It happens quite often when reading about the history of the financial markets or about the Federal Reserve. Two company names are often mentioned: Drysdale Securities and Lombard-Wall. Not only are these two names important in the U.S. financial system, but they are also key developments for the modern Repo market, and for that matter, the entire “shadow banking” system.
The History of Repo
Though collateralized lending can be traced back to China about 3,000 years ago, the modern use of Repo financing began with the Federal Reserve in 1917. During the First World War, a war-time tax instituted by the United States made the issuance of commercial paper more costly for banks. In order to provide banks with more liquidity, the Fed stepped in and bought banks’ government securities with the agreement to sell them back days later – the Repurchase Agreement was born. In the 1920s, just as the federal funds market was developing, the Fed added Banker’s Acceptances (BAs) to the list of collateral they would accept in Repurchase Agreements, which also helped create a secondary market for banks’ BAs.
The inter-dealer Repo market began to develop after World War Two and with the passage of the Treasury-Federal Reserve Accord of 1951. In this act, the Fed was given the mandate to control inflation and they began using Repurchase Agreements to inject cash into the market to control interest rates more actively. At the same time, Repo became a financing tool for securities dealers and money-center banks. The first use of Matched-Sales occurred in 1966 when the Fed needed to drain excess liquidity that resulted from a sudden surge in bank reserves. The Repo market continued to grow and was even given a boost after the Fed exempted Repo from banks’ reserve requirement calculations in 1969. At this time, there still were no “customers” in the market, it was an inter-dealer market primarily used for funding long positions, or for money-center banks to invest cash. Short positions were covered through “box loans” from a firm’s clearing bank. If a dealer was short a Treasury they would keep that security uncovered and then leave a long position of another U.S. Treasury in their “box” (securities account) as collateral. After the wire closed, the clearing bank would swap the long position for the short position through the bank’s Securities Lending Department, which became the predecessor of the Bonds Borrowed/Bonds Loaned transaction.
By the end of the 1970s, fundamental changes in the market had begun. Growth in the Repo market followed the growth of U.S. government debt and new institutions began entering the Repo market; the size of the U.S. Repo market was at $14.8 billion in 1977.
Higher Interest Rates
Runaway inflation and the Fed’s attempt to control that inflation by managing the nation’s money supply led to much higher short-term interest rates after October 1979. Whereas banks were only allowed to pay 3.00% interest on savings accounts, overnight Repo rates were trading between 10% and 20% between 1980 and 1982. Institutional investors were obviously attracted to the Repo market to enhance their returns on cash. At the same time, the increased volatility of interest rates led many in the market to realize the need for better risk management. More and more long positions needed to be hedged by shorting Treasurys and dealers began covering their shorts in the inter-dealer Repo market. By 1981, the size of the Repo market stood at $111 billion.
By the early 1980s, the Repo market was the largest single securities market in the world. Since government securities were still “exempt” from most provisions of federal securities laws, even more participants were attracted to this unregulated market. Regional banks, as well as large corporations and municipalities entered the Repo market. In Europe, after the “Big Bang” of deregulation hit the banking sector in London in 1986, investment banks began using Repo to finance long bond positions, with trades predominantly booked as buy/sell-back transactions. The high cost of borrowing securities from Euroclear and Cedel further increased inter-dealer short-covering in the developing European Repo market. By 1986, the size of the Repo market in the U.S. had crossed $200 billion.
Along with the rapid rise in the Repo market came flaws that were mainly overlooked, or maybe just ignored. Case in point: accrued interest. When someone bought and sold a U.S. Treasury outright, the securities was delivered with the accrued interest added to the purchase price. That is, when you bought a U.S. Treasury, you had to pay for the amount of interest which had already accrued on the bond since the last coupon payment date. When interest rates were high and coupons above 10%, it meant a lot of coupon interest was accruing on the bonds each day. Repo pricing was a little different. Since Repo was generally just overnight and settled for cash, including the accrued interest in the Repo price did not seem important.
David Heuwetter, the head trader at Drysdale Government Securities, put together a trading idea. It was really scheme to take advantage of differences in the market convention between outright Treasury purchases and Repo trades. He started short-selling U.S. Treasurys outright, where he received the price plus the accrued interest. Then he borrowed the securities to cover his shorts in the Repo market, paying only the market price. He was getting the full use of the accrued interest on the bonds at no cost. In order to maximize to amount of cash he was collecting, he concentrated on shorting Treasurys about half way between the semi-annual coupon payment dates. With years of declining bond markets, he had made a lot of money shorting the Treasury market and by February 1982 his trades reached $4.5 billion in short positions and $2.5 billion in long positions, as one Wall Street trader later remarked, “It was the most astonishingly leveraged operation that I have ever seen.” Overall, it was not a bad trading strategy since he won under two out of three possible market scenarios. If the bond market went down, he made a lot of money. If the market stayed the same, he earned free interest on the cash the trade generated. If the Treasury market rallied, he risked a significant amount of money. As it turned out, between February 1982 and May 1982, the long-end of the Treasury market rallied considerably. [See graph below]
When the May 15 coupon interest payments were due on Monday, May 17, Drysdale had been wiped out and did not have enough money to make the payments. That Sunday evening, Heuwetter called Drysdale’s clearing bank, Chase Manhattan, and informed them that “we may have a problem” meeting the $160 million interest payment due the next day on $3.2 billion U.S. Treasurys. Could Chase possibly lend Drysdale $200 million to tide them over? Yes, the market rally had wiped them out, but the problem was even worse than that. Drysdale had conducted their Repo trading mostly through Chase’s Securities Lending Department and Drysdale’s counterparties believed they were facing Chase and not Drysdale. Chase believed they were acting “as agent” for Drysdale, so they would not accept responsibility for the $160 million in coupon interest payments on that Monday, but the problem grew even worse. Up until then, the government securities market had operated under the assumption that the buyer in a Repo trade was entitled to liquidate the trade in the event of a default by the seller. There was no law on books differentiating a Repo from a collateralized loan and there had never been a case in court to set a precedent. The market was unsure whether they could liquidate the Drysdale/Chase Repo trades.
The lack of clarity as to the bankruptcy status of a Repo left Chase with a major dilemma. If the bank refused to pay the coupon interest to the Street and that contributed to any of those securities dealers going bankrupt, a future court ruling against Chase could expose them to liability for the damage they caused. However, if Chase made the coupon interest payments for Drysdale, they were clearly taking a loss and would line up as a creditor of Drysdale in the bankruptcy. It was a lose/lose situation.
On Wednesday, the Fed intervened and called together the heads of the 20 largest banks and securities dealers for a meeting at the Fed’s New York office. Not only was there pressure put on Chase from the dealer community, but also by the Fed itself. Though the Fed would later announce that their only involvement was hosting a meeting, quite similar to the future meetings for Long Term Capital Management in 1998 and Lehman Brothers in 2008. The next day Chase relented and, as they say, the rest is history. In most text books, that’s where the story ends, but in the context of the Repo market, it was just the beginning.
Repo Accrued Interest Pricing
One week after Drysdale collapsed, the executive committee of the Primary Dealer’s Association met to discuss a new convention for Repo and proposed including accrued interest on Repo trades. The Association formally adopted the rule on June 14, 1982 and by October, 1982, the New York Fed ordered accrued interest to be included on the pricing of all Repo trades, moving to “full accrual pricing” as the standard market practice. But the story still did not end.
Repo Property Rights
Since there was no actual law on the books defining Repo and no court cases to establish precedence, the legal status of Repo remained in question. The key issue was the treatment of Repo during a bankruptcy or a default. If Repo was technically a “collateralized loan,” then securities held by a creditor would be part of the bankruptcy proceeding, meaning the securities could be tied up for months, or even years. If a Repo was technically a sale and a repurchase transaction, then the securities could be immediately liquidated in the event of a default. It was a major issue that would have ramifications going forward. On one hand, a Repo looked like a single transaction (a collateralized loan) since the buyer had the right to the coupon payments and the right to substitute collateral. On the other hand, a Repo was also a pair of transactions (one buy and one sell) where the buyer took title of the securities, could sell them or pledge them to another party and had the right to return identical securities (e.g. U.S. Treasury 10 year notes) when the trade matured. Since these factors were also true, Repo trades should be free from the automatic stay provisions in a bankruptcy and could be immediately liquidated.
Following the collapse of Drysdale, the government securities market and Repo market were frozen. Though Chase covered the loss and took over Drysdale to liquidate them, there still was no real resolution of the Repo property rights issue and it weighed on the market. That issue came up again just three months later when another firm, Lombard-Wall, collapsed in August 1982. Once again, Lombard-Wall had Repo transactions with the Street and with customers, but in this case, there was no large bank to cover the losses. Immediately following the bankruptcy, Lombard-Wall’s Repo trades were tied up in bankruptcy court with no one allowed to liquidate them. In September 1982, one month later, the Federal Bankruptcy Court of New York ruled that a Repo was a buy and a sell, making it two separate transactions and thus not a collateralized loan. The court recognized that allowing prompt liquidation was necessary to continue the orderly functioning of the markets. Two years later, in 1984, Congress passed an extension of the Federal Bankruptcy laws so that Repo on Treasurys, Federal Agencys, CDs and BAs were exempt from automatic stays in a bankruptcy by law. Those extensions were expanded to include more types of Repo collateral in the mid-1990s and again in 2005.
It is often overlooked how important the bankruptcies of Drysdale and Lombard-Wall were for the Repo market. The question of the legal status of Repo in a bankruptcy became a crisis in Drysdale, then became established precedence in court by Lombard-Wall. That new legal status allowed for the creation of the entire “shadow banking”* industry. Imagine if the judge had ruled that Repo was technically a collateralized loan and the Repo financing was therefore tied up in bankruptcy court potentially for years. There would clearly be no “shadow banking” industry without the ability to quickly liquidate Repo trades in the event of a bankruptcy.
* Shadow Banking is defined as banking functions that are performed by non-bank financial institutions, for example, a REIT that owns mortgage-backed securities and finances them through Repo, a hedge fund that owns corporate bonds and finances through Repo, or even financing CDOs through the Repo market, as in the case of Bear Stearns Asset Management.