The Prevention Curve
The economics of acting before the roll
The first paper in this series, “The Averages Are Lying,” makes three points that set up this one: consumer stress is concentrated beneath calm averages, the behaviour that predicts default appears one to two years before the missed payment, and credit scores say little about how a customer responds to treatment. This post covers what to do about it.
What is delinquent debt worth once it rolls?
Very little. The Federal Trade Commission’s study of the debt-buying industry found charged-off debt sells for about four cents on the dollar: 7.9 cents when fresh, 3.1 cents at three to six years old and 2.2 cents beyond that.¹ The older a balance gets, the less of it is left to recover.
Resale price is not recovery value, but it is the market’s verdict on what an aged claim is worth.
Why pre-delinquency intervention pays
No public study runs prevention against collection head-to-head, but the components all point the same way. McKinsey’s work on digital-first collections found pre-delinquency contact strategies cut charge-off losses 8% at a North American lender. Digital-first programs delivered 20–25% reductions in non-performing loans, and customers reached digitally made 12% more payments than those contacted through traditional channels.²
reduction in charge-off losses from pre-delinquency contact strategies at a North American lender.
Source: McKinsey & Company, 2021
“The cheapest dollar to protect is the one that never rolls.”
A three-step plan for moving upstream
The response we recommend is a sequence, not a project.
1. Instrument the early-warning layer
Utilization creep, minimum-payment persistence, engagement decay and payment-hierarchy shifts exist in every issuer’s data today. Build the layer that reads them continuously, because you cannot intervene on what you cannot see. The next stress leg will show up in these signals before it shows up in the loss line.
2. Re-segment on behaviour
Replace score-band treatment waterfalls with segmentation built on behavioural response: how a customer engages, what resolves them, what backfires. Measure the program on cure rates and roll rates, never on contact volumes. This is the approach SymendCure is built around.
3. Move upstream
Shift budget and capability from post-delinquency collection toward pre-delinquency intervention. For most delinquent customers, the evidence suggests, the problem is cash flow. They do not need pressure; they need a path: the right structure, offered with dignity, before the first missed payment. The BNPL repayment evidence shows the same consumer will perform when the structure fits.
Prevention doesn’t stop at cure, either. For card issuers and lenders, the moment a customer cures is a chance to protect them from rolling again, which is where SymendPrevent’s bill payment protection comes in. Our post on why cured customers still churn makes the case.
“See earlier, treat smarter, act sooner. Each step compounds the last.”
One caution: AI won’t fix the wrong playbook
The instinct across the industry right now is to hand this entire problem to artificial intelligence, on the assumption that a sufficiently powerful model pointed at collections will fix it. It will not, for a reason the first paper established: much of collections practice today is built on the wrong assumptions, including lagging indicators, score-based segmentation and willingness framing.
An engine is only as good as the map it is given. AI applied to an incomplete playbook automates the incompleteness. What separates the deployments that move portfolios from the many that stall is the philosophy that steers the engine. That is the subject of the second paper in this series, and of the next three posts.
The bottom line
Value drains out of a balance at every stage it rolls, so the economics favour acting early. The full paper, “The Averages Are Lying,” carries the complete argument and sources. Its companion, “AI Is an Engine,” picks up where this one ends.
A 7-part series
The New Delinquency Playbook
Seven short posts drawn from two Symend research papers. The first, “The Averages Are Lying,” argues that consumer stress has become concentrated rather than widespread, that delinquency is a lagging indicator, and that credit scores have decoupled from behaviour, so issuers need to read behaviour early, re-segment on it, and move intervention upstream. The second, “AI Is an Engine,” argues that collections AI amplifies whatever operating philosophy it is given, so it only works when behavioural science steers it, outcome objectives replace deflection metrics, and credit decisions stay on the deterministic side of the line.
From “The Averages Are Lying”
- Part 1 of 7 Delinquency Is a Lagging Indicator The behaviour that predicts default shows up one to two years before the missed payment
- Part 2 of 7 The K-Shaped Consumer Why every aggregate dashboard is wrong about consumer stress
- Part 3 of 7 Same Score, Different Customer Credit scores have decoupled from behaviour — and treatment built on them shows it
From “AI Is an Engine”
- Part 5 of 7 The 95% Problem Collections is adopting AI faster than it is getting results
- Part 6 of 7 The Stigma Exception Most consumers dislike service AI. Debt may be different — if the engine is steered right
- Part 7 of 7 The Deterministic Line What AI must never decide in collections — and where it earns its keep
Frequently asked questions
What is pre-delinquency intervention?
Engaging customers who are still current but showing signs of financial strain, before the first missed payment, with a reminder, a better payment structure or another form of support.
Why is early intervention cheaper than late-stage collection?
Because value decays as debt ages. Charged-off debt sells for about four cents on the dollar on average, and the older it gets, the less it is worth.
How much is charged-off debt worth?
The FTC found it sells for about 4 cents on the dollar: 7.9 cents when fresh, 3.1 cents at three to six years old and 2.2 cents beyond that. The data is from 2006–2009, but the decay-by-age pattern is the durable finding.
What signals should trigger pre-delinquency outreach?
Utilization creep, minimum-payment persistence, declining engagement with statements and outreach, and shifts in which bills a customer pays first.
How should a pre-delinquency program be measured?
On outcomes such as cure rates, roll rates and charge-off losses, not on contact volumes or how many messages were sent.
Endnotes
- Federal Trade Commission, The Structure and Practices of the Debt Buying Industry, 2013 (2006–2009 portfolio data). Prices have likely shifted as the resale market matured; the decay-by-age pattern is the durable finding.
- McKinsey & Company, “Holistic Customer Assistance Through Digital-First Collections,” 2021. Consulting case data; pre-generative-AI.
