The signals that save you basis points before a single payment is missed.
Most credit teams watch the wrong clock. Delinquency, roll rates, vintage curves — that's the rear-view mirror. Essential, but it tells you about losses that have already started. By the time a customer misses a payment, your options have narrowed: the cure is harder, the loss is closer, and you're reacting instead of steering.
The teams that protect a book best do something different. They watch the signals that fire before the missed payment — and, crucially, they act on them. That's what an early warning indicator (EWI) framework is for. Here are the signals worth watching, and how to turn them into something other than a dashboard nobody reads.
1. Score migration — watch the slope, not the level
A customer sliding down two behavioral or bureau score bands over a few months is telling you something long before they ever miss you. A single score is a snapshot; the trajectory is the warning.
2. Bureau triggers — what they're doing elsewhere
A missed payment at another lender, a flurry of new inquiries, loan stacking, rising external balances. People usually default on someone else first — a bureau refresh often sees the crack before your own book does.
3. Payment-behavior shifts
Full-pay turning into minimum-pay. Partial payments. Autopay quietly switched off. Or paying later and later within the grace window — day 28 instead of day 3. The amount and timing of a payment is a confession; read it.
4. Utilization & exposure creep
Revolving utilization climbing toward the limit, sudden full drawdowns, repeated limit-increase requests. Reaching harder for credit is often a liquidity signal, not a growth one.
5. Cash-flow signals (if you have transaction data)
Salary landing late or shrinking, balance hitting zero before payday, rising overdrafts, a jump in gambling or BNPL spend. For digital lenders this is the highest-signal, earliest layer you have — use it.
6. Engagement & contactability
App logins dropping off, bounced emails, a disconnected number, sudden contact-detail changes. Disengagement quietly precedes distress more often than people expect.
(And a thin overlay on top: segment and macro stress — layoffs in a sector, regional shocks — to tilt your thresholds when the environment turns.)
From signals to action — the part most teams skip
A signal you don't act on is just trivia. An EWI framework needs three things beyond the indicators themselves:
A watchlist: combine signals into one tiered flag — not 20 separate alerts, one prioritized list.
Mandated actions per tier: proactive outreach, a limit reduction, re-underwrite, a pre-emptive hardship offer, tighter monitoring. The action is defined, not improvised in the moment.
A feedback loop: measure whether flagged accounts that got an intervention actually defaulted less than a control group. If you can't show the lift, you're guessing.
The mistakes I see most
Watching only lagging indicators — you're reading the autopsy, not preventing it.
Signals with no owner and no required action — the watchlist nobody works.
Too many alerts, no prioritization — alert fatigue, so everything gets ignored.
Never testing whether the interventions pay for themselves.
If you only do one thing
Pick your three highest-signal, earliest indicators — for a digital lender that's usually bureau-trigger inflows, a payment-pattern shift, and a cash-flow signal — and stand up a single weekly watchlist with one mandated action per tier. One list that gets worked beats ten dashboards that get admired.
Next week: BNPL economics — where the loss really comes from, and the lever most teams miss.
If this was useful, forward it to someone watching a book.
Views are my own and don't represent my employer.