Financial Shenanigans

Financial Shenanigans

Transocean's reported numbers are not obviously cooked, but they are aggressively framed. Underlying cash conversion is honest — operating cash flow ex-impairment tracks management's Adjusted EBITDA reasonably well, days sales outstanding is falling, and there is no goodwill, no factoring disclosure, and no auditor turnover. The forensic risk sits in three places: a recurring "non-recurring" impairment series ($57M, $772M, $3,049M over FY2023-FY2025) that coincides with a CEO transition and is excluded from the bonus metric the comp committee paid at 138% of target; a pending securities-class-action that names the exact rigs later impaired; and a capex-to-depreciation ratio that has collapsed to 0.19x as the fleet is being wound down ahead of the pending Valaris merger. The one data point that would most change the grade is whether the FY2026 10-K shows a further round of held-for-sale impairments on the remaining ultra-deepwater fleet.

The Forensic Verdict

Shenanigans scorecard

Breeding Ground

The breeding-ground signals are mixed but tilted toward "watch." The 2025 CEO change (Jeremy Thigpen to Executive Chair; Keelan Adamson to CEO) coincided with the largest impairment in the company's recent history and lifted the FY2025 net loss by $3.05B. The bonus plan continues to be 65% weighted to Adjusted EBITDA, and the comp committee paid 138% of target on that line by certifying a $1.39B outcome (slightly above the press-release Adjusted EBITDA of $1.37B) after applying "a further adjustment to reflect the effect of lower expected return due to realigned investment strategies." Auditor independence, board size, and audit-committee structure look ordinary; Ernst & Young remains and is up for re-ratification at the 2026 AGM.

The most material item in this section is not any single bullet — it is the alignment of incentives. The CEO and four other named executives received a bonus equal to 150% of target in a year that produced a $2.92B GAAP loss, $3.05B in asset impairments, 26% share dilution, and a 27% backlog decline. The mechanism that makes this possible is the deliberate exclusion of impairments from Adjusted EBITDA and the long-term incentive's anchor on TSR (where credit-rating-driven rallies count) and Free Cash Flow (where a capex collapse counts as a win).

Earnings Quality

Reported earnings are dominated by the held-for-sale impairment line. Strip it out and the underlying P&L is improving: revenue +13% to $3.965B, operating-and-maintenance cost only +9%, depreciation falling because the fleet is shrinking, and adjusted operating margin clearly positive. The issue is not the existence of charges — it is the cadence. Impairments have grown every year for three years on the same asset group, and the FY2024 and FY2025 charges hit rigs that were carried at multiples of realizable value.

The green bar is operating income with the impairment line added back. Underlying operations turned positive in FY2024 and stepped up materially in FY2025 — Adjusted EBITDA grew 19% to $1.37B and Adjusted Net Income flipped from $-54M to $+37M. Layered on top, the red bar is the impairment charge, growing four-fold each year. Three "non-recurring" charges in a row, on the same asset group, are a pattern.

This is the cleanest test in the report. Revenue is up 56% over five years while receivables are down 7%. DSO has compressed from 77 days to 51 days. There is no evidence of channel stuffing, of bill-and-hold, of unbilled receivables accumulation, or of any other classic premature-revenue mechanism. Collections are getting stronger, not weaker.

Capex/D&A of 0.19x in FY2025 is the lowest reading in the dataset and well below the 0.5-0.8x range typical of a steady-state offshore driller. Two explanations compete. The benign one: the newbuild Deepwater Aquila delivered in FY2024, six ultra-deepwater floaters were sold for $71M in FY2025, and depreciation is high because nine rigs are still on the books being impaired. The skeptical one: capex deferral inside an acquirer's last full year ahead of a stock-for-stock merger is a known pattern, because the combined entity will absorb the deferred maintenance. The merger announcement (February 2026) makes this a watch item rather than a confirmed red flag.

Cash Flow Quality

Operating cash flow is real, but it is not as durable as the headline suggests. FY2025 CFO of $749M was reported as +68% year-over-year. Roughly $204M of that increase comes from a non-cash mark on the bifurcated derivative embedded in the 4.625% exchangeable bonds — it shows up as interest expense reduction in the prior-year comparison and as add-back in non-cash items. Working capital drained $109M, a smaller drag than the $242M drain in FY2024. The MD&A explicitly cites "decreased cash paid to suppliers" as a CFO tailwind, which is consistent with accounts payable falling from $255M to $242M (a use, not a source) — so the language is misleading in direction.

The chart shows the central forensic problem: net income is shaped by impairments, CFO is reasonably steady, and FCF benefits from a capex collapse in FY2025. The three-year (FY2023-FY2025) FCF total of $556M against cumulative net losses of $4.38B is an artifact of the impairment accounting and is not by itself a red flag. The yellow flag is the trajectory of FCF: $193M in FY2024 became $626M in FY2025 partly because capex fell from $254M to $123M, a 52% cut that would be unusual outside of a pre-merger wind-down.

The CFO line is a non-cash story: $3.77B of add-backs (impairment, D&A, derivative mark, debt-exchange loss) bridge a $2.9B loss to a $749M CFO. Holders of recurring cash flow theses on RIG should price that the engine is the add-back, not the underlying P&L.

Metric Hygiene

The largest forensic risk in this report sits in the non-GAAP framing. Adjusted Net Income, Adjusted EBITDA, and Free Cash Flow are the metrics management leads with, and they are the metrics the bonus and PSU plans pay against. The reconciliation is disclosed (per the FY2025 earnings release), but the gap is unusually large and the items adjusted out are recurring.

A bridge that turns a $2.9B GAAP loss into a $37M adjusted profit is, in the language of the playbook, a misleading metric. The $3.0B impairment add-back is treated as non-recurring even though impairments have grown every year for three years on the same asset group. The $99M debt-to-equity conversion loss is also recurring in spirit — Transocean issued 73.3M shares to bondholders during 2025 alone. The $179M of "discrete favorable tax items" was deducted from Adjusted Net Income — appropriately so under the rules, but worth noting that the company removes favorable tax noise from the adjusted number while leaving in the operating beat.

Total weighted bonus payout: 150% of target — in a year that produced a $2.9B GAAP loss, a $3.05B impairment, a 27% backlog decline, and a 26% share count increase. None of those items is in the bonus formula.

The backlog disclosure deserves a separate look because it is the single best forward indicator for this business.

Backlog dropped from $8.33B to $6.06B in 12 months — a $2.26B decline against FY2025 revenue of $3.97B. The company added $839M of new bookings at an average dayrate of $453K in the year. Run-off exceeded additions by roughly $1.4B. This is the metric management is least promotional about; it is also the most diagnostic.

What to Underwrite Next

The forensic risk in Transocean is real but bounded. It is not a fraud story — there is no restatement, no auditor turnover, no off-balance-sheet receivable program, and no related-party revenue. It is a "stretching" story: management is steering investor attention toward Adjusted EBITDA and Free Cash Flow while the GAAP P&L absorbs a recurring impairment cycle that will compound through the Valaris merger.

The two signals that would change the grade in opposite directions are symmetric. A downgrade signal would be a fourth consecutive year of held-for-sale impairments on the residual fleet in FY2026, or any unfavorable ruling at the motion-to-dismiss stage in the securities suit — either would push the grade into "High." An upgrade signal would be the FY2026 10-K showing capex returning to a $300M+ run-rate, combined with a quiet impairment line — that would indicate FY2025 was a one-time pre-merger reset rather than a recurring problem.

For a PM, the practical implications are: this is a position-sizing limiter and a multiple haircut, not a thesis breaker. The non-GAAP gap means earnings-based valuation should reference Adjusted EBITDA minus a normalized impairment reserve (a conservative reserve would be ~$300-500M per year against the remaining fleet book value). The FCF figure that supports the bull case should be discounted by the capex deferral element until post-merger capex disclosure normalizes. Covenant cushion is adequate but the rating is CCC+, so any reported EBITDA miss against the $1.034B bonus threshold has outsized credit-spread implications.