Whether you're staring down a call this week or just want a plain-English answer without waiting on the risk desk — MarginLens is built for both.
Getting from a margin number to a decision a trader can act on has traditionally meant two separate jobs, often run by two separate teams, on two separate timelines.
A vendor, clearinghouse, or in-house quant model runs the actual initial and variation margin calculation — SIMM, SPAN, CCP methodologies, or an internal model.
A quant analyst manually turns that raw figure into something a trader or client can use — breaking it down, explaining what moved it, building the report. Often a day or more of turnaround.
MarginLens collapses step 2 into instant, self-service intelligence — everything a quant analyst used to do by hand:
Upload the margin figures you already have and get ahead of the call.
Step 1 — the calculation — still runs on the source you already trust, connected however it lives: file upload, database, or REST API. Today, MarginLens is where the number becomes a decision — instantly, not tomorrow.
Margin requirements are essential — but across most financial firms, they're still managed through manual workflows and fragmented systems. The cost of that gap is measurable.
DTCC raised margin collateral requirements to 100% overnight4, forcing Robinhood to halt trading and Melvin Capital to absorb a $2.75B emergency capital injection. Firms that lacked real-time margin stress-testing had no time to respond.5
Institutional desks — banks, prop shops, hedge funds, asset managers, investment managers, and family offices — are still managing margin exposure through fragmented spreadsheets and manual workflows. The exposure is real. The tooling hasn't caught up. MarginLens was built to close that gap.
1 FINRA, margin statistics, recent data 2 ISDA, margin survey, March 2020 3 LCH / CCP industry data, March 9 2020 4 DTCC margin notice, January 28 2021 5 SEC/DTCC filings, January 2021
Margin intelligence for institutional desks: analytics based on standard industry margin methodologies, plain-English queries, and scenario stress-testing — purpose-built for traders, risk, and ops teams.
Purpose-built by quants, engineers, and former trading desk professionals.
Your data, organized. Total IM, variation margin, and utilization at a glance — OTC and cleared margin broken out by position, plus an IM-by-position chart.
Counterparty risk scoring. A machine learning model scores every counterparty's netting set — call probability, alert tier, and top driving factor, with full feature attribution.
Cheapest-to-deliver allocation. Ranks the collateral you hold by funding cost and allocates the cheapest mix that satisfies each counterparty's IM requirement.
Stress-test before it happens. Adjust position size and a market shock to see directional margin impact before markets move.
Natural-language queries. Ask questions like "which positions are closest to a margin call?" or "explain my binding stress scenario for ES." Grounded in your own margin data — not a generic chatbot.
Structured exports. Six report types: VM, Cleared Margin Decomposition, OTC Margin Decomposition, Risk Sensitivity Export, Historical Margin (Illustrative), House IM. Account-stamped and generation-timestamped for internal review.
We're onboarding founding customers across banks, prop shops, hedge funds, asset managers, investment managers, and family offices — priority onboarding, preferred pricing, direct founder access. If margin intelligence is on your radar, let's talk.