Financial Shenanigans

Figures converted from INR at historical FX rates — see data/company.json.fx_rates. Ratios, margins, and multiples are unitless and unchanged.

The Forensic Verdict

The accounting at Indian Oil reads cleanly on the income statement and the cash-flow statement, but the FY2025 audit report contains two live CARO qualifications, an open FCPA-adjacent matter the Directors' Report does not name, and a disclosed fraud bundle of $15.3M concentrated in one geography. Earnings are cash-backed (3-year CFO/NI = 1.97×, accrual ratio negative), revenue is not stretched (debtor days 6-14 across 12 years, no contract-asset build), and the C&AG of India has issued a NIL comment for 19 consecutive years on the standalone financials. The risk is not P&L manipulation. The risk is a structural $7.17B short-term-funds-funding-long-term-assets mismatch that the auditors flagged this year, repeated SEBI LODR fines for unfilled independent-director seats, and the fact that the ADT-4 process under Section 143(12) was still "in process" at sign-off for the bottling-plant frauds. The single data point that would most change the grade: a separate, named disclosure of the Albemarle internal review's findings (or its closure with no exposure) in the next Directors' Report.

Shenanigans Scorecard

The three red flags cluster at the bottom of the table — they are governance and disclosure failures, not earnings manipulations. The earnings statements themselves pass the standard accrual, working-capital, and revenue-recognition tests. This is the right shape for a Maharatna PSU operating under administered pricing: low incentive to overstate earnings (no equity-linked pay), but elevated risk of policy-driven balance-sheet distortion and locality-level operational fraud.

Breeding Ground

The conditions that make accounting shenanigans more likely are partially present, but they push toward governance failure rather than earnings management.

The breeding ground tells us where to look. Compensation does not pull management toward accounting aggression. CAG supplementary audit (separate from statutory audit) has issued NIL comment 19 years running — a hard, external, non-negotiable check that most listed Indian companies do not face. What is unhealthy is the structural dependence on a single shareholder for board-composition decisions: when MoPNG is slow to nominate independent directors, the audit committee's check on management thins. The ~$6,280 per-exchange-per-quarter fine pattern is small in dollar terms but signals five consecutive quarters of unattended fiduciary architecture.

Earnings Quality

Earnings are reported approximately as they are earned. The standard tests for revenue stretching, soft-asset buildup, and reserve gymnastics all come back clean.

Revenue is not running ahead of receivables

Debtor days have lived between 6 and 14 days for nine years. Even in the COVID year (FY21), the rise to 14 was modest. There is no quarter-end revenue-stuffing pattern, no pull-forward into Q4, no contract-asset line that could hide unbilled revenue. Inventory days ran 46-111 across the cycle, which tracks crude-price volatility, not earnings management.

Operating margins swing with the cycle, not with the storyline

Other income is mostly dividends, interest, and fair-value gains on investments. It rises in absolute dollars over time because the group's investment book grew from $2.65B (FY14) to $7.86B (FY25). The proportion of operating income coming from Other Income is the test that matters: 7.1% in FY24 (a strong year) versus 19.7% in FY25 (a weak year). That is a yellow flag — not because management is fabricating the income, but because the headline operating result becomes more dependent on non-core sources when cracks compress. Strip Other Income from FY25 and pretax goes from $1,996M to $1,164M.

Capex is running 2× depreciation, but with a visible pipeline

Capex has run at 1.7-2.1× depreciation for nine straight years. That gap could be a sign of capitalised maintenance, but in IOC's case it is consistent with disclosed major projects: Panipat refinery expansion (~$4.10B, commissioning Jun-2026), Gujarat petrochem expansion (~$2.09B, commissioning Jun-2026), green hydrogen, and the LNG/CBG networks. CWIP grew from $2.59B (Mar-17) to $9.12B (Mar-25) and sits at 38% of fixed assets — a visible bulge that should normalise when assets commission. The forensic concern is not that capex is fictitious. It is that depreciation will step up materially in FY27-28 once these projects commission, compressing reported margins. That is a forecasting question more than a shenanigans one.

Cash Flow Quality

Operating cash flow is consistently larger than net income, which is the right shape for a heavy-D&A business. The mechanism is not aggressive — it is genuine working-capital efficiency from a sector that operates with negative cash conversion. But two episodes deserve scrutiny.

CFO leads net income, but the gap is mostly D&A

CFO/NI in three rolling years is 1.97×, in five years 1.82×. The accrual ratio (NI - CFO)/Avg Total Assets is negative for FY23, FY24, and FY25 — earnings are being collected, not booked. FY24's $8.53B CFO is the one outlier worth pausing on: it was roughly twice FY23 and twice FY25. Inventory days dropped from 84 in FY22 to 70 in FY24 (and further to 63 in FY25), and debtor days went from 7 to 6 — a working-capital release worth approximately $1.8-2.4B that flattered FY24 CFO. By FY25, working capital had largely normalised. This is not a manipulation flag; it is a reminder that a single year's cash flow does not generalise.

Free cash flow after capex is thin

Five-year FCF is $9.52B; five-year net income is $14.58B. FCF/NI = 0.65×. That is below 1.0× because growth capex is a real cash outflow whose benefits accrue in future periods. The company has paid roughly 30-50% of net income out as dividends over the same period, so a meaningful portion of the dividend has been funded by debt growth — total debt rose from $15.93B (Mar-21) to $17.82B (Mar-25). This is consistent with the CARO ix(d) qualification: short-term funds have been bridging the long-term capex / dividend gap, $7.17B of which is still outstanding at year-end.

Metric Hygiene

IOC does not run an aggressive non-GAAP narrative. The metrics it surfaces in earnings releases and the integrated annual report are operational throughput and pricing variables, not adjusted earnings constructs. There is no "adjusted EBITDA", no "cash earnings", no "core operating profit" exclusion. The headline numbers reconcile to the audited financials.

The cleanest part of the IOC disclosure is also its biggest test: the same fact appears in two different forms across the FY2025 Annual Report. The CARO Annexure (page 197) reports $15.3M of fraud across 11 instances; the Directors' Report (page 123) reports $12.24M across 7 instances. Both are accurate within their own scope — CARO captures all reported frauds; the Directors' Report captures only those auditors have referred to MCA under Section 143(12) — but a reader scanning either disclosure in isolation gets a different headline. We treat the discrepancy as a disclosure-clarity issue, not a manipulation, but a careful PM should reconcile both totals before quoting either.

Where the LPG bottling fraud sits

Two-thirds of the disclosed fraud — $10.06M — sits in a single LPG bottling plant under the North-East Integrated State Office, with another $1.64M across two more LPG bottling sites. The character of the fraud is operational (pilferage masked by inflated transit losses and inflated market-return cylinder claims), not financial-statement-level. The auditors confirm in CARO that "no report has been filed by the auditors in Form ADT-4 as prescribed under Rule 13 of Companies (Audit and Auditors) Rules, 2014 … However, we are in the process of complying with the provisions of section 143(12) of the Act with respect to the fraud reported under sub-clause (a) above." The audit-committee-level investigation is still procedurally open at the date of audit sign-off (30 April 2025). A clean ADT-4 outcome in FY26 would be a positive — a delayed or expanded outcome would not.

What to Underwrite Next

The forensic work changes how a PM should size IOC, not whether they can own it. The numbers are real; the cash is real; the cycle is real. What is not real-yet is the resolution of three open threads.

The accounting risk in IOC is not a thesis breaker. It is a position-sizing limiter. A reader who is bullish on the structural BPCL-discount narrative should accept that the next 6-12 months carry a procedurally open ADT-4 process, an unresolved internal FCPA review, a 5-quarter SEBI LODR fine pattern, and a $7.17B balance-sheet structural footnote that the auditors qualified for the second consecutive year. None of these change FY27E EPS materially. All of them widen the appropriate margin of safety. We would underwrite IOC at the lower end of the historical 4-7× P/E band until ADT-4 closure and the FY26 Directors' Report disclose the Albemarle review's outcome.