Leverage Decay (J-factor)

Automated exposure reduction driven by market microstructure.

J-factor is Multiply's dynamic risk engine. It continuously monitors depth profile, odds dynamics, open interest microstructure, and resolution conditions, defining when leverage should be reduced.

For users, J-factor is what quietly works in the background to keep your position safe, proactively reducing your leverage down as conditions worsen.

What J-factor governs

J-factor is a set of parameters that jointly define:

• The maximum leverage permitted at any point in the position's lifecycle

• The conditions that trigger reduction: depth thresholds, spread widening, order book imbalance, fill rate degradation, etc.

• The execution path for unwinding the corresponding hedge when a threshold is breached

Dynamic behavior

Because reduction is conditional on microstructure rather than scheduled against time, J-factor produces fundamentally different outcomes on the various sides of a market.

On the winning side, depth typically holds or improves as resolution approaches. Participants want exposure to the likely outcome, thresholds are rarely breached, and positions frequently carry their full initial leverage through resolution.

On the losing side, counterparty depth drains, spreads widen, and fill rates deteriorate. These are exactly the signals J-factor monitors. Thresholds are breached earlier, triggering progressive reduction. Each step unwinds a proportional share of the hedge on the venue, returning UF capital and shrinking the position's footprint against a thinning book. This also protects the user: each reduction pushes the liquidation price further away, letting them stay in the market longer than they would under a static leverage framework.

This microstructure-driven approach has two additional properties.

• First, because there is no fixed exit schedule, there is no predictable pattern that external participants can front-run or position against, resulting in better execution for both the trader and the Facility.

• Second, because the engine does not rely on time-to-resolution to govern leverage, the same architecture that manages a multi-week election market can underwrite short-duration markets of 5 or 15 minutes without modification. What changes is the speed at which microstructure evolves, not the logic that governs the response.