Suneel ReddySuneel Reddy
FrameworkCredit & Risk

B2B Credit, Part II — Monitoring the Book: Aging, DSO, CEI and the Signals That Arrive Before the Default

August 9, 2026 · 6 min read
B2B Credit, Part II — Monitoring the Book: Aging, DSO, CEI and the Signals That Arrive Before the Default

Part II of a three-part series on B2B trade credit. [Part I](#/post/b2b-credit-part1-granting-terms) covered the grant decision. Part II covers monitoring. [Part III](#/post/b2b-credit-part3-collections) covers collections and the cash conversion cycle.

A receivables book is a loan portfolio that most companies never treat like one. It has vintages, roll rates, concentration risk and expected losses — and in most mid-market firms it is monitored by a single aging report that nobody reads past the first column.

Monitoring exists to answer three questions: is the money arriving on time, is it arriving from everyone, and which account changed? Different metrics answer each.

The aging report, read properly

The standard bucketing — current, 1–30, 31–60, 61–90, 90+ days past due — is the foundation, but the number that matters is not the balance in each bucket. It is the roll rate: what percentage of last month's 1–30 bucket moved into 31–60 this month?

Roll rates are the receivables equivalent of a credit portfolio's transition matrix. A stable book has a steep decay — most of what enters 1–30 is collected before 31–60. A deteriorating book shows the decay flattening months before the write-offs appear. Track roll rates by month and you have a leading indicator; track only balances and you have a photograph of the past.

Two refinements worth building:

  • Age from due date, not invoice date. Mixing Net 30 and Net 60 customers in the same invoice-date aging makes the report meaningless. Everything should be measured as days beyond terms (DBT).
  • Segment the aging. By customer tier, by sales region, by product line, by salesperson. Concentrated problems hide inside blended totals — one region running 40 days beyond terms disappears when averaged with three that pay early.

The core metrics

Days Sales Outstanding. DSO = (Accounts receivable ÷ Total credit sales) × Number of days. Simple, universal, and easy to misread — DSO moves with sales seasonality even when collection performance is unchanged. A revenue surge in the final month of a quarter mechanically inflates DSO; a slump deflates it. Use the countback method for a cleaner read on volatile books.

Best Possible DSO. BPDSO = (Current receivables ÷ Total credit sales) × Number of days. This is the DSO you would have if every customer paid exactly on terms. The gap between actual DSO and BPDSO is the part of your working capital that is genuinely being lost to late payment — as opposed to the part that is simply the terms you granted. Reporting DSO without BPDSO conflates a policy choice with a performance failure.

Collection Effectiveness Index. CEI = [(Beginning receivables + Credit sales − Ending total receivables) ÷ (Beginning receivables + Credit sales − Ending current receivables)] × 100. Where DSO measures speed, CEI measures quality — of what was available to collect, how much did we actually collect? A CEI above 80% is generally considered strong and 85%+ very strong. CEI is the better management metric precisely because it is far less distorted by sales volatility than DSO.

Average Days Delinquent. ADD = DSO − BPDSO. The single cleanest number for “how late is late, on average.”

Bad debt ratio and provisioning. Write-offs ÷ credit sales, tracked by vintage. And where accounting standards apply, an expected-credit-loss provision matrix — historical loss rate per aging bucket, adjusted for forward-looking factors — turns the aging report directly into the balance sheet number.

MetricMeasuresWatch for
DSOCollection speedDistortion from sales seasonality
BPDSOTerms-implied floorWidening gap vs DSO
CEICollection qualitySustained drop below 80%
ADDAverage latenessTrend, not level
Roll rateDeterioration momentumFlattening decay curve
ConcentrationSingle-name riskTop-5 buyers as % of AR

The behavioural signals

Metrics tell you the book is deteriorating. Signals tell you which customer is about to stop paying. In roughly descending order of predictive value:

  • Payment pattern change. A customer who paid on day 32 for two years and now pays on day 48 has told you something. Absolute lateness matters less than the change — this is the receivables equivalent of a bureau score drop.
  • Partial payments and unexplained short-pays. Paying ₹8 lakh against a ₹10 lakh invoice with no dispute raised is a liquidity signal, not an administrative one.
  • Dispute frequency rising. Disputes are often a delay tactic dressed as a quality complaint. A customer who never raised issues and now queries every invoice is buying time.
  • Order pattern shifts. Sudden order surges from a stretched account are dangerous — it can mean other suppliers have cut them off and you are the last line still open. Check whether your share of their wallet is growing for reasons you did not earn.
  • Instrument failures. Bounced cheques, failed NACH mandates, requests to re-present. In India, a cheque return is also a legal event worth preserving.
  • Communication change. Calls unreturned, the accounts contact replaced, requests routed to a new entity or address, or a request to bill a different group company — the last of these is frequently an attempt to strand liability in a shell.
  • External signals. GST filing lapses, MCA charge creation (they have pledged assets to someone else), credit-insurer cover withdrawal, adverse press, or your own bureau alert on the buyer's commercial file.

Run your own book

The metrics above are available as an interactive tool: the working capital & DSO calculator.

Enter your sales, receivables and payables and it computes DSO, Best Possible DSO, Average Days Delinquent, CEI and the full cash conversion cycle — then shows what hitting a target DSO frees up in cash.

The output worth studying is the split it puts on any DSO improvement. Some of the gap is late payment you can collect; the rest is terms you granted and can only change by renegotiating. Most collections targets quietly assume the first when they actually require the second — and the calculator refuses to let you conflate them.

Review cadence and portfolio hygiene

Limits and terms must be re-underwritten, not just granted:

  • Annual full review for all accounts above a materiality threshold — refreshed financials, bureau pull, limit recalculation using the four methods from Part I.
  • Event-driven review on any of the signals above, on a limit-increase request, or on any credit note above threshold.
  • Quarterly portfolio review at management level: concentration, aging shape, CEI trend, provision adequacy, and the list of accounts on hold.
  • Dormant limit cleanup. Every unused limit on an inactive account is free risk you are carrying for nothing. Sweep them annually.

The organisational point matters as much as the metrics: credit monitoring should sit with someone whose compensation does not depend on the sale. Where credit reports into sales, limits expand quietly and holds never happen. Where it reports into finance with a documented escalation path to a credit committee, the book behaves.

Continue to [Part III — Collections and the Cash Conversion Cycle](#/post/b2b-credit-part3-collections).

References

AccountingTools, Quadient, Billtrust, Allianz Trade — Collection Effectiveness Index definitions, formula and benchmark ranges. Standard AR practice on DSO, Best Possible DSO and Average Days Delinquent. Ind AS 109 / IFRS 9 — simplified provision-matrix approach for trade receivables.

#TradeCredit#DSO#CEI#AccountsReceivable#Monitoring#WorkingCapital
Written by Suneel Reddy — read more at suneelreddy.com →