Don’t run CCP auctions by fear, study argues

Paper by BoE economist and co-authors backs ‘second-price’ auctions and limited penalties

auction-penalties montage

Threats may be an effective way for central counterparties (CCPs) to get their members to bid on defaulted portfolios but they could backfire in a crisis, new research claims.

Instead of penalties for non-participation, the research argues for an overhaul of the mechanics of default auctions, with one eye on making the price easier to digest for the winner, and another on limiting the pain for the losers.

The research was conducted by Gerardo Ferrara, an economist at the Bank of England focused on financial market infrastructure risk and research and CCP policy, and co-authors Xin Li of the University of Westminster and Daniel Marszalec of the University of Tokyo. The paper, published in the Journal of Financial Market Infrastructures, and reflects the personal views of the authors.

When it comes to auctioning off the portfolio of a defaulted member, a clearing house has to serve multiple interests. Its primary aim is to return the CCP to a matched book – a sale at any price will achieve that. But a very low price could produce losses that have to be mopped up by the surviving members.

Competition for the defaulted portfolio is usually achieved via the stick, rather than the carrot, with penalties applied in a variety of combinations – to non-participants, to those who submit low-ball bids, or even to all losing bidders. The paper argues this approach could in extreme circumstances increase systemic stress.

“[Applying] a penalty may add an unexpected stress to the balance sheet of a surviving clearing member. Additionally, a penalty can increase the risk that members may not be able to meet margin calls, default fund additional margin and default fund replenishment,” say the authors.

What counts as success?

Price theory normally judges the success of an auction by two criteria, the authors state: first and foremost, maximising revenue; and second, efficiency. But for CCP default auctions, they argue, it is necessary to add a third factor: the effect on systemic stability.

At present, CCP auctions tend to employ one of three methods. The simplest is a straightforward first-price auction, in which the highest bidder wins and pays what they have bid. First-price auctions have several major drawbacks, the authors argue, not least that they encourage overbidding – the “winner’s curse” in which, if the average bid reflects the market’s valuation of the portfolio, then the winning bidder will likely have overpaid.

In a period of generalised stress, the use of a penalty can potentially contribute to a number of serious consequences such as exposing participants to an unexpected liquidity stress
Gerardo Ferrara, Xin Li and Daniel Marszalec

A second category of approach, employed by many CCPs, is a first-price auction with the addition of some form of penalty for non-participation. At Ice Clear Europe, members that sit out the auction or offer a weak bid will see their default fund contributions ‘juniorised’, meaning they will be drawn on first to cover any losses; at Eurex Clearing, the penalty is a straightforward fine. Some CCPs apply a penalty to all non-winning bids, meaning all participants bar one face a cost.

Introducing penalties to a first-price auction provides an incentive to take part, but arguably at the cost of systemic stability, say Ferrara and his co-authors. In a scenario involving defaults at multiple CCPs, these penalties would add to the stress on member firms, they argue – though it should be noted opt-outs are usually offered to members unable to participate.

“In a period of generalised stress, the use of a penalty can potentially contribute to a number of serious consequences such as exposing participants to an unexpected liquidity stress, particularly when they do not consider this within their bidding function. As such, it is questionable whether the use of a penalty mechanism or an unclear set of rules defining the auction format actually contribute to achieving the objective of facilitating the correct functioning of the market infrastructures,” they write.

The third way

A third route already taken by some CCPs is to use a second-price auction, in which the highest bidder wins but pays only as much as the second-highest bid – and the study backs this mechanism, when combined with a limited set of penalties.

The auction would start with the CCP announcing two parameters: a loss-sharing threshold, and a proportion of total losses to be shared. A benchmark valuation of the portfolio would also be made available. The highest bidder would pay the second bid, and receive the portfolio. All other bidders would pay nothing – as long as their bids were over the threshold.

Below-threshold bidders would collectively have to pay the difference between the benchmark valuation and the loss-sharing threshold; if all bids were over the threshold, the auction would have been run as a second-price auction with no penalties.

As a worked example, consider a second-price auction for a portfolio with an offered valuation of $200, and a pre-set loss-sharing threshold of $100. There are four participants: participant A, who bids $180 for the portfolio; participant B, who bids $150; participant C, who bids $80; and participant D, who bids $50.

A clear set of auction rules would help bidders to have more clarity on how to build their bidding function
Gerardo Ferrara, Xin Li and Daniel Marszalec

The outcome of this auction would be participant A winning the portfolio, and paying the amount of the second-highest bid, $150. Participant B loses the auction, but since its bid exceeds the pre-set threshold, it pays nothing. Participants C and D are both under the threshold, however: they must collectively pay 50% of the difference between $200 and $100, in other words $25 each.

The best strategy for a participant in such an auction is to bid at their honest valuation, the authors argue. Bidders who value the portfolio over the threshold will bid honestly in any case – there’s no advantage to underbidding below the threshold and having to pay a penalty – and bidders whose own valuation is below the threshold would not underbid, as this wouldn’t avoid the penalty but would reduce their chance of winning. If their valuation was close enough to the threshold, they might even be best advised to overbid slightly to clear it.

The result is higher revenue than both first-price and simple second-price rules, the authors claim, as well as less risk of overbidding than in a first-price auction – and limited penalties for losing participants, meaning a reduced burden in times of stress.

Quite apart from the issue of picking the best set of rules, Ferrara adds, there are advantages to simply picking one common set of auction rules and making them explicit in advance.

“A clear set of auction rules would help bidders to have more clarity on how to build their bidding function. The current assumptions that the bidding function must be built by using first-price auction are neglecting the possibility of receiving a form of penalty or juniorisation… especially when we are in a situation in which we see multiple defaults across different CCPs that have common clearing members, these clearing members as bidders would need to navigate different auction rules at the same time because they would be participating in multiple auctions across the different CCPs.”

Editing by Tom Osborn

  • LinkedIn  
  • Save this article
  • Print this page  

Only users who have a paid subscription or are part of a corporate subscription are able to print or copy content.

To access these options, along with all other subscription benefits, please contact [email protected] or view our subscription options here:

You are currently unable to copy this content. Please contact [email protected] to find out more.

You need to sign in to use this feature. If you don’t have a account, please register for a trial.

Sign in
You are currently on corporate access.

To use this feature you will need an individual account. If you have one already please sign in.

Sign in.

Alternatively you can request an individual account here: