What is the Asset Performance Index?
Every month, every machine in DualMint's portfolio receives a score between 0 and 100. That score is built from IoT telemetry, not operator reports. If it drops below threshold, the operator gets reassigned and the machine keeps running.
The Asset Performance Index is how DualMint determines vault eligibility, sets monitoring intensity for each asset, and keeps yield continuous even when individual operators fail. It's also why 65% of the signal comes from the machine itself rather than the person running it.
What goes into the score
Four inputs: cash flow reliability, asset health, operator execution, and location stability. Cash flow reliability measures whether the machine earns what's expected given its location, size, and asset category. Asset health tracks uptime, maintenance frequency, and mechanical performance over time. Operator execution scores how closely the operator's behaviour matches what the telemetry recorded. Location stability accounts for the site's trading conditions and whether they've changed.
Sixty-five percent of the total signal comes from machine-generated data. The operator's self-reported revenue carries 35% of the weight. That ratio exists because operator-reported numbers have an obvious conflict of interest. IoT telemetry doesn't.
Why the machine gets most of the signal
In conventional asset financing, the lender reviews the borrower's books: revenue reports, bank statements, expense records. All of it submitted by the party that benefits from the numbers looking good. The Asset Performance Index inverts that. The primary evidence is what the machine physically recorded, not what the operator chose to submit.
A laundromat that ran 5,000 cycles in a month has a specific expected revenue given its pricing and location. If the operator reports 30% above that, the discrepancy shows up in the score before any distribution moves. The machine is the check. This is usage-risk underwriting: the asset's physical behaviour, not the operator's creditworthiness, determines whether yield is earned.
What a falling score triggers
There's no manual review when a score drops below threshold. The scoring model defines the floor, and crossing it activates a reassignment protocol. DualMint's operator pipeline surfaces qualified local replacements through three independent channels: active multi-site operators notified of an available asset, regional equipment dealer networks (Speed Queen and Continental Girbau for laundromats), and the peer contacts every onboarding operator identifies at origination.
Redeployment typically completes within 7 to 14 days of a default trigger. The revenue gap is absorbed by the vault's liquidity buffer. The machine keeps earning during transition. A reassignment compresses but doesn't eliminate the affected month's distribution.
The vault gate
New operators don't enter the vault directly. They run in the marketplace first, where machines build an Asset Performance Index score over time. Only assets with a demonstrated scoring history at or above threshold are eligible for vault inclusion. A machine that scores consistently earns its way in. One that fluctuates doesn't.
The marketplace isn't a separate product from the vault. It's the proving ground the vault depends on. Every distribution from the Boring Index Vault draws from assets that have already survived a live scoring history.
Twelve months of data
Since May 2025, the Asset Performance Index has scored every asset in DualMint's portfolio monthly. Twelve consecutive distribution cycles. Zero operator defaults. No score has triggered a reassignment. That's not a test of whether the system handles edge cases. It's evidence that the underlying asset category, physical machines with non-discretionary demand, scores well because it earns well.
The Boring Index Vault opens in Q3 2026. Every asset entering it will have a scoring history. The score is published. The methodology is documented. The machine's IoT data is the verification. That's what 13-15% from a laundromat actually means: not a projection, but a number that has to survive a monthly machine-level audit before it distributes.