1. The Federal Reserve and the creation of Repo market 1917
The modern use of Repo financing began shortly after the U.S. entered World War I in 1917. The Federal Reserve was created three years earlier in 1914, and part of it’s mandate was to “provide an elastic currency,” which included providing liquidity to the banking system. During its first few years, injecting cash into the system exclusively involved the “rediscounting” of commercial paper; the Fed loaned money to banks using commercial paper as collateral. With the growing issuance of Liberty Bonds by the U.S. government, banks began presenting government securities to the Fed for rediscounting. Soon after, the Repo market for government securities was born.
2. Drysdale: Reforming the Repo market 1982
Before 1982, “accrued interest” was ignored by Repo market participants for trade pricing, and the bankruptcy court system was not sure whether a Repo was a collateralized loan or a sale and a buy-back.
David Heuwetter, the head trader at Drysdale Government Securities, put together a scheme to take advantage of differences in the market convention between outright Treasury transactions and Repo pricing. He short-sold U.S. Treasurys outright, where he received the price plus the accrued interest, then borrowed the securities to cover his short in the Repo market, paying only the market price. The scheme allowed him the full use of the accrued interest on the bonds at no cost.
By May 15, 1982 Heuwetter was wiped out and didn’t have enough money to make coupon payments. 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 it in the event of a default. There was no law on books differentiating a Repo from a collateralized loan, and a case was never tested in court. The market was unsure whether they could liquidate the Drysdale Repo trades.
After Drysdale’s collapse, the Primary Dealer’s Association formally adopted accrued interest pricing and by October 1982, the New York Fed ordered accrued interest to be included on all Repo trades, moving to “full accrual pricing.” But the story still didn’t end.
The government securities and Repo markets were frozen. In September 1982, the Federal Bankruptcy Court of New York ruled that a Repo was a separate buy and sell transaction, meaning it was not a collateralized loan. The court recognized that allowing prompt liquidation was necessary to continue the orderly functioning of the market. 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.
3. Broker-Dealer bankruptcies and the creation of Tri-Party 1985
Surprisingly, it took the bankruptcies of an assortment of small government bond dealers in the 1980s to create the Tri-Party market. As “customers” were entering the Repo market in the early 1980s, they routinely gave their cash to securities dealers to hold in a safekeeping account. Sometimes called “Hold-In-Custody” (HIC) Repo and unaffectionately known as “trust me” Repo. The cash investors had a lot of faith in their securities dealers, trusting them to hold the cash, hold the securities, and price the securities accurately. As it turned out, many of those small bond dealers end up defaulting and abusing the collateral deposits. Lombard-Wall, Lion Capital Group, E.S.M Government Securities, RTD Securities and Bevill Bresler and Schulman Group were some of them. Between 1977 and 1985, failures by government bond dealers resulted in almost $1 billion in losses.
Not surprising, customers were unwilling to invest their cash in “Hold-In-Custody” Repo after 1984. The Street had a major problem, HIC Repo was an easy source of funding for dealers and good investment for cash customers. The Street needed a safer way of transacting safekeeping Repo and they realized an agent was needed to stand in between the cash provider and the securities provider. A few dealers approached a clearing bank to set up “safekeeping” arrangements where the bank would act as a joint custodian in the transaction. Some time in 1985 or 1986, the Tri-Party Repo market was born.
4. Central clearing of Repos
In the 1990’s, the Repo market began to see the need for a Central Counterparty (CCP). Before the CCP, dealers were “grossing up” their assets with inter-dealer trades spread across the entire market in a web of trades. Once the CCP was established, it allowed offsetting trades to net moving from a gross to net asset business.
In the mid 1990s, two rival clearing houses, Delta Clearing Corp. and Government Securities Clearing Corp. fought for control of the Repo market between 1997 and 1999. Back then, when you traded on the broker screens, you had three options: “Give-up,” “GSCC,” or “Delta.” By 1999, all dealers and liquidity migrated to GSCC. Delta Clearing and “give-up” markets were dropped. Today, any bank trading on the broker screens must be a member of FICC. GSCC won the CCP battle and is now the market standard.
5. Brokertec And Electronic Trading
With the dawn of the internet age in the late 1990s, electronic trading platforms began to develop. The first one to hit the Repo market was Brokertec, backed by broker ICAP. Up until then, voice brokering was the heart and soul of the Repo market. The Repo market was more personal, with information passing through the brokers and well as red wine and steaks passing through the traders.
By 2001, the simplicity and speed of electronic execution had won out. The vast majority of repo trading volume had emigrated to Brokertec. The voice brokers remained important for market color, but in a fast paced market, trading with a single click and the ability to leave orders was a major efficiency improvement.
6. Elimination of Repo Market Squeezes 2006
Before 2006, squeezes were common in the Repo market. In fact, in many trader’s eyes, the sole purpose of the Repo market was for squeezes. Large dealers would accumulate large amounts of a Treasury issue and control it in the financing market – moving rates up and down by adding and removing supply from the market.
This strategy worked for many years, however, by the early 2000s, regulators became more conscience of manipulation in the Repo market and tried to let the dealer community know they didn’t like it. First, they required dealers to submit “large position reports” of how much of a particular Treasury issue they owned or controlled. That was suppose to wave a flag, saying, “we’re watching you.” It was still a game of cat and mouse. The cat could chase the mouse around the market, but as long as the mouse had a place to hide, it kept coming back.
Then, the Treasury Department set out to do something about it in 2006, when James Clouse, a Treasury official, on Sept. 27, 2006 called Repo market squeezes “an exercise in monopoly pricing.” He clearly wanted to clean up the Repo market. The Fed met with the 22 Primary Dealers in November 2006, and when the dust settled, one Credit Suisse and two UBS Repo traders had been fired and no blatant dealer squeezes have occurred since then.
7. Negative Repo rates 2003-06
Before 2003, negative rates in the Repo market were quite rare because market participants always had the option to “fail” (fail to deliver the agreed upon trade). A fail had no costs, or at least, the costs were nominal.
The “big break” for negative Repo rates came in 2003, when the 3.625% 5/13 failed for four months, between August and November. It took an extreme event like that to illustrate the true cost of fails, or at least, show the costs of fails over a long period of time. There was still no Fail Charge and failing was still the equivalent of 0.0% interest rate, but this time, it was different. Broker-dealers began accumulating a regulatory net capital charge, fails became assets on the balance sheet, labor costs in the back-office increased, and at times, the fails were annoying good customers. One issue that quickly became apparent was that no one could call margin against the fails.
To help dealers clean up problem fails, the Repo market devised the Guaranteed Delivery (GD) Repo trade. The “seller” was supposed to have the securities in their “box” and would guarantee delivery by the end of the day. Later in 2005, the Repo market tweaked the GD Repo trade, but this time the seller was required to pay a penalty if they failed to deliver. It was a better way of trading negative rates because it created a cost for not completing the trade. The Bond Market Association officially adopted the new type of trade in April 2006, calling it a Negative Rate Repo (NRR).
8. Broker averages and indexing
For years, the Repo brokers compiled averages of where Specials and GC traded on their screens each morning, they became known as the “10 o’clock averages.” As Brokertec emerged as dominant trading platform for Specials, the “Brokertec 10:00 average” became the standard currency for describing where an issue traded. As the “average” became more popular, dealers began quoting trades with their customers at fixed spreads off the average, for example, using the Brokertec average + 5 basis points for for long positions and Brokertec – 5 for short positions.
As the whole idea of Repo market indexing began to emerge, it was realized that averages became the equivalent of a “floating rate” for the Repo market. Of course, once there’s a fixed rate and a floating rate the next step is a swap between the two rates. At some point in the near future, formalized Repo interest rate swaps will be standard trading mechanisms for managing short term interest rate risk.
9. Financial crisis and the “Fail Charge”
During the height of the Banking Crisis in 2008, the amount of fails in the Treasury market were astronomical. The Treasury put pressure on the bond dealers to institute a mechanism to finally deal with the fail problem.
In January 2009, the TMPG announced plans to institute a steep penalty for fails – a charge of 300 basis points – making a fail the equivalent of covering a short at -3.00%. Once Fixed Income Clearing Corp (FICC) was on board, the whole market was assured to adopted it. The new practice also included allowing for margin calls on fails.
Clearly, the Fail Charge has created a more fluid Repo market and became a mechanism to price failing securities – negative Repo rates. Since the Fail Charge began, there has been no period of market-wide protracted fails. Yes, there are fails in individual issues, like 10-Year Notes, but those fails remain a temporarily phenomena.
10. New Regulation: Dodd-Frank, Basel III, FTT, And FSB
Over the past few years, new regulation has chipped away at many aspects of Repo market stability:
- Dodd-Frank Section 165 applies to foreign banks in the U.S. and it means that 26 foreign banks will need to hold more capital in the U.S. It was estimated it will create a $330 billion reduction in the size of the U.S. Repo market.
- Since 1985, the Repo market has been exempt from the “automatic stay” provisions of the U.S. bankruptcy code. This means is that Repo transactions can be immediately liquidated once a counterparty is in default. A report from the Securities and Exchange Commission (SEC) in July 2013 put the “automatic stay” exemption for Repo in doubt under certain circumstances.
- There’s talk about a Leverage Ratio being imposed on banks by Dodd-Frank, which will require them to hold a minimum of 5% capital on all of their assets (including Repo) compared to estimates of about 1% currently.
- Basel III – Additional Capital Required For Repo - the Financial Times estimated that $180 billion in additional capital is needed moving from Basel II to Basel III, estimating the average capital set aside for Repo trades will be about 2.5%.
- Financial Transaction Tax – The FTT could possibly put some European banks out of the Repo business. There are 11 EU member countries, led by Germany and France, who have pledged to move forward with it, but implementation looks doubtful after EU legal experts view the tax as illegal.
- Mandatory Repo Haircuts – In August 2013, the Financial Stability Board (FSB) in the U.K. released a negative study on securities re-hypothication. The FSB solution is to require mandatory haircuts on securities financing, and so far from what I’ve read, the proposed haircuts do not exceed those already imposed by the market.
Here’s what it all means:
- Some Repo matched-books will be eliminated
- The financing business will book significantly more trades with central counterparties (CCPs) and exchanges going forward.
- More Repo business will evolve between end-buyers (e.g. money funds) and end-sellers (e.g. hedge funds) – effectively eliminating some of the middle-men in the Repo market.
- Regulatory Arbitrage - Wherever there’s a regulation, there’s a loophole or an exemption. There will be increased Repo market-making activity away from large banks to entities that can better compete in the securities financing markets.
- Shrinking Repo Market - Many pundits have projected that the Repo market will shrink drastically. Added regulation and capital requirements will certainly mean a smaller market, but claims of a major hit to the Repo market are deceiving. As more Repo business is done direct, it will appear the Repo market is shrinking when fundamentally it is not.