SIMM vs grid approach for IM — which should we use?

Question

AcadiFi · Accepted Answer

SIMM (Standard Initial Margin Model) is an ISDA-developed risk-sensitivity-based model that computes IM from delta, vega, curvature, and base correlation sensitivities across asset classes (interest rate, credit, equity, commodity, FX). The standardized grid applies fixed percentages to gross notional by asset class and residual maturity. Tradeoffs. ```mermaid flowchart TD A[IM Method
Choice] --> B[SIMM
Risk-Based] A --> C[Grid
Notional-Based] B --> D[Lower IM
Complex Ops] C --> E[Higher IM
Simple Ops] ``` SIMM advantages: recognizes netting and hedging within and across asset classes, typically produces 30-60% lower IM for well-diversified books, industry-standard so counterparties rarely dispute. SIMM disadvantages: requires daily computation of thousands of sensitivities, annual recalibration of parameters (required), complex operational infrastructure (risk engine, dispute resolution, parameter updates), and subject to ISDA governance process. Grid advantages: simple to implement, deterministic from trade terms alone, no calibration needed. Grid disadvantages: no netting recognition within asset class, higher IM (often 2-3x SIMM for multi-product portfolios), can become prohibitive for directional books. Practical decision: firms with $100M+ derivatives books typically choose SIMM because the IM savings easily justify build costs. Firms with small, simple books use the grid. Counterparties must agree on the method bilaterally. For Argent Stonehill Markets with a complex rates-credit-FX book, SIMM might save $35M versus grid, making the $2M/year operational cost trivial. Test both on your book before committing.

SIMM vs grid approach for IM — which should we use?

Master Part II with our FRM Course

Related Questions

Practice Questions