09.7 · The forensic checklist¶
Why this matters: the six preceding lessons contain more tests than anyone can hold in their head. A checklist turns them into a procedure: forty items, each with where to look, what threshold, and how bad. Run it on every company before the valuation, and run it again every year you hold the stock. Applied to Kaveri Pumps it produces the module's verdict — and a template you will reuse in the capstone and the private casebook.
Learning objectives — after this lesson you can:
- Run a 40-point forensic checklist organised by statement and by governance, with sources and thresholds.
- Score it consistently (0–3 per item), aggregate to a RAG rating, and translate the rating into analytical actions (adjust, discount, monitor, avoid).
- Complete the checklist for Kaveri FY26 and defend the result.
Prerequisites: 09.1–09.6 · Time: ~90 min (plus ~3 hours to run it on a real company the first time)
1. Scoring¶
| Score | Meaning |
|---|---|
| 0 | Clean — the test passes clearly |
| 1 | Watch — mildly adverse, explained, or one-year |
| 2 | Concern — adverse trend or unexplained; requires adjustment of the numbers or a valuation discount |
| 3 | Disqualifying — evidence of manipulation, fraud, or governance that removes minorities' claim; do not invest until resolved |
Aggregate: sum of scores (max 120), count of 2s and 3s, and an overall RAG: Green (total ≤ 12, no 3s, ≤ 2 twos), Amber (13–30, no 3s), Red (any 3, or > 30). The aggregate matters less than the pattern: five 2s in the revenue section mean something different from five 2s scattered across the sheet.
2. The checklist¶
A. Revenue and receivables (10)¶
| # | Test | Where | Threshold for 2 | Kaveri FY26 |
|---|---|---|---|---|
| A1 | Receivables growth vs revenue growth (3 years) | BS, P&L | > 1.5x for 2+ years | 2 (2.1–2.4x for three years) |
| A2 | DSO trend | Compute | +10 days/yr, or > peers by 30 days | 2 (61 → 96) |
| A3 | Ageing: > 6-month bucket share and growth | Schedule III ageing | > 15% of receivables or doubling | 2 (18%, doubled) |
| A4 | Allowance coverage of overdues | Receivables note | < 15% of > 6-month bucket, or falling | 2 (6.4%, falling) |
| A5 | Unbilled revenue / contract assets vs revenue | Ind AS 115 note | Growing faster than revenue | 1 (not disclosed in the running example — the absence is the point) |
| A6 | Revenue policy changes / estimates | Accounting policies (diff YoY) | Any profit-raising change | 0 |
| A7 | Related-party revenue | RPT note | > 5% of revenue | 0 |
| A8 | Q4 concentration / Q1 drop | Quarterly | Q4 > 35% without seasonality | 0 (Q4 31% — seasonal) |
| A9 | Gross vs net presentation | Revenue note | Agent booked as principal | 0 |
| A10 | Customer concentration and new large customers | Ind AS 115 note | Top customer > 25%; new large customer with long terms | 1 (two state agencies dominate solar overdues) |
B. Costs, assets, provisions (10)¶
| # | Test | Where | Threshold for 2 | Kaveri FY26 |
|---|---|---|---|---|
| B1 | Capex/D&A vs capacity growth | PP&E note; capacity disclosure | > 2x with no capacity | 0 |
| B2 | CWIP ageing; suspended projects | CWIP ageing | > 2 years; > 25% of net block | 0 |
| B3 | Intangibles under development / capitalised development | Intangibles note | Growing faster than revenue | 0 |
| B4 | Capital advances and other non-current assets | Notes | Unexplained > 5% of assets | 0 |
| B5 | Capitalised borrowing costs | Note; implied interest rate | Implied rate < borrowing rate by 2 pp | 0 |
| B6 | Depreciation rate / useful-life changes | PP&E note; Ind AS 8 | Rate falling; life extended | 0 |
| B7 | Provision roll-forward: reversals | Provisions note | Reversals > 20% of opening | 0 |
| B8 | Exceptional items frequency | P&L, 5 years | > 2 of 5 years, same type | 0 (2 of 6, different types) |
| B9 | Inventory days and finished-goods mix | Inventory note | +10 days/yr; FG growing fastest | 0 (steady) |
| B10 | Bank stock statements vs books | Schedule III disclosure | Any material discrepancy | 0 |
C. Cash flow (8)¶
| # | Test | Where | Threshold for 2 | Kaveri FY26 |
|---|---|---|---|---|
| C1 | CFO/EBITDA, 3-year cumulative | CFS | < 50% (non-lender) | 2 (42%) |
| C2 | CFO/PAT, 3-year cumulative | CFS | < 75% | 1 (78% over three years; FY26 alone 72%) |
| C3 | Implied yield on cash | Other income ÷ avg cash | < 3% when rates 6–7% | 0 (7.4%) |
| C4 | Cash-rich but borrowing | BS | Cash > 20% of assets with debt > 1x EBITDA and low yield | 0 |
| C5 | Supplier finance / factoring / bills discounted | Payables note; contingents | Any undisclosed programme; payable days jump | 0 |
| C6 | Loans, ICDs, guarantees to related/group entities | Loans note; RPT; CARO iii–iv | Any material | 0 |
| C7 | Investments in unlisted/obscure entities | Investments note | Growing; unexplained | 0 |
| C8 | Standalone vs consolidated cash and flows | Both statements | Cash trapped; parent funding losses | n/a |
D. Governance and people (12)¶
| # | Test | Where | Threshold for 2 | Kaveri FY26 |
|---|---|---|---|---|
| D1 | Promoter pledge | SAST Reg 31 | > 10% of holding, or rising | 1 (6%, new) |
| D2 | Auditor: size, resignation, opinion, KAMs | Audit report; announcements | Mid-term resignation (3); qualified (2–3) | 1 (KAM on receivables; rotation only) |
| D3 | KMP changes | Announcements | Two in 24 months | 0 |
| D4 | Related-party purchases / expenses | RPT note | > 5% of costs or rising | 2 (10.4% of materials, rising 5 years) |
| D5 | Royalty / brand fees | RPT note | > 3% of sales or rising | 0 |
| D6 | Preferential issues / warrants to promoters | Announcements | Any before news; repeated | 0 |
| D7 | Fund-raising vs reported cash | Announcements; BS | Raising while cash-rich | 0 |
| D8 | Managerial remuneration | Board's report | > 8% of PAT; up in loss year | n/a (not in running example) |
| D9 | Board independence in substance; independent-director exits | Report; announcements | Vague-reason exits | n/a |
| D10 | SEBI / exchange / NFRA / RBI actions; ASM/GSM history | Regulators' sites | Any substantive order | 0 (fictional) |
| D11 | Disclosure quality: KPIs dropped, definitions changed | Presentations over 8 quarters | Any KPI withdrawn when adverse | 1 (receivable days withdrawn) |
| D12 | Stated vs actual promoter holding; connected entities | Shareholding pattern; MCA21 | "Public" holders at company address | 0 |
E. Scores and cross-checks (bonus items; do not add to the total)¶
| Model | Kaveri FY26 | Flag |
|---|---|---|
| Beneish M | −2.14 | No (trend adverse) |
| Altman Z | 6.99 | No |
| Piotroski F | 4 | Deteriorating |
| Accruals | 2.3% | Elevated |
| C-score (approx.) | 2–3 | Watch |
3. Kaveri — the result¶
| Section | Score | 2s and 3s |
|---|---|---|
| A. Revenue & receivables | 10 | four 2s (A1–A4) |
| B. Costs, assets, provisions | 0 | — |
| C. Cash flow | 3 | one 2 (C1) |
| D. Governance | 5 | one 2 (D4) |
| Total | 18 | six 2s, no 3s → Amber |
Verdict. No evidence of fabrication or fraud: the cash is real, the assets are real, the costs are honest, the auditor is engaged. The problems cluster in one place — a solar receivables book from state agencies that is growing, ageing and under-provided — plus a rising related-party purchase line and a small, new promoter pledge. Together they are yellow flags that need resolution, not red flags that end the analysis:
- Adjust: provide 25–30% on the > 6-month bucket in your own numbers (EPS ₹15.1 → ~₹13.6); model NWC at 26–28% of revenue in the bear case; treat the Kaveri Castings purchases as a ₹5–10 Cr/yr margin question until the supplier's margins are disclosed.
- Discount: governance/forensic discount via scenario weights (bear 35%) or WACC (+50–75 bps), not both.
- Monitor (quarterly): receivable days and ageing; provisioning; state payment news; the pledge; RPT purchase share; whether the withdrawn KPI returns.
- Escalate to Red if: the auditor's KAM turns into a qualification; a state agency disputes a claim; the pledge exceeds 15%; a CFO leaves; or contract assets appear and grow.
That verdict, with the checklist attached, is the forensic section of the memo in
11.3 and the forensic-checklist.md in the casebook template.
4. Running it on a real company — the workflow¶
- Pull the last three annual reports and the last eight quarterly results/presentations; the shareholding pattern and SAST/PIT filings; the auditor's reports; exchange announcements for three years; SEBI order search.
- Compute the ratio items (A1–A4, B1, B5, B6, B9, C1–C4) with
tools/fi— an afternoon the first time, an hour thereafter. - Read the notes for the disclosure items (A5–A10, B2–B4, B7, B10, C5–C8, D4–D7) — the longest part; keep a page reference for each score.
- Search for the regulatory and people items (D1–D3, D8–D12).
- Score, aggregate, write the verdict in the four-part form above (adjust / discount / monitor / escalate).
- Date it and file it. The next year's run compares against it; the trend of the score is more informative than its level.
Trader's lens
A checklist is a pre-trade risk check. Nobody on a desk puts on size without the risk system clearing the trade, however good the idea looks; the forensic checklist is that system for a fundamental position. The "Amber" rating is the equivalent of a reduced limit — you can trade it, smaller, with tighter monitoring.
India notes
The Indian disclosures that make this checklist workable from outside: Schedule III ageing schedules (receivables, CWIP, intangibles), the bank stock-statement reconciliation, Ind AS 115 contract balances, Ind AS 24 RPT notes, CARO 2020, LODR Reg 30 event disclosures, SAST Reg 31 pledges, PIT Reg 7 insider trades, half-yearly RPT filings, and SEBI's public order database. Most of them date from 2015–22 — the Indian forensic toolkit is much stronger than it was when Satyam happened.
Common mistakes
- Treating the total score as the answer instead of the pattern of 2s and 3s.
- Scoring without page references (unrepeatable next year).
- Running the ratio items and skipping the reading items — the reading is where the 3s are.
- Applying manufacturer thresholds to lenders, platforms or project companies without adjustment.
- Letting a Green rating end scepticism; letting an Amber rating end the analysis.
Key terms¶
| Term | Meaning |
|---|---|
| Forensic checklist | A structured set of tests with sources, thresholds and scores |
| RAG rating | Red / Amber / Green summary of the checklist |
| Disqualifying (3) | A finding that removes the investment case until resolved |
| Adjust / discount / monitor / escalate | The four responses to forensic findings |
| Page reference | The citation to the filing page supporting each score |
Check your understanding¶
- Kaveri scores four 2s in section A and none in B. What does the pattern tell you that the total does not?
Answer
The issue is concentrated in one mechanism — customer credit on solar tenders and its provisioning — rather than spread across the accounts. That points to a specific business risk (state payments) with a specific fix (provisioning, monitoring), not to pervasive manipulation.- Which single new fact would move Kaveri to Red, and why?
Answer
A qualified audit opinion on receivable recoverability (D2 → 3), or a state agency formally disputing a claim (A3/A4 → 3): either converts a provisioning judgement into an acknowledged loss and questions the revenue recognised on it.- Adapt A1–A4 for an EPC contractor.
Answer
Use unbilled revenue + retention + receivables as the base; compare growth with revenue; age unbilled by contract; test coverage of disputed/arbitration receivables; add a test for cost-to-complete estimate changes (margin revisions on contracts).- Why is "no 3s" a necessary but not sufficient condition for Green?
Answer
Several 2s in one section (Kaveri's A1–A4) indicate a real, if not disqualifying, problem that changes the numbers and the risk; Green should mean "use the reported numbers as-is with normal monitoring", which is not the case here.- What goes in the casebook after the checklist is run?
Answer
The scored sheet with page references, the RAG, the four-part verdict (adjust/discount/monitor/escalate) with specific triggers and dates, and the score's history — so that next year's run is a comparison, not a fresh start ([15.2 templates](../15-capstone/02-templates.md)).Go deeper¶
- Schilit, Financial Shenanigans, Part Five — the "Fraud Detection Checklist".
- Charles Mulford & Eugene Comiskey, Creative Cash Flow Reporting — cash-flow section items.
- The Kaveri running-example pages — rerun every item with the tables there and confirm each score.
- Module 15 capstone — the checklist as part of the full case-study method.
← Previous: 09.6 Forensic scoring models · Module index · Next: 10.1 Model architecture →