Collections

Collections is the function responsible for securing repayment on loans — spanning everything from reminding a borrower before an instalment falls due, through following up on arrears, to negotiating workouts and pursuing recovery on accounts that have already been written off.

It is distinct from the metric used to measure it. Collection efficiency is the percentage of amounts due that were collected; collections is the operational function that produces that number.

Collections is often treated as a back-office cleanup function. In practice it is one of the highest-leverage operations in a lending business: in unsecured lending, credit losses are typically the largest single cost line after operating expense, and the difference between a competent and an incompetent collections operation shows up directly in the loss rate, the cost of funds, and whether the institution can price competitively.

Collections vs Recoveries

The two terms are frequently used interchangeably but describe different stages:

• Collections deals with accounts still on the balance sheet — delinquent but not yet written off. The objective is to bring the account back to current (a cure) and preserve the client relationship where possible.
• Recoveries deals with accounts already written off. The objective is purely cash recovery; the relationship is generally over, and expected recovery per account is low.

The operating models differ accordingly. Collections is relationship-preserving and cure-focused; recoveries is cost-driven, often outsourced, and measured on cash collected per unit of effort.

The Collections Lifecycle

Collections activity is organised around delinquency buckets, defined by days past due (DPD). Treatment escalates as accounts age, because both the probability of cure and the value of preserving the relationship decline with time.

• Pre-delinquency (Before due date) — Objective: Prevent the miss. Actions: Due-date reminders by SMS, app or call; confirming payment method is funded.
• Early stage (1–30 DPD) — Objective: Cure quickly, diagnose cause. Actions: Automated reminders, first calls, group meeting follow-up, promise-to-pay capture.
• Mid stage (31–90 DPD) — Objective: Prevent roll to non-performing. Actions: Intensified contact, field visit, guarantor or group contact, affordability discussion, workout options.
• Late stage (91–180 DPD) — Objective: Structured resolution. Actions: Formal demand, negotiated settlement, restructuring where viable, collateral discussion.
• Recovery / legal (180+ DPD) — Objective: Maximise cash recovery. Actions: Legal demand, collateral realisation, litigation where economic, external agency, portfolio sale.
• Post-write-off (After write-off) — Objective: Residual cash. Actions: Low-cost periodic contact, settlement offers, agency placement.

The economics change sharply across these stages. Cure probability is highest and cost per account lowest in the early bucket, and both deteriorate as accounts age — which is why the strongest collections operations are disproportionately weighted toward pre-delinquency and early-stage work rather than late-stage escalation.

Collections Strategy

Risk-based prioritisation

Not every delinquent account warrants the same effort. Effective operations segment the queue by expected value of contact — combining exposure size, probability of cure, cost of the contact channel, and the reason for the miss.

A borrower who has missed one instalment for the first time in nine cycles and a first-cycle borrower who has never paid are the same DPD bucket and entirely different problems.

Segmentation dimensions

• Delinquency history — first-time misser versus chronic
• Loan cycle number — first-cycle borrowers behave very differently from repeat borrowers
• Exposure size — justifies field visit cost or does not
• Reason for non-payment — cannot pay, will not pay, or a process failure such as a failed debit mandate
• Channel responsiveness — which contact methods have worked before
• Product and methodology — group versus individual

The three causes of non-payment

Separating these is the single most useful diagnostic in collections, because each requires a different response:

1. Cannot pay — genuine capacity loss. Response: workout, rescheduling, reduced instalment.
2. Will not pay — capacity exists, willingness does not. Response: escalation, consequence, enforcement.
3. Process failure — the borrower intended to pay and something broke: failed debit, wrong account, misapplied payment, agent unavailable, meeting missed. Response: fix the process. This category is consistently larger than institutions expect and is the cheapest to resolve.

Contact strategy

Channels vary enormously in cost and effectiveness: automated SMS and app notifications are near-free; voice calls cost staff time; field visits cost hours and transport. Strategy should match channel cost to expected recovery, escalating only where cheaper channels have failed and the exposure justifies it.

Key operational measures include right-party contact rate (how often contact reaches the actual borrower), promise-to-pay (PTP) rate, and PTP kept rate — the last being one of the better predictors of eventual cure.

Workout and Restructuring Options

Where the diagnosis is cannot pay rather than will not pay, enforcement destroys value. Common alternatives:

• Rescheduling — extending tenor to reduce the instalment.
• Payment holiday or moratorium — temporary suspension, usually with interest continuing to accrue.
• Instalment reduction with a balloon — lower payments now, larger settlement later.
• Partial settlement or discounted payoff — accepting less than the full balance to close the account.
• Refinancing — a new facility replacing the old. This carries the highest risk of misuse.
• Voluntary asset surrender, in secured or asset finance lending.

Two cautions:

Evergreening. Repeatedly refinancing a distressed borrower to prevent an arrear from appearing conceals non-performance, resets the aging clock, and typically leaves the borrower worse off. A new disbursement used to settle an existing balance is a restructuring, not a new loan, and should be classified and reported as such.

Staging and provisioning consequences. Restructuring for credit reasons is a significant increase in credit risk indicator under IFRS 9, and a distressed restructuring is generally a default indicator. Rescheduling does not reset a loan's stage, and treating it as though it does understates provisions.

Organisational Design

Loan officer collects own portfolio. Standard in group lending and field microfinance. The officer knows the borrower and the local context, and accountability is clear. The weakness is a conflict of interest: the same person who approved the loan is judged on its performance, which creates incentive to conceal arrears, advance funds personally, or refinance rather than report.

Specialised collections team. Separating collections from origination removes that conflict, allows specialist skills and dedicated tooling, and produces better data. It costs local knowledge and relationship continuity, and requires clean handover processes.

Hybrid. Loan officers handle early buckets where relationship matters most; a specialist unit takes accounts beyond a defined DPD threshold. This is the most common arrangement at scale.

Outsourced agencies. Typically used for late-stage and post-write-off accounts, on a contingency fee. The institution retains reputational and regulatory responsibility for agent conduct — outsourcing the activity does not outsource the accountability.

Incentives. Collections incentives are unusually prone to perverse effects. Targets set purely on amounts collected or on headline collection efficiency can drive pressure tactics, misallocation of payments across accounts, and staff funding instalments themselves to protect a target. Incentive design should reference cure quality and vintage performance, and should be paired with conduct monitoring.

Ethical and Regulatory Requirements

Collections is the part of lending with the greatest potential for direct harm to clients, and it is where most consumer protection regulation and most reputational damage in the sector originates.

Practices that are prohibited or restricted in most jurisdictions and under recognised client protection standards include:

• Contact outside permitted hours, or at unreasonable frequency
• Threats, intimidation, abusive language, or threats of legal action not actually intended
• Disclosure of the debt to employers, neighbours, community groups or contacts, or any form of public shaming
• Accessing a borrower's phone contacts or media without lawful, informed consent — a documented pattern of abuse in digital lending
• Seizure of assets without legal process
• Misrepresenting the borrower's legal position or the consequences of non-payment
• Collecting from vulnerable individuals without appropriate safeguards

Sound practice requires: a written collections policy and staff code of conduct; documented, auditable contact records; disclosure to the borrower of the amount owed and how it was calculated; a complaints and redress channel the borrower can actually reach; data protection compliance; and applying the same standards contractually to any outsourced agent.

Group lending adds a specific dimension: peer pressure is part of the methodology's design, but it can escalate into coercion, asset seizure by fellow members, or public humiliation. The line between social accountability and harassment needs to be drawn explicitly in policy rather than left to field discretion.

Requirements vary by jurisdiction; institutions should confirm applicable rules with their regulator.

Technology in Collections

• Automated queue allocation by risk segment, rather than alphabetical or branch-based distribution
• Workflow and case management with full contact history, PTP tracking, and next-action scheduling
• Automated pre-due and early-stage messaging, the cheapest intervention available
• Digital repayment channels — mobile money, standing instructions, payment links — removing cash handling and travel from the process
• Collections scoring — models predicting cure probability and optimal treatment, allowing effort to be concentrated where it changes the outcome
• Field collection apps with real-time receipting and geolocation, closing the gap between cash received and cash posted
• Champion-challenger testing — running an alternative treatment on a matched subset to measure whether it actually outperforms current practice

Key Collections Metrics

• Collection efficiency — Share of amounts due collected in the period
• Roll rate — Proportion of accounts moving from one delinquency bucket to the next
• Cure / resolution rate — Share of delinquent accounts returning to current
• Right-party contact rate — Share of contact attempts reaching the actual borrower
• PTP kept rate — Share of promises to pay that were honoured
• Cost to collect — Collections cost as a percentage of amounts recovered
• Recovery rate — Cash recovered as a percentage of amounts written off
• First-payment default rate — Share of loans missing their first instalment — an underwriting signal surfacing in collections