermore, the definition of swaps contract has always explicitly included spot foreign exchange agreements.  The text of the law with revisions indicated is available here:  http://www.law.ttu.edu/lawlibrary/library/research/BAPCPA_Library/code-01.htm#101.50

Extending Safe Harbor to Repos #1-3

The first unregulated financial contract to be granted safe harbor was the repurchase agreement (or repo).  Two parties enter into a repurchase agreement when a security is sold and simultaneously the parties agree to reverse the transaction at a specific future date and price.  Thus, a repo transaction is comparable to a secured loan; in fact, repo contracts typically include a mark-to-market clause requiring additional margin if the value of the security falls below the value of the loan.  On the other hand, the repo transaction is like a sale, because the buyer-lender has complete control over the collateral including the right to sell it on to somebody else:  this process is called rehypothecation.  The buyer-lender’s only obligation is to replace the rehypothecated collateral with an equivalent security by the date of the repurchase contract.

Repos were used by the New York Federal Reserve Bank starting in the early years of the 20th century and also by other firms that faced restrictions on secured lending.  Since the 1950s the Federal Reserve used repurchase agreements as a principal tool when implementing monetary policy.  For this reason, the repo market was a crucial concern of the Fed.

The private sector repo market did not grow large until the 1970s when short-term interest rates rose to unprecedented levels.  This phenomenon drove firms and local governments – which often were not authorized to lend – to place funds in the repo market as way of earning interest income.[1] The use of repos by the private sector grew from $2.8 billion of outstanding repos in 1970 to $45 billion by 1979 to hundreds of billions in 1983.[2]

The problem with repos in the 70s and early 80s was that there was a strong likelihood that they would be considered secured loans under the terms of Article 9 of the Uniform Commercial Code, which applies to “any transaction (regardless of form) which is intended to create a security interest in personal property.”[3] If repos are secured loans, then in bankruptcy they are subject to the automatic stay:  in other words, the counterparties to a bankrupt firm cannot dispose of the collateral they hold as it is the property of the bankrupt estate.

In 1982 when a small government securities dealer, Lombard Wall, declared bankruptcy, the bankruptcy judge found that Lombard Wall’s repo contracts were secured loans.  While the status of repos had been uncertain before, market participants in the past had held out the hope that “friend of the court” briefs from the Federal Reserve and major investment banks would convince bankruptcy judges not to risk disrupting this huge market.  Lombard Wall made it clear that some judges were not swayed by these arguments.   Buyer-lenders in the repo market risked finding that collateral that had been pledged by a bankrupt firm had suddenly become a frozen asset, and, if the price of the collateral moved quickly, this illiquidity could lead to losses.[4] Lombard Wall, however, was a small affair and the collateral was released within days, so it did not involve significant losses.

Whether this ruling had any quantitative effect on the repo market is very difficult to determine, because it followed less than three months after the failure of Drysdale Government Securities – a large bankruptcy which made it clear that contemporary repo market practices could be gamed and which led to major changes in the market.  Thus, while there is evidence that the rate of growth of the repo market slowed between June 1982 and May 1983,[5] it is not at all clear that this slowdown can be attributed to the Lombard-Wall ruling.

In the absence of safe harbor protections for repurchase agreements, buyer-lenders were exposed to more credit risk than they would be if they controlled the collateral free and clear.  The repo dealers believed that the best solution to this state of affairs was for Congress to amend the Bankruptcy Code and expand the safe harbor provisions to repo contracts.  In their pursuit of a legislative solution the dealers had an impressive ally: the Federal Reserve.

In 1984 the repo amendment to the Bankruptcy Code was passed.  The amendment applied only to the most common repurchase agreements, specifically those that were backed by Treasury and Agency securities, bank certificates of deposit and bankers’ acceptances.  Repo traders were now permitted to seize collateral, liquidate it and net the proceeds against the bankrupt counterparty’s obligations without interference from a bankruptcy trustee, as long as the collateral fell into one of the protected categories.

Because the rules of repo were not established by a self-regulatory organization, the contractual rights that were exempt from bankruptcy had to be broader than those granted in 1982.  Specifically the repo amendment states that it applies to contractual rights including any right “whether or not evidenced in writing, arising under common law, under law merchant or by reason of normal business practice.”[6] Thus the 1984 law set a precedent for the expansion of safe harbor provisions to unregulated financial contracts.[7]


[1] Kenneth C. Kettering, March 2008, 29 Cardozo L. Rev. 1553  “Securitization and its discontents:  the dynamics of financial product development”, p. 1640

[2] Norman Bowsher, “Repurchase Agreements,”  Federal Reserve Bank of St Louis Review, Sept. 1979, p. 19 and for the 1983 data Kettering, 2008, note 287 quoting Peter Sternlight, Executive Vice President of the Federal Reserve Bank of New York statement for Bankruptcy Reform: Hearings Before the Senate Judiciary Comm., 98th Cong. 328 (1983).

[3] Jeanne Schroeder, “Repo Madness: The characterization of repurchase agreements under the bankruptcy code and the UCC,” 46 Syracuse Law Review, 1996, p. 1007.

[4] Kenneth Garbade, “The Evolution of Repo Contracting Conventions,” NYFRB Economic Policy Review, May 2006

[5] Kenneth Garbade, “The Evolution of Repo Contracting Conventions,” NYFRB Economic Policy Review, May 2006, p. 37.  Raw data on repos is available here:  http://fraser.stlouisfed.org/publications/frb/page/31488.

[6] Hance p. 743.

[7] Of course, the 1978 Bankruptcy Act also granted an unregulated market safe harbor, but forward contracts were exempt from regulation, because they were commercial — not financial — contracts.  (Ramdhanie, p. 277).

The First Safe Harbor Exemptions #1-2

In 1978 the Bankruptcy Act of 1898 was thoroughly revised.  The revised Act granted special privileges to both commodity brokers and forward contract merchants in the event of a customer bankruptcy.[1] Commodity brokers dealt in exchange traded derivative contracts and had been subject to regulation since the Commodities Exchange Act of 1936.  Forward contract merchants, by contrast, did not trade in derivatives; they traded in goods where the contract had a settlement date in the future – that is they traded in commercial contracts for future delivery. [2]

The privileges granted by the 1978 Bankruptcy Code to commodity brokers and forward contract merchants were the following: they were exempt from the automatic stay that prevents creditors from seizing and liquidating collateral, they were protected from the bankruptcy trustee’s power to avoid recent pre-bankruptcy transactions and they were permitted to offset claims against debts.  Thus, they were allowed to seize margin funds (i.e. collateral), close out customer positions and net the value of margin against the customer’s holdings.[3]

Overall the effect of the 1978 Bankruptcy Code was to give special protection to two categories of contracts:  exchange traded commodity contracts and commercial contracts for future delivery.  In order to understand why these privileges were granted, we need to understand how these markets operated.

An exchange sets rules for a market to ensure that every trader can honor his obligations.  Typically the value of every position is calculated on a daily basis in a process called mark to market, and traders are required to post margin sufficient to cover their obligations under the contract at current prices.  Since the margin posted against exchange traded contracts is the property of the trader who posted it, in the absence of special bankruptcy procedures these margin payments could easily be tied up in bankruptcy proceedings.  Furthermore, the obligations of a bankrupt party are frozen at the date bankruptcy is declared, so standard bankruptcy procedure disrupts the process of marking and margining contracts.

Another complication is that exchanges are central counterparties.  Thus, all trade on the market is trade with the exchange.  Because the exchange is backed by a partnership of the largest dealers in the market, everyone can trade on the market with confidence that every contract will almost certainly be honored.

Standard bankruptcy procedure can easily impose significant losses on the exchange, particularly if a large trader fails and triggers the failure of a few smaller traders.  To minimize the possibility that large losses could destabilize an exchange the 1978 bankruptcy law allows brokers to close out a bankrupt’s position, seize any margin posted and determine the net obligation without delay.  In this way, bankruptcy law validates the process of marking and margining contracts, by permitting the margin funds to be used to settle the debts against which they have been posted in a timely manner.  Sometimes these provisions of the Bankruptcy Code are called “safe harbor” provisions.

While these rights were only written into the bankruptcy code in 1978, the fact that exchanges were self regulatory organizations (SROs) may well have served as an informal mechanism protecting margin payments from inclusion in bankruptcy estates in earlier years:  as long as no claim was made on a bankruptcy court, there may have been no obligation to resolve margin payments through bankruptcy.  This view is supported by the fact that in 19th century Britain the law viewed the resolution of a defaulter’s Stock Exchange accounts as determined by the rules of the Stock Exchange and thus protected from the trustee of the bankruptcy estate.[4] It seems plausible that a similar system may have been in effect in the United States prior to the 1978 law.  If this speculation is correct, we may also speculate that commercial contracts for future delivery were granted the same treatment because there was a tradition of allowing forward contract merchants to seize and liquidate the margin posted by bankrupt counterparties in a timely manner.

In 1982 safe harbor privileges were also granted to securities brokers and clearing agencies, thus expanding the bankruptcy exemptions to securities-based derivatives.  The 1982 amendment was careful to specify that the exemption from the automatic stay only applied to contractual rights “set forth in a rule or bylaw of a national securities exchange, a national securities association or a securities clearing association.”[5] In other words, the 1982 amendment was designed to protect practices sanctioned by the rules of self regulatory organizations, not the terms of individually negotiated contracts.[6]

In short, after 1982 derivatives subject to regulation by self regulatory organizations and commercial contracts for future delivery, as well as the margin payments related to these contracts, were formally excluded by acts of Congress from resolution in bankruptcy court.  It is possible that this represented nothing more than the codification of existing norms.  Whether this special treatment was a continuation of past procedures or not, the fact that it was written into the law may have encouraged the growth of exchange-traded derivatives.


[1] Note that forward contracts were carefully defined in the Commodities Exchange Act to include only transactions where future delivery is intended (i.e. they are commercial rather than financial transactions).

[2] Karen Ramdhanie, 1999, “Derivatives Contracts of Insolvent Companies”  18 NY Law School J of Intl and Comp Law, p. 277.  If the intent of a contract was to exchange in cash the difference between the spot price on the settlement date and the contracted price, the contract did not qualify as a forwards contract.  In fact, such contracts were unlawful under the terms of the Commodity Exchange Act.
Because the meaning of the term “forward contract” has changed, I am going to adopt the modern usage in this paper.  Thus, from this point on the contracts referred to as “forward contracts” in the 1978 law will be called “commercial contracts for future delivery,” and the term “forward contract” will be reserved for a derivative contract that may be settled in cash.

[3] Ramdhanie, p. 284 – 288.

[4]The settlement of debts incurred on an exchange separately from the settlement of a bankruptcy estate was upheld under the law in 19th c. Britain.   Ex parte Grant re Plumbly (1880) is explained in The Law Relating to Betting, Time-bargains and Gaming, George Herbert Stutfield, Waterlow & Sons, 1884, p. 78 – 79.

[5] Jonathan Hance, 2008, “Derivatives at Bankruptcy:  Lifesaving knowledge for the small firm” 65 Washington & Lee Law Review p.  741.

[6] The Securities Exchange Act of 1934 recognized that self regulatory organizations played an essential role in American markets and made them the partners of the SEC.

The Legal Foundations of Financial Collapse (Intro) #1-1

As economists, financiers, journalists and many others struggle to understand the financial crisis with which the 21st century opened, derivatives inevitably enter the conversation.  Unfortunately derivatives – especially those that are traded only through dealers on what are called “over the counter” markets – are not well understood outside the financial industry.  Perhaps, the fact that regulators and legislators permitted the development of markets that they did not understand is an important cause of the crisis.

Starting in the early 80s, the financial industry lobbied aggressively for legislation that favored youthful markets in repurchase agreements and derivatives.  Initially the industry was supported by the Federal Reserve, because the financial system was already fragile and there was genuine concern that losses in these new markets could be destabilizing.  In later years, the industry was careful to frame its requests for special privileges under the law in the same terms:  the stability of the financial system could only be ensured if these privileges were granted.  Neither regulators nor legislators were willing to challenge the expertise of the financiers.[1]

In 1984, in 1990 and in 2005 Congress passed laws exempting certain financial contracts from the standard provisions of the bankruptcy code.  In each case, the effect of the law was to protect collateral securing the contract from those provisions of the bankruptcy code that allow a judge to review the claims of secured creditors and to protect the interests of other creditors whenever necessary.  By granting derivative counterparties special treatment under the law, these exceptions guarantee that no contract is more secure than a repurchase agreement or an over the counter derivative – this guarantee comes at the expense of other secured creditors and bondholders.  In short, in modern bankruptcies, there is one class of claimant that doesn’t even need to show up in court to prove its claim – the counterparties to collateralized derivative contracts.

The introduction of inequitable treatment into the bankruptcy code would be acceptable, if in fact the financial contract exemptions worked to protect the stability of the financial system.   Recent experience indicates, however, that the special treatment granted to repurchase agreements and over the counter derivatives tends to reduce the stability of the financial system by encouraging collateralized interbank lending and discouraging careful analysis of the credit risk of counterparties.  The bankruptcy exemptions also increase the risk that creditors will run on a financial firm and bankrupt it.  Thus, the bankruptcy code has been rewritten to favor financial firms and this revision of the law has had a profoundly destabilizing effect on the financial system.


Introduction to the Bankruptcy Code

The Bankruptcy Code is designed to distribute the assets of a firm that cannot meet all of its obligations as equitably as possible across the firm’s creditors.  Thus, as soon as a firm files for bankruptcy, creditors are prohibited from collecting on the firm’s debts, and recent payments made by the firm may be recalled by the trustee of the estate.  The principle is that the trustee needs to take the time to realize the value of the firm’s assets and evaluate the full extent of creditor claims.  Only after the trustee has the whole picture of assets and liabilities can the bankrupt estate be distributed fairly to creditors.  Because payments made just prior to bankruptcy can potentially have the effect of favoring one creditor over another, the trustee has broad powers to void such transactions.

The prohibition against collecting on a bankrupt firm’s debts is called the automatic stay.  The trustee’s power to cancel recent transactions is called avoidance.  Prior to the special exemptions enacted by Congress over the counter derivatives were subject to the same treatment in bankruptcy as all other contracts.  The automatic stay ensured that netting of offsetting derivatives and termination of derivatives could only take place after receiving the approval of the bankruptcy court.  Furthermore, collateral that had been posted against derivative contracts over the last 90 days was subject to avoidance – and could thus be reclaimed as the property of the estate.  Collateral that had been posted over the previous year could be avoided if the transaction had an adverse effect on the equitable distribution of the estate.

In general, the purpose of the bankruptcy exemptions enacted by Congress was to ensure that participants in the markets for over the counter derivatives and repurchase agreements were not subject to the automatic stay or to the avoidance powers of the trustee.  The bankruptcy exemptions guarantee that over the counter derivatives and repurchase agreements can be netted and terminated instantly when bankruptcy is declared.  Any collateral can be seized and sold, free and clear of encumbrances.  Effectively under current law once collateral is posted against a derivative or repurchase agreement, it cannot be touched by a bankruptcy trustee – unless the recipient of the collateral was aware that the transfer was fraudulent.

Before examining the effects of these changes to the bankruptcy code, we will outline the history of the changes and then analyze why the exemptions were put in place.  Understanding the reasoning behind these bankruptcy provisions is essential to any discussion of the role that derivatives and repo markets played in destabilizing the financial system.


[1] There were exceptions amongst the regulators, such as Brooksley Born.  They were, however, unsuccessful in their endeavors.