Accenture at 7x FCF — the cheapest quality name in IT services, or a franchise whose unit of sale is being automated?

**Quick background**: Accenture (NYSE: ACN) is the largest IT-services and consulting firm in the world — \~780,000 people, roughly half strategic consulting and half multi-year managed services, capital-light, ROIC near 20%, net cash on the balance sheet. After a steep de-rating it trades around $129 (≈$79B market cap, as of 26 Jun 2026), about 9.5x forward earnings, \~7.1x free cash flow, a \~14% FCF yield and a \~5% dividend — broadly in line with Cognizant, a lower-quality peer, and below the IT-consulting median. Q3 FY26 (reported 18 June) was operationally fine: revenue +6% USD, 17.0% operating margin, EPS +9%. The stock still fell \~15% on the day.

The interesting part of the bull case: the claim is this is the 2014 cloud short all over again — back then the thesis was "cloud means 15-25% less integrator work," it was wrong, and the stock 5x'd over seven years. The variant view is that AI eats the low-value mechanical work while strategic design and project coordination become more valuable, and that companies hire Accenture for organisational/social reasons that don't change with the tooling.

Where I kept poking holes:

* Bookings are already rolling: Q3 new bookings $19.3B, -3% LC, book-to-bill \~1.0. Consulting only +3% LC ex-FX (+1% LC in the soft spot). The leading indicator is softening before any macro shock.
* \~Half of revenue is managed services that reprices at renewal on a 1-3 year cycle. Advanced-AI revenue is only \~$2.7B (FY25) against a multi-tens-of-billions installed base — the offset may not scale before the base reprices.
* The margin rests on a pyramid of low-cost juniors billed at a markup — exactly the layer AI removes first. Hold margins only if they move 800k people up-value, fast.
* $25.3B goodwill (37% of assets) from \~$3B/yr of M&A blends bought growth into the organic line; FCF looks pristine partly because acquisitions sit in investing cash flow.

**One genuinely interesting angle**: a reverse DCF (FCF \~$11B, 9% cost of equity) says at \~$79B the market is pricing cash flows to shrink \~4% a year in perpetuity. Flat FCF forever is worth \~$122B (\~$200/sh). So you're not paying for growth — you're being paid a \~12% shareholder yield to take the other side of "AI kills consulting." For contrast, Adobe sits at almost the same \~13% FCF yield right now, but its AI exposure is two-sided (Firefly is a product it monetises inside recurring \~92%-retention ARR), where Accenture's AI hits the billable hour and passes savings to clients. Same cash yield, opposite shape of risk.

**Closing question**: the whole thing turns on one timing question I can't resolve from the disclosure — does the installed managed-services base reprice downward faster than advanced-AI revenue scales to replace it? For those closer to enterprise IT budgets: are you seeing renewal pricing actually compress yet, or is the bookings dip still just lumpiness?

Full write-up with more context: [https://fmarinisecondopinion.substack.com/p/acn-accenture-plc](https://fmarinisecondopinion.substack.com/p/acn-accenture-plc)