Margin rules

Definition

US securities margin is regulated borrowing: Regulation T (Federal Reserve) sets INITIAL requirements — 50% for long equity purchases — and FINRA Rule 4210 sets MAINTENANCE minimums (25% long, higher for shorts and concentrated positions), with brokers free to impose stricter “house” requirements at will. Facts, never advice: this entry states the mechanics the engine must model — buying power, margin calls, forced liquidation, and the portfolio-margin alternative — because leverage mechanics decide outcomes at exactly the moments theses are tested.

How it works / structure

• Reg T initial margin: 50% of a long equity purchase may be borrowed; short sales require 150% (proceeds plus 50%); options are generally not marginable as collateral under Reg T (long options paid in full).
• Maintenance (FINRA 4210): equity must stay above 25% of long market value (30-40%+ typical house levels; short positions 30%+ with per-share floors); breach triggers a MAINTENANCE CALL — deposit or the broker liquidates, WITHOUT being required to give notice or choose positions the client prefers (the agreement’s fine print, uniformly).
• Portfolio margin (the risk-based alternative): FINRA 4210(g) — margin computed from scenario stress (typically ±15% equity moves) on the NETTED book (port-exposure-netting recognition); hedged books margin far lower; minimum account equity applies ($100k+ typical); leverage capacity rises with hedging quality, and so does the sophistication assumed.
• Futures margin is different plumbing: performance bonds, not borrowing (ms-futures-margin); daily variation settlement; no Reg T.
• Engine-executable facts: per-position margin class, house-requirement snapshots, buying-power computation, and the liquidation waterfall assumption (in stress the model assumes the broker liquidates the most liquid collateral first — worst case honesty).

When it applies

Every leveraged or short position (the rules ARE the position’s failure mechanics); options spreads (defined-risk structures margin at max loss — the platform’s preference has a margin-efficiency leg); sizing frameworks (risk-fixed-fractional computes on equity, never on buying power — the distinction is the entry’s practical core).

Risk profile & failure modes

• Forced liquidation at the low: margin calls cluster at drawdown extremes — the mechanism converts paper drawdowns into realized bottom-ticks; sizing below call-distance is the structural counter.
• House-requirement surprises: brokers raise requirements on volatile names mid-episode (documented in every squeeze) — the call arrives from the requirement change, not the price change.
• Buying-power illusion: 2:1 (or 6:1 portfolio-margin) capacity treated as sizing guidance — capacity is a ceiling set by regulation, not a recommendation set by evidence.
• Cross-margin contagion: one position’s loss liquidates unrelated positions in the same account — the account, not the position, is the risk unit.

Evidence & limits

Reg T and FINRA 4210 are the primary-source rules (cited); broker house policies vary and change — the platform treats them as per-broker data, refreshed, never assumed. Margin- call clustering at drawdowns follows mechanically from the rules; no behavioral claim is needed.

Falsifiable-thesis examples

Illustrations only, not signals:

• “This book survives a ±15% single-day index move without a maintenance call at current house requirements (call-distance audit)” — falsified by the scenario computation.
• “Portfolio margin reduces this hedged book’s requirement by at least 40% vs Reg T (netting-recognition check)” — falsified by the paired computation.

Cross-references

• The futures counterpart: ms-futures-margin
• The day-trading overlay: acct-pdt-rule
• The netting it rewards: port-exposure-netting
• The sizing doctrine it motivates: risk-fixed-fractional; account context: acct-account-types