The recurring fourth-quarter compression — operating margin of 11.6% in 2025Q4 and 13.8% in 2026Q4 against 16.8–17.0% in adjacent quarters — suggests seasonality in utilization, year-end bonus costs, and project ramp-downs rather than deterioration, so a mid-cycle operating margin near 15–16% is likely the more defensible measure of true earning power. Net margin of 8.0–12.9% and free cash flow margin swings of 4.9–21.6% appear largely driven by working-capital timing rather than operating performance, and the isolated $817.3M charge booked in 2025Q4 distorts that quarter's comparison. The absence of a rising gross margin trend around 32% implies limited evidence that the consulting-to-operations mix or offshore leverage is structurally lifting unit economics.

With ROIC falling to 5.2–6.3% in both 2025Q4 and 2026Q2/Q4, the return profile appears to be maintained by margin seasonality rather than improving capital efficiency, since asset turnover simultaneously slipped from 0.31 to 0.26 over the same period. That turnover decline coincides with total assets expanding roughly 36% against only about 14% revenue growth, driven substantially by goodwill accumulation to $26.8B — approximately 36% of total assets — which inflates the denominator without a demonstrated proportionate contribution to earnings. Analysts should watch whether ROIC recovers as acquired capabilities are absorbed; if it does not, the interpretation that goodwill-heavy expansion is diluting rather than enhancing returns on invested capital would gain weight.

The forward P/E of 15.3x versus 17.5x trailing embeds roughly mid-teens EPS growth, which appears to exceed recent organic revenue expansion of 5.6–7.4% and therefore seems to rely on margin normalization and buyback-driven per-share accretion rather than top-line reacceleration. The 4.16x price-to-book multiple, roughly double Cognizant's 1.95x and above Infosys' 4.61x only marginally despite comparable or superior growth optics, suggests the market is paying for perceived quality of the client franchise — not for a widening return spread. A 10.0x trailing EV/EBITDA against IBM's 15.2x looks inexpensive in isolation, but with D/EBITDA rising from 1.55x to 3.80x, EV-based multiples should be read alongside the leverage build rather than in isolation.

Even at the latest 24.9x coverage — the lowest in the series versus 235x in 2024Q3 — the level appears comfortable relative to a services model with structurally low capital intensity, since capital expenditure runs below depreciation at roughly 0.6–1.3% of revenue. The more meaningful signal is compositional: with cash more than doubling to $12.8B alongside the debt build, the firm may be carrying leverage to fund dividends, buybacks, and acquisition activity rather than operating necessity, which suggests the return on incremental debt deserves scrutiny. Given that goodwill is now roughly 36% of total assets, investors should monitor whether leverage ratios mask a balance sheet whose equity quality is increasingly intangible-weighted rather than an outright solvency concern.

The turnover decline is largely arithmetic rather than operational: liabilities grew about 65% and total assets about 36% against revenue expansion of roughly 14%, so the efficiency signal here reflects acquisition pace and cash accumulation more than delivery productivity. DPO has been stable in a narrow 19–23 day range, indicating modest supplier leverage rather than a company squeezing its vendor base, and the days-sales-outstanding spike in 2026Q4 versus 73–79 days in prior quarters appears inconsistent with the underlying working-capital pattern and may reflect a partial-period or data anomaly rather than a genuine collections improvement. Cash conversion cycle data is unavailable given the absence of inventory disclosures, so conclusions about end-to-end working capital efficiency should be treated as provisional.