Screen

Reserve-adequacy screen
who's persistently under-reserved, relative to peers.

For every covered institution, the loan and lease allowance is compared against the net charge-offs realized over the following four quarters. The first list below shows institutions where that forward coverage ratio has sat below their own asset cohort's 25th percentile for four or more consecutive quarters — a descriptive, reproducible pattern in already-reported figures, not a regulatory rating, failure prediction, or investment conclusion. A second, independent list further down looks at each institution's own trend rather than its peers: a declining dollar allowance alongside rising realized charge-offs. Institutions with a genuinely immaterial reported allowance are excluded from both lists, since the underlying ratios would be noise there, not a signal.

A limitation worth reading before this list, and how it's handled: consumer/card-heavy lenders structurally charge off faster than commercial/mortgage lenders. Each cohort below is split by charge-off velocity before computing a threshold, so a fast-recognizing bank is compared only against similarly fast-recognizing peers — not one flat asset-cohort percentile. Checked live, that correction is real (it moved Capital One's threshold from 2.72x to 2.38x) but does not fully explain every listing: some institutions, including Capital One, remain below even their better-matched peer group, an open question rather than a settled one. A listing here means "below this institution's own peer group", not "below the 1.0x line that would mean realized losses exceeded the reserve" — those are shown as separate columns below for exactly this reason.

Running the reserve-adequacy screen…