Sponsored by IHS Markit
In preparing for UMR compliance, how much of what you need to do can be seen as a tick-box exercise? Overall, how would you describe the challenge of preparing?
Cristina Grigore: It’s quite complex. Some aspects can be looked at as ‘tick the box’ exercises: Have you identified all the in-scope counterparties? Have you done all your legal documents? Have you set up your custodial relationships? Once you have the framework, do you have the operational capacity to support initial margin?
Once you’ve done all of that, the important question is, “How do you calculate the requisite amounts?” That’s where it’s not such a straightforward tick-the-box exercise, because there are a lot of nuances. First, how are you going to calculate your initial margin amount? Second, are you going to breach the 50 million threshold? You need to look at these points from a long-term perspective. So, there are grey areas. It’s not so straightforward.
Are there any areas that, in your experience, people tend to overlook, or not give enough importance to in their preparations?
Cristina: Definitely calculations. There are two main methodologies, the most popular being SIMM though Grid or Schedule also receives a lot of consideration, and whatever methodology is chosen in the end it needs to be agreed with your counterparties. Choosing a methodology depends on how the portfolios are constructed.
In-scope entities really need to understand their portfolio composition and test with counterparties. If you do that work now, you can understand what’s going to hit you come September 1st this year or next year.
Let’s say you have two counterparties who are following SIMM as closely as it’s humanly possible to do. Certain things are standardised and agreed. How important are the possible variations that are not standardised?
Hiroshi Tanase: It’s definitely good news that the industry has a standardised model in SIMM, but what’s not standardised is how the input to SIMM is calculated. By inputs, we mean the risk sensitivities. There’s no standardisation in how those sensitivities are to be calculated.
There can be differences between counterparties or calculation agents in both the models and market data that are used to produce the risk sensitivities. So, for example, options can be valued differently due to, say, volatility inputs being different between the counterparties or calculation agents. And for more complex products or exotics, even the choice of model may be different.
And how relaxed are regulators about these differences in inputs?
Hiroshi: Regulators are unlikely to take a prescriptive approach when it comes to the calculation methodologies of the risk sensitivities. But, ultimately, they care about the soundness of the collateral management process or the collateral exchange between firms as a whole, because UMR is an important element of the G20 initiatives set out after the financial crisis. Unless the collateral exchange takes place on a sound basis between firms, the main G20 objectives won’t be met. So, in that sense, the regulators will care that the calculation is done correctly, or accurately enough, between the firms involved, so that the final collateral exchange takes place without any material problems.
There will inevitably be some differences. Is there anything you can do short of adopting the same calculation methodology exactly to minimise or pre-empt those differences?
Cristina: I think pre-empting is where the work needs to be done. We really think testing in advance to be able to understand the factors driving the initial margin numbers will help with that. The methodology for initial margin is a risk-based methodology, quite different from variation margin, which is based on mark- to-market. And that means that you need to run different calculations in advance. The more diverse your portfolio, the more types of asset classes you have in your portfolio, the more testing needs to be done, because they will all have different risks sensitivities allocated based on the instrument type. You need to think of current trading strategies and future ones and find a few friendly counterparties willing to test with you to ensure you minimise the disputes once you go live.
And it’s not just the final numbers you need to consider. The numbers are the outcome. To get to that outcome, no matter what method you choose to do those calculations, you need to ensure that you have all the right inputs and a good database where you have stored all the trade details and from which you can easily export them.
In the event that differences and therefore disputes do arise, does the industry have a good way of dealing with that situation?
Cristina: There are tools offered by the industry and there’s a lot of very good guidance from ISDA as well. What is probably lacking a bit is that practice from an in-scope entity’s point of view. Yes, you have all the information there, but how do you apply it? Does your collateral team have a good understanding of what needs to be done?