A colleague of mine (kudos for you know who you are) once told me that in his competition law class he has a part called “how to make a good cartel?” A thought-provoking academic exercise in many aspects, indeed. When analyzing cartels in the financial sector this popped in mind to raise other questions – how participants in cartels in sophisticated markets (like Euro interest rate derivates) remain on the right side of the game and whether such markets make the underpinning of cartels different.
There is hardly anyone not aware of the EURIBOR scandal. Part of it was settled in 2013 when the Commission fined Barclays, Deutsche Bank, Royal Bank of Scotland and Société Générale a total of EUR 1.49 billion in a cartel settlement decision for participating in a cartel related to Euro interest rate derivatives (EIRDs) and manipulation of the EURIBOR benchmark in 2007.
Another part of the saga was that banks, namely HSBC, Crédit Agricole and JPMorgan, appealed the decision of the European Commission.
HSBC challenged the decision finding a restriction of competition by object and imposing a fine of EUR 34 million for the infringement of Article 101 TFEU. In September 2019 the General Court upheld the Commission’s decision and found infringement ‘by object’ of Article 101 TFEU. Also, HSBC’s allegations that the Commission adopted an unfair procedure were dismissed.
Essential aspects of the HSBC judgment such as ‘by object’ infringements, single and continuous infringement, the Commission’s obligation to give reasons, hybrid settlement procedure could not have been better discussed than in this case annotation.
This post will look at the specifics of Euro interest rate derivates market, cartel incentives and detection in this market.
The Euro interest rate derivates are financial instruments with a value linked to the movement of interest rates. Among those financial instruments are futures, options and swap contracts. There are two main aims of such financial agreements between counterparties – to manage risk due to interest rate fluctuations or to speculate.
As you may know, the EURIBOR is set by a designated panel of banks. Each bank of the panel submits their daily quotes to a calculation agent.
The essential thing is that contributing banks also trade in financial instruments linked to the EURIBOR. Thus, benchmarks impact their exposures.
In regular cartels participating undertakings usually aim at increasing prices. That is different from banks in a panel. They have divergent interests as their exposures to the rate are dynamic and fluctuating.
Furthermore, those interests are uncertain in both the short and the long run. That is because the position of a bank depends on many transactions of their traders and trading desks around the world. Therefore banks “will regularly find themselves on opposing sides of the market, where some gain from an increase in one or more of the rates, while others benefit from a decrease” (see more: here). Briefly, those interests differ in business cycles.
Diversity of opposing interests and uncertain payoffs at a particular time determine that a cartel in a market of interest rates derivates is dynamic. Instead of sticking to one-direction aims and actions, banks do deviate from a cartel to ensure its stability and bare short-term losses from time to time.
Cartels in interest rate derivatives are extremely difficult to underpin. First, that is due to the diverse interests between cartelists and even of the same participant of a cartel but at different time. Second, even traders’ interests may be opposite to the interests of a bank employing those traders. Third, collusion is costly as the payout is uncertain (both in terms of time and benefits) and may require exceeding the cartel timeframe. These features imply that collusion may be episodic and leave no traces in benchmark rates and variation patterns.
Efforts to figure out the incentives of banks to collude in benchmark rates and to model how collusion works are still ongoing. One of the latest research describes two models of facilitating collusions: (1) transaction-based – when banks are rigging a transaction to distort the rates; and (2) front running – when banks get an advantage from self-created inside information about the new rate before it is published.
The models give insights on the ‘administration’ of cartels in interest rate derivates market. However, cartel mechanisms react to regulatory and market changes, thus, larger research on collusions is necessary to understand how manipulation of benchmark markets may be understood and detected earlier.
In the EURIBOR cartel traders exchanged information on their desired or intended EURIBOR submissions, trading positions and pricing strategies.
However, not all information exchanges fall within infringement ‘by object’. The General Court concluded that the exchange of information could be a mere observation any market observer could make. It has to be connected to manipulation of the EURIBOR rate and “<…> removing uncertainty <…> as regards the timing, extent and details of the modifications to be adopted by the undertakings concerned in their conduct on the market.” (HSBC, paras. 184, 193).
That implies the Commission has to prove that information exchange gave the trader advantage which allowed to adjust their trading strategies as a result (HSBC, paras.187-188). Having in mind the discussed features of collusion in interest rate derivates market (e.g. detecting at which point the bank gets a payout), that is a very challenging standard of proof.
In any case, financial investment market participants tend to make deals via chatrooms. I have discussed issues of chatroom and FX market trading here. In EURIBOR, traders also used chatroom which served the Commission well in tracking the behavior and incentives.
The establishment of single continuous infringement is also more challenging than in conventional cartels. Taking into account dynamic derivative markets, not matching interests on cartelists on a particular rate in various time periods and diverging costs, gains or losses on particular transactions, collusion in benchmark rates make the exercise of proving that single continuous infringement not that straight forward.
The evidentiary standard comprises of three elements. It has to be proved that (1) the conduct in question must form part of an ‘overall plan’ with a single aim. Moreover, alleged cartelists have to both (2) intend to contribute to that common objective and (3) be aware of the conduct planned or put into effect by other parties in pursuit of such objective (HSBC, paras. 198, 208).
The General Court ruled that HSBC’s awareness of conduct and ‘overall plan’ in which the bank did not directly participate was not found. The argument to support this was ‘fragmented information’ exchanged with HSBC (HSBC, para 268). This information was for participation in a particular episode of the manipulation. Also, a trader could have not been aware of a stable group of traders participating in other prohibited conduct in the market (HSBC, para. 271).
HSBC participated in a cartel just for a month. Thus, it was relatively easier to prove those three elements of the evidentiary burden. However, looking at the features of the collusion in the market at stake, neither ‘overall aim’ nor direct participation (contribution) or awareness is easily established.
The story is not over. The HSBC judgment is appealed both by the Commission and HSBC. Also, cases challenging the Commission’s EURIBOR decision by Credit Agricole and JP Morgan Chase are still pending. They will provide further guidelines on hybrid settlement procedures and single continuous infringements in benchmarks and related financial instruments.
By the way, manipulation of benchmark interest rates is tackled not only by antitrust rules but regulation as well. Cartels in benchmark interest rates implied a change of regulation. In June 2016, the European Parliament and the EU’s Council of Ministers adopted a Regulation on benchmarks which forbids and sanctions manipulation of benchmarks as a violation of capital markets rules.