CDS Credit-Event Auctions
Click to Print This Page
Introduced in 2005 to facilitate cash settlement in the multi-trillion dollar credit default swap market, credit-event auctions have a novel and complex two-stage structure that makes them distinct from other auction forms. Examining the efficacy of the auction's price-discovery process, we find that the auction price has a significant bias relative to pre- and post-auction market prices for the same instruments, and that volatility of market prices often increases after the auction; nonetheless, we find that the auction generates information that is critical for post-auction market price formation. Auction outcomes are heavily influenced by strategic considerations and "winner's curse" concerns. Structural estimation of the auction carried out under some simplifying assumptions suggests that alternative auction formats could reduce the bias in the auction final price.
With a notional outstanding measured in the tens of trillions of dollars, credit default swaps (CDSs) are today among the most important of all financial instruments. A CDS is a financial security that offers protection against the default 1 of a specified reference entity. Central to the value of such protection is the manner in which contracts are settled following a default, in particular, the payment to be effected from the protection seller to the buyer. Since 2005, a novel and complex auction mechanism has been at the heart of this process; its performance forms the subject matter of our paper.
Some background is useful. For many years, CDS contracts were "physically settled," meaning that the protection buyer delivered the defaulted instrument|or any instrument from the same issuer that ranked pari passu with the defaulted instrument—and received "par" (i.e., the instrument's face value) in exchange. However, cracks in the system surfaced following the extraordinary growth of the CDS market in the early 2000s: For many names, the volume of CDSs outstanding far outstripped the volume of deliverable bonds. Particularly dramatic was the case of Delphi Corporation which, at its bankruptcy in 2005, had an estimated $28 billion in CDSs outstanding against only $2 billion in deliverable bonds (Summe and Mengle, 2006). Such mismatches created the evident potential for market-disruptive squeezes following a default.
In response to these developments, major changes were introduced to the CDS settlement process beginning in 2005. A specially-designed auction mechanism was instituted to identify a fair price for the defaulted instrument; and the market moved to a "cash settlement" system in which protection sellers pay buyers par minus the auction-identified price. This paper investigates the auction's performance over a multi-year horizon, including the efficiency of the auction's price-discovery process among many other questions.2 It represents, to our knowledge, the first detailed empirical investigation of this subject.3
CDS CREDIT-EVENT AUCTIONS: A BRIEF DESCRIPTION
CDS credit-event auctions are two-stage auctions. Stage 1 identifies an indicative price, called the initial market mid-point or IMM, for the defaulted instrument, while Stage 2 identifies the definitive price to be used for cash-settling CDS contracts. The auction has the unusual feature that both the amount auctioned in the second stage, and whether that quantity is for sale or purchase (i.e., whether the second stage is a "standard" or "reverse" auction) are endogenously determined from the first-stage submissions. We provide a sketch of the auction process here; a detailed description including the rationale for the auction structure is provided in Section 2.
All submissions to the auction must go through designated dealers. In Stage 1, dealers make sealed-bid price and quantity submissions. In brief, the price submissions are used to identify the indicative price (the IMM) for the defaulted instrument, and the quantity submissions are used to determine if the second stage of the auction will be a standard or reverse auction.
The Stage 1 price submissions are two-way prices at which the dealers are willing to make markets in the defaulted instrument. They are for a specified quotation amount (say, $5 million) and are subject to a specified maximum bid-offer spread (typically $2 per $100 face value). After eliminating crossing bids and offers, the IMM is identified from these prices by averaging the "best halves" of bids and offers, as described in Section 2. The price submissions are also carried forward to Stage 2 as limit orders.
The Stage 1 quantity submissions are called "physical settlement requests" or PSRs. PSRs must be specified as requests to "buy" or to "sell," and represent undertakings to buy or sell the submitted quantity at the auction-determined final price. PSRs can be submitted by dealers on behalf of themselves and/or their customers, but there are some restrictions. Sell-PSRs may only be submitted by dealers/customers who are net long protection, and buy-PSRs by those who are net short protection. Also the submitted PSRs may not exceed the size of the existing net CDS exposures of the dealer/customer.
By netting the buy-PSRs against sell-PSRs, the auction's "net open interest" or NOI is determined. The NOI is the amount auctioned in the second stage. Since the NOI could be to buy or to sell, both the quantity on offer in the second stage and whether that quantity is for sale or purchase are endogenous consequences of first stage behavior. If the NOI is zero, the auction is over and the IMM acts as the auction's final price; else it proceeds to the second stage.
In the second stage, a standard uniform-price auction is held for the NOI. Dealers submit limit-order prices and quantities on behalf of themselves or their customers, without any restrictions on participation; the (appropriate side of the) dealers' Stage 1 price submissions are also brought forward as limit orders. The price at which the NOI is satisfied is the auction's final price, with one final restriction: the final price cannot exceed the IMM (resp. be less than the IMM) by more than a pre-specified amount for sell-NOI auctions (resp. buy-NOI auctions).
In this paper, we investigate behavior and outcomes in the auctions conducted from 2008-2010. Our analysis is in three parts. Section 4 examines perhaps the most important intended contribution of the auction: price discovery. Building on this, Section 5 looks at bidder behavior in the auctions, including (a) the impact of "winner's curse" and strategic considerations on liquidity provision in the auction, and (b) intra- and inter-auction learning dynamics. Finally, Section 6 carries out a structural estimation of the auction under some simplifying assumptions and uses it to examine the impact of alternative auction formats. A summary of our findings follows.
Figure 1. Average Prices Pre- amd Post-Auction
This figure describes the behavior of the average (log-)price of the deliverable instruments in the CDS credit-event auctions with a sell NOI 5 trading days before and after the auction date. Day-0 is the date of the auction and the day-0 price is the auction-determined final price. The data is described in Section 3 below and the calculation of average prices in Section 4.
Our study opens in Section 4 with a study of the auction's price discovery. The preliminary evidence is discouraging: Market price data on the deliverable instruments indicates that, even after a careful elimination of outliers, auction prices appear to have a significant bias. For instance, in auctions with an NOI to sell (which are the vast majority of auctions in the data), both pre-auction and post-auction market prices are, on average, sharply higher than the auction-determined final prices (Figure 1).
It is tempting to conclude from Figure 1 that the auction does not work well, but economic theory has suggested many reasons why auctions may be informative and yet result in underpricing. So, to get a better feel for the informativeness of the auction, we turn to econometric analysis. And indeed, we find, in contrast to the impression given by Figure 1, that information revealed in the auction—in particular, the auction's final price—is a key determinant of post-auction price behavior. In particular, in the presence of auction-related information, no pre-auction price or quantity information is significant in explaining post-auction price behavior.
What then could explain the observed pricing bias? An obvious suspect is the winner's curse4 problem that may induce conservative bidding (see, e.g., Nyborg and Sundaresan, 1996). A second and more subtle possibility, suggested by the theoretical work of Wilson (1979) and others, is strategic behavior by bidders.5 Yet a third possibility, raised by Bajari and Hortacsu (2005), is risk-aversion on the part of bidders. In Section 5, we return to these issues and show that the winner's curse and strategic behavior indeed have significant impacts on auction outcomes.
In Section 4, we also examine a second, indirect, test of auction informativeness, this one using pre- and post-auction market price volatilities. Intuitively, if the auction were fully (or even considerably) informative in identifying the "fair" price of the defaulted instrument, one might expect that post-auction volatility of market prices be lower than pre-auction volatility. We find that this is not the case. To the contrary, we find that price volatility actually goes up after the auction, both on average and for over two-thirds of individual names. This finding is puzzling and difficult to reconcile with efficient price discovery. One possible explanation, suggested by our conversations with market participants, is that many informed and specialized traders (hedge funds, firms with workout desks, vulture investors) enter the market only after the auction; consistent with this possibility, we find a sharp increase in the volume of trading post-auction.
In Section 5, we turn our attention to bidder behavior. We begin with the factors that affect liquidity provision in the auction. The liquidity provided by a dealer is proxied by the slope of the demand (or supply) curve submitted by the dealer in the auction's second stage: intuitively, the steeper the submitted curve, the lower the level of liquidity provision. We examine how liquidity provision in the auction is affected by (a) the possible winner's curse effect, and (b) by strategic considerations, i.e., by the behavior of other dealers in the auction?
Section 5.1 looks at the impact of the winner's curse. In principle, more dispersed information entering the auction should lead to a greater anticipated winner's curse, in turn causing dealers to bid more cautiously, i.e., to submit steeper demand curves. We proxy pre-auction information dispersion with the variability of first-round inside-market price submissions. We find a strong and significant effect exactly along the expected lines: that a higher level of information dispersion leads to steeper demand curves. Motivated by this, we revisit the pricing bias issue and find that the most significant explanatory variable for underpricing is indeed our winner's curse proxy; indeed, it is the only price or quantity variable that is significant across the board.
Section 5.2 examines the role of strategic considerations, an issue highlighted in the theoretical models of Wilson (1979) and Back and Zender (1993). Motivated by the constructed equilibria in these papers (see footnote 5 for the driving ideas), we examine whether the slope of a bidder's demand curve increases in the slope of the aggregate demand curve submitted by the other bidders. We find the hypothesis strongly confirmed.
Sections 5.3 and 5.4 turn to learning dynamics within and across auctions. Section 5.3 looks at how information revealed in the first stage of the auction affects how much a bidder deviates from its own first round bid; a greater deviation from own first-round bid indicates more weight being placed on the "public" information revealed compared to the "private" information that led in the first-round bid. The findings are subtle with a key and interesting role played by winner's curse considerations. In Section 5.4, we examine how past wins and inventory affect current bidding; we find that more past wins leads to less aggressive current bidding.
In the final part of the paper in in Section 6, we carry out a structural estimation of the auction in an asymmetric information setting. The estimation is carried out under some simplifying assumptions that enable us to focus on the second stage of the auction. Utilizing the first-order conditions defining best responses, the estimation uncovers the distribution of signals that drive observed bids in each auction. Using the estimated signals, we then examine the counterfactual of what auction prices would have resulted under truthful bidding (i.e., under a Vickrey auction). We find that the extent of underpricing in equilibrium would be reduced substantially. Under (much) stronger assumptions, we find that switching to a discriminatory auction format would have a minimal impact.
The rest of this paper is organized as follows. Section 2 describes the auction mechanism in detail, highlights its unique characteristics, and provides a brief literature review, as well as a summary of comments from market participants concerning the auction. Section 3 describes the data sources we tap and the features of the data obtained. In Section 4, we test the efficiency of the auction's price discovery process, while Section 5 looks at bidder behavior in the auction. Section 6 carries out the structural estimation of the auction and counterfactual experiments. Section 7 concludes with a discussion of further avenues of research. The appendices carry material that supplements the presentation in the main body of the paper.
II. THE CREDIT EVENT AUCTION
Credit-event auctions were designed by the International Swaps and Derivatives Association (ISDA) in collaboration with the auction administrators CreditEx and Markit. A major motivation behind the auction's unusual format is allowing those investors who wish physical settlement of their existing CDS exposures to replicate such an outcome via the auction. The previous section provided a brief introduction to these auctions. This section presents a detailed description.
As noted above, the credit event auction has two stages. All submissions to the auction in either stage must go through dealers; 12-14 dealers, all of them large banks, participate in each auction. Prior to the auction, a "cap amount" is specified which limits how much the auction's final price may differ from the indicative price, the IMM, identified in Stage 1. The cap amount is typically set at 1% ($1 per face value of $100).
Stage 1 of the Auction
In Stage 1, dealers make two sealed-bid submissions:
- Two-way prices, called "inside-market prices," for the underlying deliverable obligations.
- Physical settlement requests (PSRs) on behalf of themselves and their customers.
The submitted prices are for a specified quotation amount which is announced ahead of the auction. If the quotation amount is (say) $5 million, then the dealer is undertaking to buy up to $5 million at the submitted bid price or to sell up to $5 million at the submitted ask price. (Whether the dealer will actually have to buy or sell at the quoted prices depends on what happens in the second stage of the auction to which these price quotes are transferred; see below.) The quotation amount may vary by auction; for example, it was $10 million in the case of Washington Mutual in 2008, and $5 million in the case of CIT in 2009. The bid-offer spread in the submitted prices is also required to be less than a maximum amount which too is specified ahead of the auction. This maximum may vary by auction, but is typically 2%. That is, assuming a par value of $100, the ask price can be no more than $2 greater than the bid price.
The submitted PSRs represent quantities of the underlying deliverable bonds that dealers commit to buying or selling at the auction determined final price. The submissions must obey certain constraints. Only dealers with net non-zero CDS positions may submit PSRs. Moreover, the PSRs must be on the side of the market that would be used to physically settle a dealer's trades. For example, a dealer who is net long protection can only submit sell-PSRs, since the dealer would have been required to deliver bonds under physical settlement. Lastly, the submitted PSR cannot exceed the dealer's total net exposure. For example, a dealer who is net long $10 million of protection can only submit PSRs to sell $m million of bonds where
Customer PSRs are subject to the same two constraints and must be routed through a dealer. Customer PSRs are aggregated with the dealer's own PSR and the net order is submitted in the auction. Since only the dealer's net PSR is observed, it is impossible to tell what part of a submitted PSR represents customer orders and what part the dealer's own request. (Nor is this data collected by ISDA or the auction administrators.)
PSRs enable investors to replicate the outcomes of physically-settled CDS contracts. Consider, for example, an investor who is long protection and long the underlying bond. Under physical settlement, the investor would be left with cash worth par (say, 100) following a credit event. The same outcome can be achieved in the auction by submitting a PSR to sell the bond; in the case, if P is the auction final price, then the CDS is cash-settled for 100 – P while the bond is sold for P, leaving the investor with cash worth par. Absent PSRs, there is no guarantee that the investor will be able to sell the bond at the auction-determined price.
Once the first-round prices and PSRs have been submitted, three quantities are computed and made public by the auction administrators:
- The initial market mid-point (IMM), determined from the submitted prices.
- The net open interest (NOI), calculated from the submitted PSR quantities.
- Adjustment amounts, computed using the submitted prices and the NOI.
To calculate the IMM, all crossing or touching bids and offers are first eliminated from the given list. (A bid b is crossing or touching with an offer o if ) From the remaining bids and offers, the best halves—highest bids and lowest offers—are chosen to calculate the IMM. The IMM is just the arithmetic average of these best halves. Thus, if there are n bids and offers remaining, the highest n/2 bids and the lowest n/2 offers are averaged to obtain the IMM. (If n is odd, the best (n + 1)/2 bids and offers are used.)
To calculate the NOI, the buy-PSRs are netted against the sell-PSRs to identify the remaining net position. Thus, for example, if a total of $100 million of "buy" and $140 million of "sell" orders were received as PSRs, then the NOI is to sell $40 million.
The Adjustment Amounts
The adjustment amounts are penalties levied for being on the wrong side of the market, that is, for bids that are higher than the IMM when the NOI is to sell, or for offers that are lower than the IMM when the NOI is to buy. To compute the adjustment amount, the difference between the submitted price and the IMM is applied to the quotation amount. For example, suppose the IMM has been determined as 50.00 and there is a net open interest to sell. Assume the quotation amount is $2 million. Then, a dealer who submitted a bid of (say) 52.00 pays an adjustment amount of $(0.02 x 2,000,000) = $40,000. This penalty is not levied if the bid or offer in question did not cross with another offer or bid.
With this, Stage 1 of the auction is complete. If the calculated NOI at the end of Stage 1 is zero, then the IMM acts as the final price for cash settlement of all CDS trades, and the auction is concluded. If the NOI is non-zero, the auction moves to Stage 2.
Stage 2 of the Auction
In Stage 2, a uniform-price auction is held to fill the NOI. Dealers may submit limit orders on behalf of themselves or their customers; there is no limitation on participation in this stage. In addition, the relevant side of the price submissions from Stage 1 are also carried forward into the second part of the auction as limit orders for the specified quotation amounts. Since customer orders are routed through dealers, it is not possible to disentangle the two and to identify which of the (new) limit orders originate from the dealer and which from the dealer's customers.
If sufficient limit order quantities are not received to fill the NOI, then the final price is set to zero if the NOI is to "sell," and to par if the NOI is to "buy." Otherwise, the auction's final price is determined from the limit orders as the price that fills the NOI, but with one additional constraint: If the NOI is to sell, then the final price cannot exceed the IMM plus the cap amount, while if the NOI is to buy, the final price cannot be less than the IMM minus the cap amount.
We spoke to a number of major market participants (dealers, customers, and administrators) to get a feel for such issues as dealers' pre-auction CDS positions and the impact of the "adjustment amounts." We summarize their consensus opinions here. They offer useful pointers for analysis and modeling, but we note that data does not exist to independently validate these opinions.
Concerning net dealer positions, it is generally believed that dealers are roughly "net flat" entering the auction, i.e., that their long and short CDS positions offset. So PSRs submitted in the first round are not generally dealer orders but pass-throughs from customers. What types of customers? A major source of sell-PSR orders are believed to be "basis traders," investors who are long protection and long the underlying deliverable instrument, and who wish to replicate the outcome from cash settlement. Buy-PSRs may have multiple origins, from investors with correlation desks dealing with structured products to ones with workout desks looking to take speculative postions. In the data, auctions with sell-NOIs outnumber auctions with buy-NOIs by almost 3-to-1.6 Finally, regarding the "adjustment amounts," while the penalties are not large in dollar terms, consensus opinion is that they have a greater impact than the immediate monetary value because of the reputational consequences of being seen to be off-market.
Relation to Other Auction Forms
The credit-event auction format shares features in common with some other auction forms but is distinct from all of these, and is significantly more complex than most. We have already highlighted its key feature, the endogeneity of the second-stage auction. In contrast, most auctions (in theory and practice) deal with a fixed quantity on offer that is specified in advance as being for sale or purchase, and have as their objective the maximization of the auctioneer's revenue (if a "sell" or standard auction) or minimization of the auctioneer's cash outflow (if a "buy" or reverse auction). There is no analog of this situation in credit event auctions; rather, price-discovery and smooth CDS market settlement are the key goals.
Broadly speaking, there are two kinds of auctions to which CDS auctions bear some similarity: two-stage auctions and Treasury auctions. Two-stage auctions, studied in Ye (2007), are employed to sell complex and high-valued assets. Like CDS auctions, they have a first-stage used to identify an indicative price, and a second round that identifies the definitive final price. However, the similarities end here. Two-stage auctions are commonly single-unit auctions with a single winning bidder; there are no first-stage quantity submission decisions to be made by the participants. More importantly, in two-stage auctions as currently used in practice, the only role of the first-stage bids is to restrict participation in the second round to those submitting the highest first-stage bids; the bid has no other payoff consequence.
Auctions of Treasury securities worldwide resemble the second stage of credit-event auctions with a sell-NOI: in both cases, there is a given quantity being auctioned, bidders submit limit orders, and the final price is determined by matching the aggregate demand curve to the available supply. Treasury auctions worldwide have been widely studied in the literature; see, e.g., Nyborg and Sundaresan (1996) on US auctions; Nyborg, Rydqvist, and Sundaresan (2002) on Swedish auctions; Keloharju, Nyborg, and Rydqvist (2005) on Finnish auctions; and Hortacsu or MacAdams (2010) on Turkish auctions.
The Literature on Credit-Event Auctions
There are, as far as we know, only four other papers on credit-event auctions. Two of them, Helwege, et al (2009) and Coudert and Gex (2010) are empirical studies. Helwege, et al, looks at various empirical features of credit-event auctions up to March 2009, including a comparison of the auction final price to the market prices on the day of and the day after the auction. A portion of our analysis in Section 4 is based on similar questions, but our analysis has the benefit of more data and is carried out in greater detail. Coudert and Gex examine the performance of the auction process in individual cases including Lehman Brothers, Washington Mutual, CIT and Thomson, as well as Fannie Mae and Freddie Mac. Their focus is more on the functioning of the market in stressful times, though they also provide some documentation on the behavior of prices including the bounce-up in prices after the auction date compared to the auction's final price.
The other two papers, Du and Zhu (2011) and Chernov, Gorbenko, and Makarov (2011) are both theoretical models of CDS credit-event auctions developed in the spirit of Wilson (1979). Both papers take the distribution of post-auction values to be exogenous and common knowledge; the focus in each case is on how the auction-determined price compares to this exogenously-specified price. Taking first stage outcomes as given and assuming only dealers participate in the auction, Du-Zhu model solely the second stage of the auction. They show that there are equilibria of the second stage in which the prices will be systematically biased, with sell-auctions resulting in prices that are too high (relative to fair value) and buy-auctions in prices that are too low. (Taking sell-auctions as the reference point, we will refer to these as "overpricing" equilibria.)
Chernov, Gorbenko and Makarov study a full two-stage game with both dealer and non-dealer participants. The(exogenously-specified) true value of the defaulted bond is taken to be common knowledge; this means, in particular, that there is no asymmetric information, so there is no role for winner's curse considerations. The paper obtains and characterizes subgame-perfect equilibria of the game. It is shown that both overpricing and underpricing equilibria are possible; and that which one obtains depends on the size of net CDS positions entering the second stage relative to the size of the NOI. Since data on dealer positions is not currently available, these implications are not directly testable, but using proxies where feasible, the authors show that the data exhibits patterns consistent with their model's implications.
III. THE DATA AND DESCRIPTIVE STATISTICS
Our auction data comes from http://www.creditfixings.com, a website run by Creditex, one of the two co-adminstrators of the credit-event auctions. The site provides considerable detail on each auction including (a) whether auction is an LCDS (Loan CDS) or CDS auction, and in the latter case, whether the underlying deliverable instruments are senior or subordinated; (b) the list of deliverable instruments in each auction identified by their ISINs, (c) the list of participating dealers, (d) the prices and PSRs submitted by each dealer (identified by name) in Stage 1 of the auction, (e) each limit order (price and quantity) submitted by each dealer in Stage 2 of the auction, (f) whether and what penalties were levied on the dealers, and (g) information on the auction's IMM, NOI, and final price.
Table 1 describes the auction types and the names involved in the auctions. There were a total of 76 auctions over the period 2008-10,7 the bulk of them (51) in 2009. Of these, 54 were CDS auctions and 22 were LCDS auctions. Our analysis in this paper focuses only on the CDS auctions. Table 1 provides a list of the underlying firms in these auctions. (Six firm names appear twice because there were separate auctions for their senior and subordinated bonds.)
Descriptive statistics on deliverable bonds and participation in CDS auctions are presented in Table 2. Panel A provides summary statistics on the deliverable bonds. On average, there were 30+ deliverable bonds per auction, but with huge variation, ranging from a single deliverable bond (in five different auctions) to a high of 298 deliverables (the CIT auction). The median number was 5.5, with six auctions (all financial firms) having more than 100 deliverable bonds.
Table 1. CDS Auctions 2008-10: List of Firms
Panel A of this table lists the auction types (CDS and LCDS) that were conducted over the period 2008-10. Panel B lists the underlying firms for the CDS auctions. The data was collected from the Creditex website, http://www.creditfixings.com. The boldfaced names in the list represent those firms on whose deliverable bonds trading data is available on TRACE, as explained in the text.
Figure 2. The Lehman Second-Stage Demand Curve
This figure describes the aggregate demand curve submitted in Stage 2 of the Lehman credit-event auction. The aggregate demand curve is obtained by summing over all submitted limit orders. The red vertical line represents the NOI, which was $4,920 million.
Panels B-D of Table 2 deal with dealer participation in the auction. 12-13 dealers participated in each auction, with the numbers remaining stable over time. Around 75% of all auctions had an NOI to "sell" at the end of Stage 1, and 25% had an NOI to "buy," with the split again remaining roughly stable over time. Dealer participation was roughly the same regardless of whether the auction turned out to have a buy NOI or a sell NOI, but, as as Panel D shows, the number of limit orders submitted in the second round was significantly higher for sell-NOI auctions compared to buy-NOI auctions. The aggregate quantity demanded in Stage 2 (summed over all prices) vastly exceeded NOI in every auction, although there were often huge bids submitted at very low prices; Figure 2 illustrates with the Lehman auction: the NOI was $4.92 billion.
Panel C of Table 2 describes the penalties (adjustment amounts) for off-market first-round price submissions. On average, 1.2 firms got penalized in each auction, with a minimum of zero and a maximum of 5. Several dealers suffered multiple penalties, with HSBC leading the list with eight penalties over the three-year span.
Where our analysis only concerns behavior within the auction, we use data from all 48 auctions involving non-subordinated bonds. Where we also use market prices of the deliverable bonds (e.g., in the analysis of price discovery in Section 4), we use market price data from TRACE. We look mainly at a horizon of five trading days before the auction to five trading days after the auction. Market price data is available (i.e., at least one deliverable bond is traded over this horizon) for 27 of the auctions; the names appear in boldface in Panel B of Table 1. The remaining auctions have deliverables such as trust-issued securities or euro-denominated covered bonds on which TRACE had no information. Twenty-two of the 27 auctions meet the stronger criterion that there is at least one trade in a deliverable bond (possibly a different deliverable bond on each day) on each of the 10 trading days in our horizon; four of these are "buy" auctions (i.e., have a NOI to buy) and the remaining are "sell" auctions.
Table 2. CDS Auctions 2008-10: Descriptive Statistics
This table describes summary statistics on CDS auctions between 2008 and 2010, such as the number of bidders per auction, the number of bids per auction in each round, etc. The data was collected from Creditex via the auction-by-auction details posted on their websitehttp://www.creditfixings.com. "Number of Firms" refers to the number of underlying firms on whom CDS contracts had been written that were settled by the auctions. The "Number of Auctions" exceeds the "Number of Firms" because some firms had more than one auction (one to settle CDS on their senior debt and one to settle CDS on their subordinated debt). The information pertains only to CDS auctions, not LCDS auctions.
Table 3. CDS Auctions 2008-10: Trading in Deliverable Bonds
This table describes summary statistics on trading in the deliverable bonds of the CDS auctions described in Table 1. The numbers pertain to only the 27 auctions for which data on trading in the deliverable bonds is available, as explained in the text. The data comes from TRACE. In Panel B, "Large Trades" refers to $1 million+ trades. In Panel C, Day A-1 refers to the day before the auction; "Normalized NOI" refers to the ratio of the NOI to the Day A-1 Trading Volume; and three outliers are excluded from the computations as noted below the table.
Summary statistics on the frequency and size of trades are presented in Table 3. Panels A and B deal respectively with the total number of trades and the number of "large" trades (i.e., trades over $1 million. TRACE provides the dollar-size of all trades under $1 million, but trades over that amount are simply reported as $1 million+ trades). Panel A shows that trading volume creeps up before the auction, and then increases sharply on the day after the auction. While trade moderates somewhat after that, the number of trades remains far higher than in the days before the auction. Panel B shows a similar trend for large trades. Finally, Panel C relates the size of the auction (the NOI) to the trading volume one day before the auction. As the numbers show, the former is typically an order of magnitude larger with the mean (resp. median) of the NOI-to-trading-volume ratio being 11.7 (resp. 7.8).
IV. PRICE DISCOVERY IN THE AUCTION
In this section, we examine the importance of auction-generated information to post-auction trading. The principal question that concerns us here is: How good is the auction at price discovery? For example, is there information in the auction's final price for subsequent trading of the deliverable bonds? Is there any more information than was already present in the pre-auction prices? How do market-price volatilities of the defaulted instruments behave pre- and post-auction? How does the other auction-generated information—PSRs, NOI, second-stage limit orders—affect post-auction behavior? We use data on market prices and traded quantities for the deliverable bonds in the 27 boldfaced auctions of Table 1 to study these questions.
Identifying a Representative Market Price
As a first step in the analysis, we need to identify from the market prices a candidate price for the deliverable instrument on each day in the horizon using the traded market prices of the deliverable instruments. We begin by eliminating the data points in TRACE that are clearly erroneous (e.g., some Lehman trades report a trade price of $100 even while most trades took place in a neighborhood of $10-$20, and the auction final price was $8.625). A second, more subtle concern shows up in the cleansed data set: For some companies, certain issues of deliverable bonds tended to trade at systematically different prices from other issues. An extreme example is Charter Communications, whose auction-determined final price was $2.375. Some of the 19 deliverable obligations for Charter (e.g., the one with ticker CHTR.HM) tended to trade in the pre-auction market at prices of $9-$10, while the other deliverables traded at prices around $2-$3, close to the auction's final price of $2.375. This suggests the existence of issue-specific influences on the prices.
There are two different approaches we use to extract a "representative" market price from the data given this problem. The first is manual: we eyeball the data, and eliminate all those de- liverable issues whose prices exhibit systematic differences (e.g., the CHTR.HM ticker mentioned above) from other deliverables on the same name. Using the remaining data, we calculate on each given day the average of the traded prices over all the deliverable bonds on that day, and treat this as the representative price for the bond on that day. (We weight the average by trade size to give large trades more importance. Our results are unchanged if we use an equally-weighted average.)8
This second approach looks to use all the data. It accommodates the possibility of systematic or persistent differences in the prices of different deliverable bonds on a given name, and distinguishes between the fundamental or "pure" price and the issue-specific effect. To identify the pure bond price in the presence of these effects, we run the following set of regressions on each day: letting i index the CDS underlying name, and j the deliverable bonds on that name, we estimate
where pijk is the log of the observed price for the k-th trade in the ,j-th deliverable bond in auction i (or "name" i).9 In words, (1) the bond price is the sum of three components: a "pure" price
, an obligation-specific term uij which is meant to capture systemic or persistent pricing biases, and a "trading noise" term σijk. The quantity is then taken to be the (log of the) market price of name i on that particular day; we refer to it as the "estimated price."
Importantly, the two approaches yield very similar results for our analysis. While we do not report the numbers here, the levels of the prices estimated under the two methods are very close, and, in many cases, almost identical.
1The contingency that triggers payment in a CDS is called a credit event; it includes, but is not limited to, traditional default events (e.g., failure to pay or bankruptcy); e.g., in European and pre-2009 North American corporate CDS contracts, restructuring also constitutes a credit event. For simplicity, we use `default' and `credit event' interchangeably. We note too that neither buyer nor seller of CDS protection need have any actual exposure to the underlying bonds, i.e., the CDS may be "naked."
2The original auction was modified in mid-2006. It is the modified procedure that is described below and is the subject of this paper. In April 2009, the auction was "hardwired" into all new CDS contracts as the default settlement mechanism. While participation in the auction was voluntary until April 2009, it is estimated that parties holding over 95% of the outstanding CDS instruments participated in each auction to that point.
3The literature on CDS auctions is discussed in Section 2.
4Loosely put, the "winner's curse" in a common value auction is the observation that, by definition, the winning bid is the most optimistic of the submitted bids, so the expected valuation of the item conditional on winner's information is less than the expected valuation conditional on the combined information of all bidders. For more details and a formal analysis, see, e.g., Milgrom and Webber (1982).
5In essence, this argument notes that the marginal cost curve facing a bidder is endogenously determined by the residual supply curve that obtains after subtracting the aggregate demand curve of the other bidders. The submission of suitably steep demand curves by other bidders can cause marginal cost to escalate very rapidly for the last bidder, so the bidder cannot increase his own allocation substantially with a small increase in the price. This makes it optimal for the last bidder too to submit a steep demand curve, and the result is underpricing.
6Since there is a positive supply of bonds but a zero net supply of CDSs, it is plausible that some of the long protection CDS positions go to hedge existing long bond positions, while the corresponding short protection positions are naked. Thus, it is natural to expect sell-PSR orders to dominate buy-PSR orders on average.
7There were only three auctions in 2006 and a single one in 2007. Since the format of the auction was changed in late-2006, we focus our analysis on the period 2008-10.
8Since there are several deliverable bonds in a given auction, there is an implicit "cheapest-to-deliver" option that should perhaps be taken into account in computing the comparison market price. In general, using an average price over all deliverables may overstate the comparison market price. Our eyeballing of data and throwing out the bonds with systematically higher prices is meant to address this issue too. Our second approach implicitly achieves the same objective by removing "issue-specific" price effects. As noted below, the two approaches yield very similar prices and analytical results.
9We are grateful to Joel Hasbrouck for suggesting this approach.