Every week the originator delivers a loan tape: every loan, its dates, amounts and
payment history. The pool compares what was collected against what was expected.
If collections fall below 90% of expectation, the pool halts new purchases by
itself and routes all incoming cash to repaying investors.
No vote, no delay, no discretion. The stop is automatic.
Why automatic matters
A discretionary stop is a stop that gets argued about. By the time a committee has
agreed that collections look bad, more capital has been deployed into a book that
was already deteriorating.
Removing the discretion removes the argument. It also removes PRISM’s ability to
be persuaded to wait, which is the point.
What happens after it fires
The pool stops buying new receivables and switches to repayment mode. Incoming
cash goes to investors in waterfall order — senior first, then down the stack.
Nothing new is funded until the situation is resolved.
What it catches, and what it does not
Catches: a loan book deteriorating faster than expected, and collections that
are quietly short.
Does not catch: a loan tape that does not describe reality. The tripwire
compares reported collections against expectations. If the report itself is wrong,
the tripwire is measuring the wrong thing.
That gap is why independent verification of the invoices by an outside firm is on
the list of things that arrive later — it costs money per pool, so
it comes once there is revenue to pay for it.
Where it has been tested
Firing the tripwire under test is an explicit deliverable of the shadow pool, and
has not happened yet. See
M2 on the roadmap. 
