IM getting serious

Click to open ISDA’s new Initial Margin For Non-Centrally Cleared Swaps Analysis

A key recommendation of the G20 2009 Pittsburgh Communique was to enhance systemic resiliency by reducing bilateral counterparty risk and mandating central clearing of “standardized” OTC derivatives. ISDA and market participants are fully supportive of the G20’s clearing initiatives. Today, more than half the market is cleared and more clearing is on the way.

However, most industry estimates expect that 20 percent or more of the notional value of OTC derivatives cannot be and will not be clearable. Such swaps play an important economic role for the global economy, ranging from housing to corporate financing. Policymakers are hoping to eliminate counterparty risk of these unclearable trades by proposing margin requirements – both initial (IM) and variation margin (VM) – for them.

We fully support mandatory exchange of VM among covered entities. Experience (good and bad) has demonstrated that the practice of frequently exchanging the unrealized mark-to-value fluctuations between two parties is beneficial in reducing counterparty risk. It avoids the build-up of large unrealized positions that could become destabilizing in periods of market stress. This is a widely adopted practice in the OTC derivatives marketplace. And, had this practice been followed – which was NOT the case with AIG and certain monolines in the US – counterparty risk would not be the issue it has become today for the OTC derivatives industry. So, the requirement for VM exchange alone would be more than enough to address counterparty risk concerns.

However, while we support the use of VM, the same can not be said for imposing mandatory IM on counterparties. A rudimentary risk/benefit exercise reveals that the approach is flawed. We see minimal benefits in terms of incremental risk reduction – above and in excess of what is already provided by capital requirements.

Instead, we see huge risks in the making. The outright quantum of margin required under these proposals, even in “normal” market conditions, is significant.

Take, for example, the data that is coming out of the QIS study that the BCBS/IOSCO is conducting. The numbers are revealing. Even if we take the most optimistic scenario (where ALL market participants use some form of internal model to maximize netting benefits), we still estimate incremental margin requirements of approximately $800 billion. This is an amount of incremental collateral which, even under normal circumstances, is challenging to raise.

What’s worse: we know that IM requirements in stressed conditions could go up by at least three times, raising potential requirements at a time when liquidity will be most needed. This is a clear recipe for disaster, only exacerbating systemic risk. And as to the proposed use of thresholds to alleviate the IM impact, at times of crisis, they just make the problem worse. It’s a message we plan to share with supervisors and regulators around the world. (ISDA’s recent analysis of data regarding IM estimates and their impact is here.)

In short, ISDA believes that mandatory, risk sensitive IM will not achieve its purported goals. We instead advocate a three pillar framework for ensuring systemic resiliency: a robust variation margin framework, mandatory clearing for liquid, standardized products and appropriate capital standards.

No Margin for Error

On Friday, ISDA submitted its response to the BCBS/IOSCO proposal on Margin Requirements for Non-Centrally Cleared Derivatives. This is one of the most important issues today in financial regulation. It is not hyperbole to suggest that the future of OTC markets depends on getting it right. Less apparent, but equally true, is that the resiliency of the financial system does too.

Broadly speaking, the BCBS/IOSCO proposal aims to reduce counterparty risk, a goal ISDA shares. But they seek to do it by imposing a universal two-way initial margin (IM) requirement on all covered entities. Covered entities are defined as all OTC derivatives participants except for non-systemically important corporates, sovereigns and central banks. The proposals also call for variation margin (VM) exchange by all covered entities. And if the above were not enough, the proposals call for collateral exchanged to remain segregated and not be re-hypothecated.

It is obvious that the BCBS/IOSCO proposal aims to replicate the mechanics found in the context of clearing. CCPs seek to completely insulate themselves through the imposition of VM (which settles unrealized gains/losses), and IM (meant to provide a cushion to absorb losses that may materialize in trying to cover its risk if a counterparty defaults). By doing so, CCPs effectively neutralize (to a 99% confidence interval) counterparty risk.

And there lies the rub.

In the context of a CCP, such a minimization of counterparty risk is appropriate and the utilization of VM and IM are tools (along with their default funds) used to accomplish this objective. CCPs are not supposed to take credit risk themselves, but to pass through the benefits associated with multi-lateral clearing. However, extending these concepts to the bilateral context is inappropriate, because the parties involved are typically creditworthy entities on their own, and back their creditworthiness with their own capital as well as the proper use of credit mitigants.

So, imposing mandatory margin requirements on bilateral trades would be tantamount to insuring for the same risk twice.

We can understand – and in fact, we fully agree with – the mandatory exchange of VM among covered entities. Experience (good and bad) has demonstrated that the practice of frequently exchanging the unrealized mark-to-value fluctuations between two parties is very beneficial in terms of reducing counterparty risk. It helps avoid the build-up of large unrealized positions that could become destabilizing in periods of market stress. This is a widely adopted practice among practitioners in the OTC derivatives markets. Had this practice been followed by AIG and certain monolines in the US, counterparty risk would probably not have become such a big issue.

While the requirement for VM exchange alone would be more than enough to address counterparty risk concerns, the BCBS/IOSCO proposal goes a step further. In order to further reduce counterparty risk in the eventuality that one of the parties in a bilateral trade defaults, they call for IM to establish a buffer. The buffer – the IM – is meant to absorb any losses realized during the time (10 days) when the non-defaulting party is closing out or replacing the risk it had with the defaulting party. This innocuous step has huge unintended consequences.

In addition to being unnecessary (as this exposure is typically covered by other risk mitigants that the two parties typically agree among them), this arrangement is inefficient. That’s because BOTH parties are required to post IM – but only one defaults. It would also require all covered entities to create new set ups to meet these requirements (be they in the form of new operational processes, cash and collateral and custodian management needs, new agreements etc.). All of this implies higher costs for derivatives users as they will bear the increased cost of doing business, irrespective of whether they are covered entities or not.

Unfortunately, the detrimental effects of this proposal do not stop here. While one cannot argue with the intentions of the regulators to reduce counterparty risk, it seems that the potential implications of this proposal have been grossly underestimated. The combination of requiring the posting of an across-the board two-way IM – which has to be segregated and cannot be re-hypothecated – leads to some very large collateral requirements.

ISDA has performed some preliminary analysis to gauge the potential demand for collateral that this proposal could generate. If this were to be implemented on the existing portfolio of non-cleared OTC derivatives outstanding, incremental collateral demand could run as high as $15.7 trillion.

Now $15.7 trillion may sound like a very large number. It is! But for all practical purpose, any number above a couple of trillion would make the cost of meeting such requirements prohibitive – even if it is feasible at all to locate so much collateral. Collateral demand in such amounts is likely to cause irreparable damage both to the OTC derivatives market, but also to the general economy.

So, once again we have come full circle. Risk does not disappear. It just changes form. Regulators have initiated margin proposals to enhance systemic resiliency by reducing counterparty risk. But in the process of doing so, they are about to decrease systemic resiliency by introducing healthy amounts of liquidity risk (caused by the shortage of collateral) and economic risk (caused either by the shortage of liquidity and/or by all the economic risks that are likely to remain unhedged).