FR2004 fails data reflect cumulative "fails to receive" and "fails to deliver" over the course of a week for the primary dealer community only. The cumulative weekly totals are calculated by summing the fails outstanding on each business day of the reporting week.¹ These totals include both fails that started during the reporting week as well as fails that started in prior weeks and have not yet been resolved. The aggregate fails data include fails associated with both outright transactions and financing transactions.

Fails data are reported for four distinct categories: Treasury Securities, Agency Securities, Mortgage-Backed Securities and Corporate Securities. Mortgage-backed securities include those issued and insured by government sponsored enterprises. Privately issued mortgage-backed securities are categorized as corporate securities. The FRBNY has collected aggregated fails data in this form since July 1990 for Treasury, Agency and Mortgage-Backed securities, and since July 2001 for Corporate securities.

Reported fails numbers sometimes can reach elevated levels due to so-called "daisy chains" and "round robins" in which an initial delivery failure causes a chain of subsequent fails as the party expecting to receive the security in the initial transaction fails to deliver to its counterpart in the second transaction, and so on. Daisy chains and round robins are ultimately not the cause of fails. Fails, at root, are caused by the core short positions of cash and repo market participants.

As described in the primer below, there are many factors that can create an initial delivery failure. Once a significant volume of fails occurs, lenders of collateral sometimes also withhold collateral because they are concerned that existing fails diminish the likelihood of that collateral being returned to them. Such withholding can be self-fulfilling because withholding scarce collateral can increase the incidence of fails in and of itself.

The importance of delivery chains and the potential for feedback effects from changes in the withholding behavior of collateral lenders also imply that relatively small amounts of collateral can settle a larger volume of failed transactions: an increase in collateral can be delivered from one party to the next to clear up a chain of failed trades and the resolution of failed trades may, in turn, make collateral lenders more willing to lend securities that had been in short supply.

Introduction
A settlement fail occurs when securities are not delivered and therefore paid for on the date originally scheduled by a buyer and seller. The transaction can be an outright sale or the starting or closing leg of a repurchase agreement. Fails are important because they expose market participants to the risk of loss in the event of counterparty insolvency. The prospect of such loss leads participants to devote resources to monitoring and controlling counterparty exposure and could, in an extreme case, lead them to limit their secondary market trading. This primer considers the reasons for settlement fails, the cost of a fail, and measures to avoid or cure fails.

Reasons for Settlement Fails
Fails occur for a variety of reasons. One source of fails is miscommunication. Despite their best efforts to agree on terms, a buyer and seller may sometimes not identify to their respective operations departments the same details for a given transaction. On the settlement date the seller may deliver what it believes is the correct quantity of the correct security and claim what it believes is the correct payment, but the buyer will reject the delivery if it has a different understanding of the transaction. If the rejection occurs late in the day there may not be enough time for the parties to resolve the misunderstanding.

In some cases a seller or a seller’s custodian may be unable to deliver securities because of operational problems. An extreme example is the September 11 catastrophe that destroyed broker offices and records, impaired telecommunications links between market participants, and damaged other critical infrastructure. Less extreme operational problems can also precipitate settlement fails, and are not uncommon.

Finally, a seller may be unable to deliver a security because of a failure to receive the same security in settlement of an unrelated purchase. This can lead to a “daisy chain” of fails; where A’s failure to deliver bonds to B causes B to fail on a sale of the same bonds to C, causing C to fail on a similar sale to D, and so on. A daisy chain becomes a “round robin” if the last participant in the chain is itself failing to the first participant.

Fails also occur “naturally” when special collateral repo rates approach or reach zero. In general, a market participant would be better off borrowing securities to avoid a fail even if the interest on the money lent in the specials market is below the general collateral repo rate, because (as explained below) the alternative is forgoing interest altogether.³ However, this incentive becomes less compelling as a specials rate approaches zero. A specials rate will approach zero if there is unusually strong demand to borrow a security, e.g. following heavy short selling by hedgers, or if holders are unusually reluctant to lend the security.⁴

The most important cost of a fail is that the seller loses the time value of the invoice price over the interval of the fail. (This cost may be reimbursed by the buyer if the buyer’s actions caused the fail, for example, by improperly rejecting securities tendered by the seller, or by a third party if the third party’s actions caused the fail, for example, a custodian who failed to deliver securities pursuant to the instructions of the seller.) This implicit penalty (which can be quantified as the interest that could have been earned in the federal funds or general collateral repurchase agreement markets) provides an incentive to sellers to avoid and cure fails. (There is an exactly offsetting benefit to a buyer who fails to receive securities and therefore does not have to pay for them as soon as originally scheduled: it can invest the invoice price until the securities arrive.)

Additionally, a fail exposes both the buyer and the seller to replacement cost risk. The buyer faces the risk that the seller becomes insolvent before settlement and that the price of the security increases prior to the seller’s insolvency. Conversely, the seller faces the risk that the buyer becomes insolvent and that the price of the security declines prior to the buyer’s insolvency. The significance of replacement cost risk exposure may be small for a fail that does not last more than a few days, but it increases as a fail continues. Aged fails generally prompt market participants to step up their monitoring of counterparties. Initially, the increased monitoring may be nothing more than a phone call to identify whether there has been a misunderstanding, but it can escalate to credit reviews. Marking fails to market can mitigate this risk. Additionally, the net capital requirement for regulated brokers and dealers adopted by the Securities and Exchange Commission assesses capital charges for fails that have gone on for more than a specified interval of time.

Fails stemming from an inability to deliver securities because of a failure to receive the same securities can be cured (or avoided entirely) by borrowing the securities needed for delivery from another party and delivering the borrowed securities. The borrowed securities can be returned when the seller finally makes delivery.⁶

___________________

¹ For example, cumulative weekly fails of $7 billion are consistent with fails of $7 billion on a single day or fails of $1 billion for each of seven consecutive calendar days (fails outstanding over a weekend or holiday are cumulated for each day).

² This primer draws upon the paper of Michael J. Fleming and Kenneth D. Garbade, “When the Back Office Moved to the Front Burner: Settlement Fails in the Treasury Market after 9/11,” Federal Reserve Bank of New York Economic Policy Review, November 2002.

³ General collateral repo rates typically closely track the Fed funds target rate. Special collateral repo rates fall below the general collateral rate by the size of the fee participants are willing to pay to borrow a particular security. For example, a general repo rate of 1 percent and a special repo rate of 25 basis points imply a fee of 75 basis points to borrow a security.

⁴ Once fails have arisen and the repo rate has fallen to zero, the withdrawal of collateral lenders may keep repo rates at zero for longer than otherwise.

⁵ FICC also marks fails to market every day and thereby mitigates the replacement-cost risk of a continuing fail. The invoice price specified in an instruction to deliver a security to FICC is revised to the contemporaneous current market value (CMV) of the security each day that a member continues to fail to deliver the security. The change in invoice price is offset with a cash payment called a “fail mark.” If the CMV increases, the member has to make a payment equal to the product of the increase in the CMV and the quantity of securities to be delivered. (The member recovers this payment when it delivers the security at the revised, higher, CMV.) Conversely, if the CMV declines the member will receive a cash payment. At the same time, FICC marks to market its own fail to the member that was to receive the security.

⁶ This method of curing a fail may not work if fails in an issue are extensive and there is an expectation of being failed to on a newly initiated borrowing.