The third phase of BCBS-IOSCO’s margin requirements for cleared and non-cleared derivatives will go live on September 1, 2018. In a recent DerivSource webinar, Shaun Murray, Managing Director, Head of Strategic Collateral Management at Standard Chartered Bank, spoke about the firm’s experiences with the Standard Initial Margin Model (SIMM) 2.0 implementation as a phase one dealer, and what firms can expect from future versions of the model. His comments are summarised here. To listen to the on-demand version of the webinar, click here.
Standard Chartered is a phase-one dealer and has been live with SIMM since September 2016. The dealer currently has 80-plus relationships with the ability to exchange initial margin (IM), and with the third phase going live in September 2018, there will be up to 50 additional relationships that will need to be re-papered.
The SIMM model is working well for the firm, which is able to perform daily back testing and benchmarking, and work across the competing priorities of meeting time-bound regulatory requirements, prioritising new products according to front-office demands, and operating within general resource and budget considerations.
Standard Chartered spent two years building its SIMM model. When considering whether to build or buy, it is essential for firms to be aware of their aggregate average notional amount (AANA), the SIMM model is brand new—to both users and the industry in general—firms cannot start early enough with their preparations.
Who to consult as firms prepare for SIMM implementation
Model development and business as usual (BAU) processing involve a multitude of different people, and there needs to be a lot of internal communication as firms decide what their SIMM model plan will look like. Equally important, some of the biggest pain points are external. Does the firm have a custody agent? If so, they should be made aware if the firm is likely to be over the IM threshold at some point in the future by looking at their AANA.
If firms have not on-boarded with a custody agent, they should get an understanding of the onboarding process. What does Know Your Customer (KYC) look like? What does Anti-Money Laundering (AML) look like? Is the firm trading through one entity, or through multiple entities or subsidiaries? There are individual requirements for each subsidiary or each entity firms want to execute through. Financial organisations should look at the IM credit support deeds (CSD), terms and conditions, collateral for eligibility, settlement processes. Is the firm in a position today to use tri-party repo? Is it considering third-party or bilateral exchange of IM?
In fact, is the firm even considering wanting to exchange IM? IM is not like variation margin (VM) which is one-directional. It is a two-way gross payment, which makes it a liquidity drain, and a funding cost. The front office needs to be consulted with regards to IM management and implementation strategies. How is the firm going to allocate the costs of posting IM? How is it going to optimise the IM? Will it look at connectivity and where derivatives transactions are executed, and the distribution of hedges? There are many questions the front office needs to understand and answer before financial organizations can make a decision about whether to continue.
On the technological implementation of the model, testing is absolutely critical. ISDA ensured the industry came together and tested with each other before the first phase went live. Firms do not know if their models are similar to the rest of the industry, so they should aim to test with three or four of their dealers, looking at the deepest, most complex portfolios. Waiting until September 1 to see what happens when it goes live is not a sensible option. Firms need to test their model, whether they built it or bought it, and to make sure it fits with what everyone else is doing to avoid unnecessary disputes.
Financial organizations should also consider whether automation is something they would like to do or have to do. A lack of automation will clearly eat up resources and costs. Firms should think about automation as a strategy and a priority, not as a nice to have.
Finally, firms need to have a handle on their legal documentation early. If financial organizations know their AANA, they can start negotiating well in advance of the implementation deadline. It will not be good for the industry if there are thousands of clients that still needing to repaper custody documents, VM documents, or IM credit support annexes (CSAs) in the summer of 2019, or 2020.
The SIMM model must stabilize as smaller firms adopt it
The SIMM model has undergone a lot of change over the last 18 months, with four deliveries plus SIMM 2.0 in 2017. Looking ahead, there need to be fewer changes to the model as the smaller firms start to be phased in to IM requirements. In 2019 and 2020, the AANA threshold for financial organizations exchanging IM will go down, there will be much larger numbers of SIMM users, and the newly impacted financial organisations will likely have smaller derivatives books, and smaller technology budgets and resource pools from which to draw expertise to deliver on SIMM. Consistency of the model will be absolutely critical going forward.
Every US registered swap dealer is required to use the latest version of SIMM. If the SIMM versioning changes four times every year, that will consume significant overhead. Firms have done this for the last 18 months as they needed to get approval, but it is not sustainable given all the complexities surrounding it. At some point, there has to be a settling down process, so financial organizations can consume the SIMM in the knowledge there will only be one delivery that year. It is incumbent on the phase one and two dealers, and ISDA, to make sure the model has been simplified as much as possible before large numbers of smaller firms are onboarding.
Mostly, the changes over the last 18 months were regulatory driven. Generally, the model has been well thought through so there have not been many other changes. Hopefully, as the model settles down and the industry takes stock, users will see that the model is working well, because they are able to reconcile, disputes are low, and the sensitivities are matching.
When it comes to developing SIMM, as well as processing it after delivery and go live, it is important to have the right people, with the right knowledge, and the correct organizational setup to deliver on the technology stack, and then to implement and operationally process it.
From the outset, firms need to identify both a model owner and a process owner. The model owner would be based in either the trading or risk function, depending on how the firm views the model. The process owner could be based in a risk function or an operations function, depending on how the firm is set up. These two people should work in tandem and drive the decision of whether to build or buy.
They need to be empowered to decide who, how and what is going to deliver SIMM, as well as to dictate how their operating model should work front to back. They need to understand where the various SIMM functions sit, who is going to be involved in delivering the SIMM, delivering the technology, and then delivering the BAU operational processes.
This is new ground, and it takes time. Ensuring that the processes are fully understood and documented, and that people understand what they are responsible for is very critical. It is important to get people involved really early in understanding how the SIMM operates, how it is going to be built, how it is going to be delivered, how it is going to sit in a technology stack, and how it is going to be optimal. This needs to be done well in advance of going live.
Talent acquisition will be a challenge as there are still only a relatively small number of firms that have gone through a SIMM implementation and every firm aims to be lean and agile and keep their teams small. A firm’s developers, project, implementation and risk teams might have a general understanding of SIMM, but there will not be large pools of talent with deep SIMM implementation experience available.
“Talent acquisition will be a challenge as there are still only a relatively small number of firms that have gone through a SIMM implementation and every firm aims to be lean and agile and keep their teams small. A firm’s developers, project, implementation and risk teams might have a general understanding of SIMM, but there will not be large pools of talent with deep SIMM implementation experience available.”
Working through which skills a firm requires across departments from front office and technology to back office and other support areas is also challenging. It requires a significant ramp up in resourcing which brings additional cost. As part of the build/buy decision, firms must consider whether they will use permanent employees or contractors.
Firms that choose not to build their SIMM model will be focused on the implementation and governance of the model rather than the development. They will still need to overcome issues around providing sensitivities, which can be challenging for many financial organizations. That is one of the reconciliation points they need to be aware of, as well as ramping up in development and model development implementation and validation. They need to understand what happens with legal documentation, in margin operations, in settlements across all the risk disciplines, as well as in the audit function. And they need supporting documentation around all of this.
ISDA SIMM may be an industry model, but regulators will still want to see that firms have tested it and gone through the validation process, and are able to show that they have standards and documentation around model development and performance as well as model governance escalation. Much will come down to the firm’s internal appetite, policies, and processes to their regulatory jurisdiction. The communications with dealers, hedging counterparts and regulators will differ from firm to firm.
Funding is another element that must be understood in the implementation stage. The front office has to be engaged in this process, and to understand that there will be a P&L event when they deal with a derivative that is non-cleared. They already understand this for cleared derivatives, because they post VM and IM to their clearinghouse.
This leads us to the margin valuation adjustment (MVA). The majority of the industry is not currently pricing in MVA. Most people are either building it or working through it now, but firms will need it in order to balance the costs of trading OTC derivatives into their P&L. Whether they do that on an individual basis, trade basis, book basis, portfolio basis, or just on a business basis, that is part of their implementation decision making. People need to understand whether they are going to get charged, whether it is going to be flat, or if it is going to be swallowed up into a black hole. There are many different ways of going through that.
Some firms may look to remain below a certain IM threshold with their individual dealers, or even stop trading if they decide the margin costs are too burdensome. The front office will have to choose carefully where they hedge and where they want their funding cost to be. There may even be different documentation requirements if firms are able to stay under their minimum thresholds, but this is something the industry is still working out.
There are several vendors providing services to help firms decide where to put their hedges and manage IM to ensure they stay below their AANA thresholds with each dealer. Some position themselves as industry utilities, others are more like tracking systems that help financial organizations optimize their trading and hedging strategies. Phase 3 and 4 firms should start sending out their requests for information (RFIs) and requests for proposal (RFPs) to work through their various options for managing IM requirements as soon as possible.