Paddling Hard for Margin Implementation

This past month in derivatives markets has felt a bit duck-on-a-pond-like. On the surface, it all looks calm, the duck is serenely gliding along, with nothing more than the occasional quack to break the silence. Under the water, though, its webbed feet are paddling furiously. So it is with the derivatives market, as the largest banks prepare for the start of non-cleared margining rules from September 1.

A couple of articles have recently picked up on the all-out effort to prepare for the rules – this one from Bloomberg, and this one from Risk. Both describe the huge documentation and system changes that have to be put in place, and the rush to get all this completed by the implementation date. They also pick up on the cross-border complexities that have emerged as a result of the fracturing of the global implementation timetable.

The articles cover a big, big issue. From next week, a first-phase group of large derivatives users will be required to post variation and initial margin on their non-cleared derivatives trades. These requirements were originally intended to be rolled out in a coordinated way across jurisdictions, but a decision by the Europe Union to defer its start date in June, followed by Australia, Hong Kong and Singapore earlier this week, means the rules will only be implemented on September 1 in the US, Canada and Japan.

Exchanging collateral on derivatives trades may not seem such a big deal. In fact, the rules will trigger the biggest transformation of derivatives markets in decades. That’s because they touch virtually every aspect of the non-cleared derivatives space – from pricing, funding and documentation, to IT, custody and collateral management. It’s required massive changes to infrastructure, technology and documentation, which have needed to be developed, implemented and tested. And it’s all had to be done very, very quickly. US prudential regulators were the first to issue final rules at the end of October 2015 – a little more than 10 months ago – while Japanese regulators published theirs at the end of March.

Given much of the detailed preparation and implementation couldn’t begin until these final rules were released, it’s meant a huge amount of complex work has had to be squeezed into a matter of months. That includes applying and testing the ISDA SIMM, a common model that will be used to calculate initial margin in the non-cleared space (here’s a short explanation), and adapting existing collateral documents. ISDA has published new credit support annexes (CSAs) for variation margin and for initial margin under various legal regimes, but each phase-one firm will likely need to execute between 100-200 CSAs with other phase-one entities.

This process has been made all the more difficult by the fact there’s now a cross-border element to this. The splintering of the timetable adds to the compliance complexity and could disrupt cross-border trading – margin requirements may or may not apply depending on the status and location of the counterparty, whether it has a guarantee from its parent, and whether it is consolidated with the parent entity for accounting purposes, among other things.

No wonder, then, that the Bloomberg and Risk articles suggest preparations are coming down to the wire. There’s going to be plenty of hard paddling over the coming days, as banks work hard to ensure they have everything in place on time.


Margin Settlement Blues

Our days at ISDA start pretty much the same as everywhere else: get into work, sip on a double macchiato, and get up to speed with the day’s events. Given the avalanche of emails and news alerts, the challenge is always deciding what to read and what to skip. But this opinion piece from Risk grabbed our attention – Regulators must scrap T+1 timezone tax.

The alliterative headline helped – we’re suckers for that kind of thing. But, more importantly, it chimed exactly with what we’ve been hearing ourselves, and have repeatedly drawn attention to. Over the past few months, derivatives users and regulators in the Asia region have become increasingly concerned – and increasingly vocal – about the timing of settlement for collateral posted under US and European margining requirements.

A quick word of explanation. As part of their forthcoming rules for the margining of non-cleared derivatives, US regulators have stipulated that initial and variation margin has to be settled on the day after execution of the trade, or T+1. Europe has taken a similar route in its proposed rules – although the final text now won’t be published until the end of the year following the announcement of a delay earlier this month.

So, why is this a big deal? In short, because timezone differences make it more or less impossible for derivatives users in Asia to meet the requirement when trading with counterparties further west. A margin call at 9am in New York, for instance, would be 9pm in Hong Kong, 10pm in Tokyo and 11pm in Sydney.

There is a way round this: the Asian counterparty could pre-fund margin to cover the time lag. But posting extra collateral comes at a cost – a “timezone tax” as Risk puts it. Asking Asian firms to pay more than other counterparties when trading with US banks, purely because of where they are located, seems deeply unfair to many Asian firms and even regulators.

Risk speculates on the ultimate outcome:

Asia-Pacific dealers have already voted with their feet with regard to the US Dodd-Frank regulation and opted to trade with European counterparties instead. If T+1 is imposed on cross-border trades, this trend will accelerate and expand to include Europe’s lenders.

Japan’s dealers don’t want to see a shift to Asian autarky, but the biggest losers in this scenario could well be Europe and the US.

This comes on top of cross-border challenges caused by the European delay. Global regulators have made real effort to ensure the margining framework in each jurisdiction is as closely harmonized as possible. After all that work to harmonize the national rules, it would surely be a failure if the end result is even more fragmentation.

Novo Banco: An Exceptional Story

Everybody loves a good credit derivatives story. And market participants and the media have had a very good one in recent weeks. It concerns CDS contracts written on Novo Banco, a Portuguese bank.

The Novo Banco story begins with Banco Espírito Santo (BES). The central bank of Portugal, in its capacity as a resolution authority, transferred various assets and liabilities from BES (which was in difficulty) to Novo Banco (a so-called good bank) in August 2014, under the Portuguese bank resolution regime. Sixteen months later, the central bank took the decision to re-transfer five senior bonds back from Novo Banco to BES, reportedly because the European Central Bank’s stress test had uncovered a capital shortfall at Novo Banco. To say this is unusual is an understatement.

Let’s now turn to the credit derivatives market. Credit derivatives contract terms set out the conditions for a credit event to occur, typically using the ISDA Credit Derivatives Definitions. Decisions about whether an event meets those conditions are made by the ISDA Credit Derivatives Determinations Committees (DCs). These committees, which each comprise 10 sell-side and five buy-side firms, make their determinations by gathering publicly available information and comparing it against the definitions to see if the relevant conditions are met (you can read more about the process here and here). A supermajority (12 out of 15 votes) is required to reach a determination. In the case of Novo Banco, the European DC was asked to resolve whether the transfer of bonds from Novo Banco back to BES constituted a governmental intervention credit event.

One of the cornerstones of the Credit Derivatives Definitions is that they are as precise as possible from a legal perspective to enhance predictability and objectivity. This protects both buyer and seller. But like any contract, unanticipated events occasionally emerge that aren’t neatly covered by the definitions. In the case of Novo Banco, the majority of DC members voted that the bond transfer did not constitute a credit event, but the majority fell one short of the supermajority threshold. As a result, the issue was (as per the DC rules) referred to an external panel of experts.

The panel unanimously agreed with the majority of the DC. The decision hinged on whether the transfer constituted a mandatory cancellation, conversion or exchange, or whether the transfer had an analogous effect to those defined events. Ultimately, they determined the transfer was neither a cancellation, conversion nor exchange, and was sufficiently different to those events to be not analogous to them. Taking a broader, catch-all interpretation of ‘analogous’ would mean this clause would “dominate the whole of the definition, which is inconsistent with the careful and detailed drafting”, the external panel decided.

So there it is. While most CDS credit event determinations in practice are clear cut, it’s clearly challenging (and perhaps impossible) to consider and explicitly address all possible future scenarios and contingencies that might occur in the credit markets. That’s why it is important to have a robust process (which includes industry definitions drafting committees, as well as the DCs and the external review panels) through which issues and uncertainties can be addressed and clarified. This enables market participants to gain the clarity they need and deserve, even in exceptional situations such as the one involving Novo Banco.

The Not-So-Secretive Circle

Thomas Jefferson once remarked: “When the press is free and every man able to read it, all is safe.” He knew how important an independent, critical press is to a free society.

He also knew the democratic process could be messy. “Error of opinion,” he said, “may be tolerated where reason is left free to combat it.”

It was in this spirit that we recently read and analyzed a Bloomberg News piece on ISDA’s Determinations Committee (DC) process. The more that people understand how the process works, the better. If the coverage leads to discussion of potential improvements (in addition to those that have been and are being made), then we welcome it. Transparency is essential to building trust in the DC process. That’s why the DC members, the DC rule book and every vote taken in every credit default swap (CDS) credit event is available to anyone on our website.

By the same token, though, when news coverage contains what we see as errors (either of opinion or of fact), then we need to exercise our right to point these out and correct them. Here are three worth noting.

The first is the article’s headline: Inside the Secretive Circle that Rules a $14 Trillion Market. We know headlines are meant to draw in readers. And we also know that Wall Street conspiracy theories are popular. But we’re not sure what the secret is here. The questions asked to the DC, the publicly available information used to make the determinations, and the CDS Definitions themselves are all available on ISDA’s website. A supermajority vote is needed for a credit event to be determined – that’s 12 out of the 15-strong panel of buy- and sell-side participants. The decision is made public. How each firm voted is also made public, which means these entities have to be sure they can justify their decisions internally and externally (and that includes to regulators). And where a supermajority decision is not achieved, the decision is referred to an external panel – the last of which was videoed and published on our website.

More importantly, though, the headline and the article make the credit event process seem arbitrary. They ignore the fact that the legal definition of what constitutes a credit event under an ISDA CDS contract is spelled out in ISDA’s CDS Definitions. The DC’s job is to look at publicly available information about a CDS reference entity, benchmark it against the Definitions and determine whether a credit event has occurred. The fact these decisions are taken by market participants with knowledge of the market and the sometimes-complex documentation used means these decisions can be taken quickly – important, as buyers of protection need to know whether their hedges will be effective. Uncertainty for months on end would dramatically reduce the effectiveness of this instrument as a hedge.

The second point relates to a quote offered by an “industry expert” that the article uses to advance its narrative: “You’ve got a self-regulatory organization that has handed authority over an entire market to those folks who have the greatest self-interest and have no prohibition for putting their interests ahead of the broader market.”

The third point relates to another quote in the story: “The problem is there’s no ability for an independent body to determine whether or not the process is fair.” This comes from the CEO of a “Washington-based non-profit watchdog group”.

The statements are so wrong in so many ways that it is difficult to figure out where to start in addressing them. But we will try: (1) ISDA is not an SRO; (2) no one handed anyone anything; (3) the CDS market is regulated; (4) in addition to regulatory oversight, market participants are also subject to fraud and anti-manipulation laws.

In other words, regulatory authorities are able to see the exposures and votes of the firms they supervise for each and every credit event.

Because the questions asked to the DC are published on the website, along with the publicly available information used to make the determinations, the CDS Definitions that set the legal definition of a credit event, and the votes of the DC members, anyone can do their own analysis and take issue with how a firm voted.

Now to be clear, we don’t think the DC process is perfect. While we do believe it’s robust and transparent, and that it has worked very well, we’re always thinking of ways to improve and strengthen it.

For example, the potential for a conflict of interest was recognized from the start. Rules were put in place to mitigate this issue (including the requirement for supermajority voting and transparency over how firms voted). But market sentiment has evolved since the DCs were created, and so additional measures are now being taken to further guard against conflicts. We think the ability to consider and make such changes underscores the strength of the DC process.

There’s more here that we could say, and we suggest that readers who want additional information click on the following links:

The Credit Derivatives Mailbag

Storm Warnings

The DC rules

Video of external review panel hearing

Summer Reading

We always like to pack a few good yarns to enjoy while on vacation in August…and this one is at the top of the list. It’s a tale of vanishing trades, shadow banking and international intrigue.

Kind of like The Da Vinci Code meets Wall Street.

Except…the trades in question did not vanish (the article admits as much, despite the scary lead) and nothing was done in the shadows.

The fact is, derivatives trades are required to be reported to trade repositories to ensure regulatory transparency, whether they are executed in the European Union or US. And a large proportion (approximately 75% of interest rate derivatives) is centrally cleared globally.

So why all the fuss?

The root of the issue lies in the need for greater cross-border harmonization of derivatives regulations – an issue that we’ve regularly flagged in research reports, magazine articles, speeches and testimonies.

The article claims US banks are shifting derivatives trades overseas in order to benefit from “weaker” regulation in Europe and elsewhere. It doesn’t mention that all trades with a counterparty classified as a US person have to comply with Dodd-Frank, irrespective of whether those trades reside at a US bank or an offshore affiliate.

The crux of the issue, then, is those trades conducted by US banks with non-US counterparties. These clients have to comply with their own country’s rules, and often prefer to trade with counterparties in their own jurisdiction to avoid the compliance burden of meeting multiple sets of regulations simultaneously (ie, their home rules and the regulations of their counterparty’s jurisdiction). Many banks have therefore organized their operations in order to serve those clients. As a result, some trades with non-US clients are being conducted through non-guaranteed, non-US affiliates. (The obligations of a non-guaranteed affiliate are not guaranteed by the parent company.)

It’s important to note that the trades in question are subject to the regulatory requirements of the location in which they and the affiliate are based. The affiliates are also subject to the regulatory and capital mandates of their host jurisdictions.

Critics, however, believe this is merely an attempt to avoid Dodd-Frank regulatory requirements and instead benefit from weaker regulation overseas.

Now, the article doesn’t actually give any specific examples of how the regulatory framework on one side of the Atlantic is ‘weaker’ or ‘stronger’ than the other. Both jurisdictions have implemented, or are in the process of implementing, the commitments made by the Group of 20 nations in 2009 to reform derivatives markets. Both have introduced regulatory trade reporting requirements. Both have committed to mandatory clearing of standardized derivatives. Both jurisdictions have passed bank recovery and resolution legislation to ensure troubled banks can be wound down in an orderly way without resorting to public funds. And stricter capital and margin rules have been developed at a global level and are intended to apply to all banks. Our recent report on progress in regulatory reform details the advances that have been made.

If there is a problem here, it’s the need to ensure regulatory frameworks are consistent across borders to support deep, liquid global markets. Global regulators have determined that their respective rules are equivalent in some areas, which would allow a counterparty to trade with an overseas entity and comply with equivalent overseas rules, without the need to also comply with its home regulations. However, equivalence/substituted compliance determinations have stalled in some cases because of technical, highly granular differences in the two sets of rules.

ISDA believes greater harmonization on the detail of the rules would help. But these equivalence decisions should be based on broad outcomes, rather than word-by-word comparisons of two sets of rules. Only then will the incentive for a fragmentation of liquidity pools be removed.

We know that story isn’t exactly a summer blockbuster. But non-fiction rarely is.

Hot, Hot, Hot!

The Basel Committee on Banking Supervision’s rules for determining the amount of capital relief banks can achieve by hedging credit risk have most likely been called a lot of things by tired and frustrated risk managers over the years. ‘Hot’ is probably a new one.

That, though, is exactly how the Wall Street Journal describes hedges that result in capital relief in a recent article (The Hot Thing for Wall Street Banks: Capital-Relief Trades). The starting point is a report published by the US Office of Financial Research back in June, which considers the use of credit default swaps (CDS) by banks. The WSJ picks up on some of the points in this report, and contains a variety of views on the extent and use of so-called capital-relief trades.

It’s an interesting article on an interesting topic. So we thought we’d add a few additional observations, building on some of the comments we made in our last media.comment post.

First, banks need to continually monitor and manage the risk posed by their loan books and other exposures. Left alone, these loan books would likely reflect the geographic and sector characteristics of their client base. Without active management of their portfolios, banks would quickly reach concentration limits, restricting their ability to lend further. Managing risks such as these is a primary motivation for banks when deciding whether to hedge.

Second, as the WSJ article points out, regulators have long allowed banks to obtain capital relief if they hedge the risks they are exposed to. The reason is intuitively quite simple. If a bank makes a loan to Company A, it’s exposed to the risk that Company A may go bankrupt and not be able to repay the loan. It has to hold capital against that risk – the riskier the borrower, the more capital the bank has to hold. By putting on a hedge, the bank is less exposed to the credit risk of the borrower, and so is able to reduce the amount of capital it has to hold against the loan. Buying CDS protection is one of the many ways banks manage the risk of their loan books, but there are strict rules governing the amount of capital relief banks can obtain.

That’s not quite the end of it, though. Banks have to hold capital against the CDS protection they’ve purchased too. That is largely determined by the counterparty credit risk posed by the protection seller. But capital requirements for single-name CDS transactions have increased across the board under Basel III.

Capital isn’t the only line of protection. The vast majority of these CDS hedges will also be backed by collateral (unlike many of the original loans). This collateral is meant to cover the possibility of a loss arising from one of the counterparties failing to meet its obligations under the derivatives trade. Under new margin rules for non-cleared derivatives, these trades will be subject to mandatory initial margin as well as variation margin from next year.

Turning to the issue of transparency: how do we know how much capital relief US banks have obtained through their CDS hedges? There are various ways. For one thing, US banks submit quarterly filings to the Federal Reserve, which includes data on the amount of credit protection recognized for regulatory capital purposes. As we pointed out in our last media.comment post, this information is publicly available. Second, those banks that apply the internal ratings based approaches to Basel III – so, pretty much all the big Wall Street banks mentioned in the WSJ article – have to publicly disclose the impact of credit risk mitigation techniques on risk-weighted assets (RWAs) within their Basel Pillar III disclosures (see page 32 of this document). Smaller banks using the standardized approach aren’t required to report pre- and post-credit derivatives RWAs, but they do have to disclose a variety of other information related to credit risk mitigation techniques.

A final point. Regulators can also discover the identity of the counterparties that have sold the CDS protection to the banks. Under the Dodd-Frank Act, all derivatives trades have to be reported to a trade repository, giving regulators the ability to scrutinize these transactions down to the counterparty level.

Summer Relief

It’s been some four months since our last posting, and we have been hoping for an equally quiet summer. But then we came across a recent article in the American Banker about a new research report by the US Treasury’s Office of Financial Research.

The story and the report are on banks’ use of credit derivatives to reduce their capital requirements. It’s an interesting topic, to be sure, and it’s one that deserves proper consideration and analysis.

So in that spirit, let us offer some thoughts.

Banks constantly search for ways to manage their risk exposures and allocations, and they have different strategies for doing so. They might, for example, make loans and hold them to maturity, and/or they might sell those loans and recycle capital to make new loans, and/or they might buy credit default swap (CDS) protection on those loans to free up capital for new purposes.  All of this is important for economic growth.

If a bank’s risk management activity reduces its exposure (and hence its risk), it follows that its capital requirements should also be lower. That’s the primary motivation behind the credit derivatives activity in question. And it’s long been recognized as legitimate by prudential regulators.

Now, as the report and article state, banks are required to disclose to the Federal Reserve the notional amount of credit derivatives for which they “purchased protection that is recognized as a guarantee for regulatory capital purposes”.

It’s important to note that this information is available for all to see at the National Information Center (which houses data collected by the Federal Reserve System). The data is available on an individual firm basis (and can be accessed under the ‘Institution Search’ tab), as well as on an aggregated basis, segmented by peer group (under the ‘BHCPR Peer Reports’ tab).

What this means: regulators, shareholders and others can see exactly how much CDS protection a bank has bought to improve its regulatory capital position.

What about the counterparties to those credit derivatives trades? Do we know who they are? Firms in the US, European Union and other jurisdictions are now required to report their derivatives trades to trade repositories. In this way, as well as through the normal course of supervision, the answer is yes – regulators have the data in order to see who those counterparties are.

Also worth noting: under the bank capital rules, a bank that enters into CDS transactions needs to take the creditworthiness of its counterparties into consideration in determining the capital it needs to back the trades.

Another key point: under the new margin rules coming into effect, counterparties are required to post initial and variation margin on all non-cleared derivatives trades. Margining helps mitigate counterparty risk; it works to help ensure that a buyer of protection will actually get paid by a seller of protection.