The Uncleared Margin Rules (UMR) Phase 6 deadline is fast approaching, with the International Swaps Derivatives Association (ISDA) estimating that 775 entities will come into scope for the last phase of Initial Margin (IM) rules. Phase 5, which had a fraction of the number of firms coming in scope and more significant operational resources, already saw hold-ups with regard to documentation, compliance, and custodial set-ups. The leaner buy-side entities, caught by Phase 6 are unlikely to have that luxury and are running out of time, as highlighted in a recent article from ISDA on IM Compliance which discusses the key themes for becoming UMR compliant.
Aside from these steps, one of the key elements for evaluating readiness for the UMR deadline is understanding which margin calculation approach is relevant for your firm. This article details the main considerations when deciding whether to use the Standard Initial Margin Model, commonly referred to as SIMM, or the GRID approach for margin calculation. We also analyze the different approaches available, using sample portfolios.
In negotiating agreements with counterparties, an important factor is to understand whether to choose the SIMM or GRID method of calculation This can have an impact on the margin that builds up over time and can be a significant decider; not only when you breach the USD 50 million thresholds (or jurisdictional equivalent), but also relevant once you are aware of how much collateral you will be required to post.
Cassini assists clients to look at this as a snapshot in time based on existing portfolios, or an ongoing basis, and how the IM amounts may look on UMR-compliant portfolios as the book builds over time. Cassini has assisted several clients in making the most efficient choice in this matter. Some of the key themes that arise are –
This is the biggest driver in the choice of using SIMM or GRID. Under SIMM, each trade gets repriced to generate sensitivities under the different risk buckets defined by ISDA. The respective sensitivities across each netting set will have a netting benefit before running through the SIMM model to produce a final IM number. Whereas, for GRID it is a function only of the trade notional, product type, and the Mark-to-Market (MTM) on the portfolio.
GRID is additive across trades regardless of any netting or hedging between them.
Another key factor is portfolios that contain long-dated rates or inflation swaps. SIMM will increase with the tenor of the trades as the Interest Rate delta or Inflation delta increases. In addition to this, the longer tenor buckets carry higher risk weights in the SIMM formula. In contrast to this, GRID charges a rate based on the product type only; therefore, we see vast differences between the two approaches where this is a driving factor of the portfolio’s risk.
One thing counterparties often overlook is that you don’t have to choose GRID or SIMM for the entire agreement. You can vary this based on product types within an agreement with your counterparties. This allows you to optimize the mix of SIMM and GRID calculations, given the types of trades that you have in your portfolio.
GRID is not just a function of trade notional and margin rate; it considers the MTM at the portfolio level and individual trades to have a net replacement cost factor before calculating the IM number. For deeply in-the-money or out-the-money portfolios, this can have an impact.
We used a sample set of trades to determine how this would look for a one-directional rates portfolio. Using a portfolio with 12 billion USD in notional on SOFR Vanilla Interest Rate Swaps, a tenor of 20 years; we varied the portfolio with trades at-the-money, in-the-money, and out-of-the-money to create a variety of scenarios where the overall MTM or net-present value of the portfolio varied across the range. See the below table for 5 scenarios across SIMM and GRID.
In all the scenarios, GRID ranged from between 1.5% to 3.5% of gross notional across the trades. SIMM, however, ranged between 12% to 13% of the gross portfolio notional. The GRID amounts varied significantly with the base scenario as the MTM of the trades and portfolio varied.
This only further solidifies how MTM, and replacement cost, impact GRID calculations, which are not always considered whilst making these decisions. On the SIMM side, the IM charge did not vary too much across the different scenarios, with the determining factor being the long duration and directional nature of the trades.
The current CME charge for the base scenario (1) is 50% cheaper than the SIMM IM amount. Although surprisingly the GRID margin is still only 20% of the cleared CME margin.
This gives a lot of food for thought. You need to consider all the factors before deciding on which model to choose and if you intend to apply GRID / SIMM for the whole portfolio or only specific trade types.
The key driver of this discrepancy is that when there is no offsetting risk between a product class SIMM will be additive. When there are offsetting risk trades, the SIMM number will give you a netting benefit; conversely, the GRID IM will remain constant.
Overall, it is very important to consider not only how your existing portfolio behaves under the two approaches, but also how the portfolio will be evolved post-go-live. This includes what type of trades will make up the portfolio and what clearing routes will be available to trade.
By knowing this, when you are negotiating your counterparty agreements, you have a forward-looking approach in mind for ensuring you have the most optimal margin agreements in place for your business.
In addition to SIMM vs GRID analysis, some of the other key features of Cassini that clients are using in the UMR space are: