Part 3 of a five-part series on the
signals a retrospective review reads. Earlier:
Coherence. Next:
Inflection.
Growth is not suspicious. A real member who is doing well borrows more as they
earn more. They ask for a limit increase, they carry a larger balance, they
take a car loan, and every bit of it is backed by rising income you can see
arriving. If you flagged accounts simply for wanting more credit over time,
you would flag your healthiest relationships. So the signal cannot be "credit
appetite is rising." It has to be "credit appetite is rising faster than
real income," and staying that way.
The two lines to watch
Picture two lines drawn across an account's life. The first is credit exposure
sought: limits requested, balances carried, new tradelines opened, the total
amount of borrowing power the account is reaching for. The second is genuine
inflow: the real money actually coming in, payroll and the other funding that
provenance taught us to look for.
For a real growing member, those two lines rise together, roughly. The
appetite tracks the income, because the income is what services the appetite.
For a cultivated synthetic, the exposure line pulls away from the income line
and keeps climbing. The account wants more borrowing power than any real money
coming in could ever support. That widening gap is the ramp.
The move that hides the gap
There is a complication, and it is the reason this signal has to be computed
carefully rather than eyeballed. Inflow can be faked, at least on the surface,
by cycling money. Funds get pushed into the account and pulled straight back
out to the same place, or moved in a loop between related accounts, so that
the raw deposit total looks healthy. On a naive read, the income line looks
high enough to justify the borrowing.
So the honest version of the signal removes that first. You net out money that
arrives and departs to the same counterparty in a short window, transfers that
round-trip rather than settle, deposits that never actually fund anything. What
is left is genuine inflow: money that came in and stayed to be spent
on a life. When you measure appetite against that number instead of the gross
one, the ramp that cycling was hiding tends to reappear.
Why this reasoning, not a threshold
Notice that none of this is a fixed cutoff. There is no universal ratio of
credit-to-income that means "synthetic," because the right ratio depends
entirely on who the account is. A high earner can carry a lot of exposure
against a lot of income and be completely ordinary. The signal is not the
level; it is the divergence, read against the account's own income
and its own past. Appetite that outruns real, non-cycling inflow, and keeps
outrunning it, is the shape we mean. A single month above trend is noise. A
persistent, widening gap after you strip out the round-tripping is the signal.
Where it sits in the lifecycle
Ramp is the middle of the story. Provenance was the origin; coherence was the
texture of the spending; ramp is the account straining against the limits of
what its real economics can support. It is louder than the first two signals
and earlier than the last one. By the time the ramp is unmistakable, the
account has usually not yet done the dramatic thing, the sudden run to the
limit, that everyone eventually notices. That dramatic thing is the next note
in the series: the inflection, when the account finally breaks from its own
pattern all at once.
Sources & notes
This is a conceptual,
operator-facing description of one behavioral signal. It makes no factual
claims about any specific institution, portfolio, or case, cites none, and
describes no Delegate engagement or result. The mechanics it names (credit
exposure sought, net versus gross inflow, transfer round-tripping) are
ordinary properties of account data. You can see the signal families run on
a simulated book in the Portfolio Explorer.
Field notes
New teardowns and notes on how synthetic losses hide, sent occasionally.
Occasional field notes on synthetic fraud, plus Delegate product updates. Unsubscribe anytime.