Forensic red flags in company accounts, with examples
What is a forensic red flag?
A measurable sign that a company's reported profit or strength may not be what it seems. It is a reason to read further, never proof of wrongdoing.
Accounts involve judgement. A company decides when a sale counts as revenue, how fast its machines wear out, and whether a customer who has not paid will pay. Most use that judgement fairly. Forensic checks look for the places where the reported picture and the underlying cash, debt or ownership have drifted apart.
The fifteen below are the red checks Valarth runs on every listed company. They fall into four groups. One flag is common and often harmless. Several at once are worth your time.
Is the profit arriving as cash?
Profit is an opinion; cash is a fact. Four checks compare the two.
1. Profit with negative operating cash flow net profit above zero, cash from operations below zero
The company reports a profit, yet its day-to-day business used up cash. This happens honestly when a business grows fast and funds stock and credit to customers. Year after year, it means the profit exists only on paper.
2. Accruals in the worst fifth accruals quintile 5
Accruals are the part of profit not backed by cash. Ranking all companies by accruals and taking the worst fifth picks out those whose profit leans most on accounting entries. Accruals quintile, explained.
3. Beneish manipulation flag M-score above −1.78
The M-score combines eight ratios, among them receivables growing faster than sales, margins falling, and depreciation slowing. Above −1.78 the accounts resemble those of companies later found to have manipulated earnings. Fast, honest growth can trigger it too, so Valarth sets sales growth aside before raising the flag. Beneish M-score, explained.
4. Montier C-score of 4 or more out of 6
Six yes-or-no signs that the books are being flattered: profit rising while cash flow falls, receivable days rising, inventory days rising, other current assets growing, depreciation falling against fixed assets, and fast asset growth. Four or more together is the warning. Montier C-score, explained.
Can the business carry its debt?
Six checks on borrowing and on the risk of failure. The leverage checks are not read for banks, lenders and insurers, whose borrowings are their raw material.
5. Debt-to-equity above 2 borrowings more than twice shareholders' funds
₹200 of debt for every ₹100 the owners have put in. Normal for a power utility, a warning for most other businesses. Debt-to-equity, explained.
6. Interest coverage below 1.5 operating profit less than 1.5 times interest
The company earns barely enough to pay its lenders. One poor quarter and it cannot. Interest coverage, explained.
7. Net debt above 4 times EBITDA
At today's operating profit it would take more than four years to clear the debt, using every rupee. Net debt / EBITDA, explained.
8. Altman Z-score in the distress zone below 1.8
A score built from five balance-sheet and profit ratios, designed to separate companies that went bankrupt from those that did not. Below 1.8 is the distress zone; above 3 is the safe zone. It was built on manufacturers and misreads asset-light and regulated businesses. Altman Z-score, explained.
9. Ohlson O-score above a 50% modelled chance of distress
A second bankruptcy model using size, debt, profit and cash flow. When Altman and Ohlson agree, take it seriously. Ohlson O-score, explained.
10. Goodwill above half of equity
Goodwill is the premium paid for past acquisitions. If those purchases disappoint, it is written off and equity shrinks with it. A balance sheet that is half goodwill is half hope. Goodwill / equity, explained.
What are the owners doing with their shares?
Promoters know the business best. Three checks watch what they do, not what they say.
11. Promoter pledge above 25% of the promoters' own holding
Promoters have borrowed against their shares. If the price falls, lenders ask for more shares or sell the ones they hold, which pushes the price down further. Promoter pledge, explained.
12. Promoter stake falling
The controlling group has been reducing its holding quarter after quarter. There are good reasons, such as meeting the minimum public shareholding rule, and bad ones. Promoter trend, explained.
13. Share count up more than 10% in a year
New shares were issued, so each existing share owns a smaller slice. Useful when the money builds something; a warning when the company issues shares every year to stay afloat. Share count change, explained.
Does the record hold up?
14. Losses in three or more of the last ten years
A business that loses money every third year has not shown it can earn through a cycle. Loss-making years, explained.
15. Piotroski F-score of 3 or less out of 9
Nine yes-or-no tests of whether profitability, debt and efficiency improved over the year. Three or fewer means most are moving the wrong way. Piotroski F-score, explained.
What have red flags looked like in Indian markets?
Two cases from the public record. In both, the signs were in filings before the price fell.
Satyam Computer Services, 2009: cash that was not there
In January 2009 the chairman of Satyam wrote to the board and the exchanges that the cash and bank balances on the balance sheet were overstated by more than ₹5,000 crore. For years the company had reported profits that never showed up as money it could spend. In the weeks before, lenders had sold shares the promoters had pledged, and the promoter stake had fallen to a few per cent. Flags 1, 11 and 12 on this page describe that pattern.
Manpasand Beverages, 2018: the auditor walks away
In May 2018 the company's auditor resigned days before the annual results were due, saying it had not been given information it had asked for. The share price fell sharply over the following sessions. An auditor resigning mid-term is not a ratio and no formula catches it, which is why the notes and announcements matter as much as the numbers.
These are historical cases, described from what the companies and their auditors disclosed at the time. They are examples of a pattern, not a comment on any company today.
How common are red flags today?
Of the 5496 listed companies with enough data to judge, as of 9 Nov 2026. Valarth turns the count of red flags into a risk level.
Companies by risk level
What is the audit-priority tier?
A second reading of the accounts, by Valarth's forensic risk engine: a score from 0 to 100 and a Green, Amber or Red tier, printed on every company page beside the red flags.
Where the fifteen flags above are fixed thresholds, the engine reads thirty-one checks in five families. A ratio is set against two reference classes: every scored company in the same sector and year, and the company's own earlier years. The larger anomaly decides what a check reads: in line, leaning adverse or out of line. A published model (Beneish, Dechow, Montier, Altman, Ohlson, Merton) reads through its own cutoff instead, and an exchange filing is read as an event.
Accrual and manipulation 12 checks
Profit that is not backed by cash: accruals against assets, receivables and inventory growing faster than sales, fixed assets growing faster than sales, a jump in gross margin, other income carrying the profit, the gap between the tax charge and the tax rate, taxes paid in cash against the tax charge, capital spending against depreciation, and the Beneish, Dechow and Montier models.
Cash integrity 5 checks
Whether the cash is real and arrives: what the cash balance earns against a bank deposit, cash piling up while debt rises, three years of operating cash against three years of EBITDA, profit with no operating cash behind it, and dividends paid with no free cash flow to pay them from.
Promoter pledge and stake 2 checks
What the owners are doing with their shares today: the share of the promoters' own holding that is pledged, and a stake falling quarter after quarter. Read on the latest year from the shareholding pattern.
Governance events 7 checks
What the exchange filings say: the statutory auditor resigning, the audit opinion, key managers and directors leaving, results and meetings slipping past their deadlines, regulators and reported defaults, suspensions and exchange queries, and credit rating actions. An event's weight halves every two years; a check is judged only where a filing stream has carried the company.
Distress 5 checks
Whether the business can carry itself: the Altman Z″ and Ohlson models, Merton's distance to default from the share price, interest coverage, and the share of the debt due within a year.
A family's score is the mean of its three most adverse checks, on 0 to 100, so one loud figure does not carry it and many quiet ones do not dilute it; the governance family combines its checks so two concerns say more than one. The company's score is its worst family, plus 10 for each other family at 50 or above, and times 1.25 when governance is at 60 or above. Green is below 70, Amber is 70 to 85 and Red is above 85, set so that Red is about the top 5% of the market. Confidence is the share of checks that had the data to be judged; a Red on thin data is shown as such.
A bank, lender or insurer has no tier until the lender engine is built, and a company whose accounts are not comparable across a demerger or divestment is read on the years that are. The tier, the score and the most severe flag are on every company page here. Every check with its figures, the sector percentiles, the five-year drift and what would change the tier are in the app's forensic report. What the checks look for, group by group, is in the forensic checks primer.
An audit-priority score from the public filings, not a finding of wrongdoing.
Companies by tier
| Tier | Score |
|---|---|
| Green | Below 70 |
| Amber | 70 to 85 |
| Red | Above 85 |
What do the green flags look for?
The opposite signs. Valarth runs 15 of these too, and reports greens minus reds beside the risk level.
- Cash and accounts: operating cash flow at least 90% of net profit; accruals in the best fifth; a Beneish M-score below −2.22.
- Strength: Altman Z above 3; debt-to-equity below 0.1; interest coverage above 10; a Piotroski F-score of 7 or more; return on capital employed above 20%.
- Owners: promoters own at least half and have pledged nothing; the promoter stake is rising; mutual funds or foreign funds are adding.
- Record: profitable in each of the last ten years; dividends paid ten years running; profit up five years in a row; shares bought back.
What can red flags not tell you?
- They are not proof. A flag says the numbers have a shape that has gone with trouble before. Many flagged companies are sound.
- They lag. Every check is computed from filings, which arrive weeks after the quarter ends.
- They miss what is not a number. An auditor's resignation, a qualified audit opinion, a related-party loan in the notes: read the annual report for these.
- Some do not fit some businesses. Leverage checks are not read for lenders. The Altman score misjudges asset-light and regulated companies. A young company can fail several checks while doing exactly what it should.
- Clean is not safe. No red flags means none of these fifteen fired. It does not mean the share is worth its price.