Full Report

The numbers behind CGI Inc.: as-reported financial statements and company metrics for FY2021–FY2025, traced to the source filings, opened with the share-price history those statements have to justify. Every linked figure opens the exact page of the filing it was printed on, with the statement row highlighted. Amounts in C$ thousands unless noted.

Reading notes: All figures are as printed in CGI's filings: in thousands of Canadian dollars (per-share amounts in CAD), fiscal years ended September 30. Display currency conversion to USD, if any, is handled by the renderer using period-end FX. Each FY2021–FY2025 column of the income statement, balance sheet, cash-flow statement and revenue-by-service-line breakdown is cited to that year's own annual report / consolidated financial statements. FY2016–FY2020 rows in the Long-Term Record are from the standardized data feed (CGI as-reported history) and are shown without page links. Revenue by service line (Managed IT BPS vs. Business/strategic IT consulting systems integration) is CGI's only revenue disaggregation reported on a consistent basis across all five years, so it is used as the primary breakdown. In FY2021 the two service lines are printed in the reverse order and the consulting line is labelled 'Business consulting, strategic IT consulting and systems integration'.

Share Price — Available History Since January 2026

The stock closed at C$94.85 on Jul 10, 2026 — down 28% over the window shown, trading between C$85.04 and C$131.51. At that close the stock trades at 13× FY2025 diluted EPS as reported below.

Source: market price feed, daily closes, Jan 2026–Jul 2026 — the feed marks this available history as partial. Price return only, excludes dividends.

FY2025 at a Glance

Source: FY2025 consolidated statements [1] [2] [3] [4]. Click any linked figure to open the filing page with the row highlighted.

Revenue by Service Line

Revenue by Service Line	FY2021	FY2022	FY2023	FY2024	FY2025
Managed IT and business process services	6,722,967	6,980,988	7,674,460	8,041,857	8,820,863
Business and strategic IT consulting and systems integration services	5,403,826	5,886,213	6,621,900	6,634,295	7,091,810
Total revenue	12,126,793	12,867,201	14,296,360	14,676,152	15,912,673
Total revenue growth, derived	—	+6.1%	+11.1%	+2.7%	+8.4%

Source: Note 29 (Note 28 in FY2021–FY2022) Segmented Information — Information about Services [5] [6] [7] [8]. Click any linked figure to open the filing page with the row highlighted.

Income Statement

Source: Consolidated Statements of Earnings [1] [2] [3] [4]. Click any linked figure to open the filing page with the row highlighted.

Columns marked E are consensus analyst estimates shown alongside reported results for direct comparison; they are not company guidance.

Estimate source: analyst consensus (claude_web), as of 2026-07-12. Forecasts carry no filing page links.

Balance Sheet

Balance Sheet	FY2021	FY2022	FY2023	FY2024	FY2025
Cash and cash equivalents	1,699,206	966,458	1,568,291	1,461,145	864,209
Accounts receivable	1,231,452	1,363,545	1,425,117	1,398,402	1,613,777
Work in progress	1,045,058	1,191,844	1,143,685	1,208,095	1,367,989
Total current assets	4,765,138	4,349,047	4,933,727	4,817,306	5,053,179
Intangible assets	506,793	615,959	623,103	718,575	888,006
Goodwill	8,139,701	8,481,456	8,724,450	9,470,376	11,744,782
Total assets	15,021,021	15,175,420	15,799,499	16,685,468	19,521,828
Current portion of long-term debt	392,727	93,447	1,158,971	999	845,253
Total current liabilities	3,803,472	3,649,331	4,645,799	3,549,091	5,102,170
Long-term debt	3,008,929	3,173,587	1,941,350	2,687,309	2,792,582
Total liabilities	8,034,789	7,902,696	7,489,205	7,257,478	9,239,496
Total equity	6,986,232	7,272,724	8,310,294	9,427,990	10,282,332

Source: Consolidated Balance Sheets [9] [10] [11] [12]. Click any linked figure to open the filing page with the row highlighted.

Cash Flow

Source: Consolidated Statements of Cash Flows [13] [14] [15] [16]. Click any linked figure to open the filing page with the row highlighted.

Revenue by Operating Segment (Geographic)

Revenue by Operating Segment (Geographic)	FY2021	FY2022	FY2023	FY2024	FY2025
Western and Southern Europe	—	—	—	2,600,198	2,679,167
U.S. Commercial and State Government	—	—	—	2,327,309	2,522,956
U.S. Federal	—	—	—	2,001,391	2,247,943
Canada	—	—	—	2,034,995	2,090,682
Scandinavia, Northwest and Central-East Europe	—	—	—	1,593,434	1,688,887
U.K. and Australia	—	—	—	1,584,833	2,020,016
Germany	—	—	—	894,565	901,609
Finland, Poland and Baltics	—	—	—	859,263	903,701
Asia Pacific	—	—	—	956,145	1,014,441
Intersegment eliminations	—	—	—	(175,981)	(156,729)
Total revenue	—	—	—	14,676,152	15,912,673

Source: Note 29 Segmented Information — operating segments restated on the FY2025 geographic-delivery basis; comparable series available for FY2024–FY2025 only [17]. Click any linked figure to open the filing page with the row highlighted.

Long-Term Record

Fiscal year	Total revenue	Net earnings	Diluted earnings per share	Cash provided by operating activities
FY2016	10,683,264	1,068,716	3.42	1,333,074
FY2017	10,845,066	1,035,195	3.41	1,358,552
FY2018	11,506,825	1,141,402	3.95	1,493,408
FY2019	12,111,236	1,263,207	4.55	1,633,919
FY2020	12,164,115	1,117,862	4.20	1,938,556
FY2021	12,126,793	1,369,072	5.41	2,115,928
FY2022	12,867,201	1,466,142	6.04	1,864,998
FY2023	14,296,360	1,631,249	6.86	2,112,249
FY2024	14,676,152	1,692,715	7.31	2,204,983
FY2025	15,912,673	1,658,285	7.35	2,234,197

Source: consolidated statements across filings; older years from the standardized feed [13] [1] [14] [2]. Click any linked figure to open the filing page with the row highlighted.

Operating KPIs

KPI	FY2021	FY2022	FY2023	FY2024	FY2025
Bookings	—	—	16,259,000	16,044,000	17,572,000
Backlog	—	—	26,059,000	28,724,000	31,451,000

Source: company-reported operating metrics [18]. Click any linked figure to open the filing page with the row highlighted.

Analyst Consensus

Street ratings: Moderate Buy / Buy consensus. TSX GIB.A: 8 Buy, 4 Hold, 0 Sell (12 analysts). S P Global poll of 15 analysts also rates "Buy" (~75% buy). US-listed GIB (USD): 15 analysts, 5 Strong Buy / 4 Buy / 5 Hold / 1 Sell.

Estimate source: analyst consensus (claude_web), as of 2026-07-12. Forecasts carry no filing page links.

Traceability

261 of 285 figures on this page (92%) link to the filing page where they are printed — click a linked figure to open the source PDF at that page with the row highlighted. Unlinked figures come from standardized data feeds or pre-filing years.

• All figures are as printed in CGI's filings: in thousands of Canadian dollars (per-share amounts in CAD), fiscal years ended September 30. Display currency conversion to USD, if any, is handled by the renderer using period-end FX.

• Each FY2021–FY2025 column of the income statement, balance sheet, cash-flow statement and revenue-by-service-line breakdown is cited to that year's own annual report / consolidated financial statements.

• FY2016–FY2020 rows in the Long-Term Record are from the standardized data feed (CGI as-reported history) and are shown without page links.

• Revenue by service line (Managed IT BPS vs. Business/strategic IT consulting systems integration) is CGI's only revenue disaggregation reported on a consistent basis across all five years, so it is used as the primary breakdown. In FY2021 the two service lines are printed in the reverse order and the consulting line is labelled 'Business consulting, strategic IT consulting and systems integration'.

• Revenue by operating segment (geographic) was restated in fiscal 2025 (new segments including a separate Germany and a combined Scandinavia, Northwest and Central-East Europe), and the segment structure also changed in FY2022 and FY2024. A like-for-like multi-year geographic series is therefore not available; the geographic segment table shows FY2024 and FY2025 only, both on the current restated basis from the FY2025 report (Note 29).

• Income-statement special charges: FY2021–FY2024 print 'Acquisition-related and integration costs' (with a separate 'Cost optimization program' line in FY2023–FY2024); FY2025 combines these into 'Restructuring, acquisition and related integration costs'. Rows are populated from each year's printed line(s); the small net foreign-exchange line is included so operating expenses reconcile.

• CGI initiated a quarterly cash dividend in fiscal 2025 (C$135,052 thousand paid); no dividends were paid in FY2021–FY2024.

• Bookings and Backlog KPIs are CGI order metrics printed in millions of Canadian dollars in the FY2025 report highlights (Bookings/Backlog cited for FY2023–FY2025); shown scaled to thousands to match the tab.

• Quarterly figures come from CGI's earnings-release Financial Highlights (single-quarter, in millions), which do not include full interim balance sheets or cash-flow statements; only the income-statement highlights and single-quarter operating cash flow are shown, scaled to thousands.

• 2 figure(s) differed between the data feed and the filing; the filing value is shown (see the run's metrics/metrics_tab.json for the audit trail).

---

The bottom line

CGI is one of the world's largest independent IT and business consulting firms — roughly C$15.9 billion of revenue, about 94,000 people, built over five decades by a Montréal founder who still controls the vote. Its record is a decade of near-double-digit per-share earnings growth funded almost entirely by its own cash. This chapter maps what CGI is and how it compounds, then frames the question the rest of this report exists to test.

What CGI is

Founded in 1976 and headquartered in Montréal, CGI sells information-technology and business-consulting services to large organizations, with approximately 94,000 consultants and professionals worldwide [1]. It is a people business: revenue comes from billing that workforce out to design, build, run and maintain the technology its clients depend on. CGI organizes delivery around a "proximity" model — teams that sit where the client lives and works — backed by lower-cost global delivery centers [2].

The work splits into two streams. Roughly 55% of revenue is managed services — multi-year contracts to run a client's IT or business processes, which produce recurring, visible revenue. The other 45% is systems integration and consulting (SI&C) — shorter, project-based work that rises and falls with discretionary technology budgets [3]. Government is the single largest end-market at about 38% of revenue, followed by financial services and manufacturing/retail/distribution — a demand base skewed toward long-cycle, non-discretionary buyers [4].

Source: FY2025 Annual Report, MD&A 3.4 Revenue by Segment [5].

No single geography dominates. CGI delivers through nine geographic segments spanning the U.S., Canada, the U.K. and Australia, France, the Nordics, Germany and Central Europe — a footprint that spreads risk but also means reported growth carries a large foreign-currency component, a point that matters below.

How it compounds — Build and Buy

CGI describes its playbook as a "Build and Buy" profitable-growth strategy, and it is central to understanding how the company creates value. It has four pillars: win, renew and extend contracts, and land new large managed-services deals (the Build, or organic, half); plus "metro market" acquisitions of local and niche firms and larger "transformational" acquisitions (the Buy half) [6]. Management is explicit that it intends to remain "a consolidator in the IT and business consulting services industry" [7].

The mechanics are what make it a compounding machine rather than an ordinary roll-up. CGI is asset-light — capital expenditure ran just 0.7% of revenue in FY2025 — and converts more than 100% of net earnings into free cash flow [8]. That cash funds acquisitions of smaller IT-services firms, which are integrated onto CGI's common operating framework, and the enlarged base then throws off more cash to fund the next deal. Over the last twelve months to April 2026, CGI announced the purchases of Daugherty, BJSS, Novatec, Momentum, Apside and a Comarch subsidiary — the BJSS deal alone lifting reported U.K. and Australia revenue by 27.5% in FY2025 [9].

The record: cash-funded compounding

The output of that engine is a decade of steady per-share growth. Diluted earnings per share rose from C$5.41 in FY2021 to C$7.35 in FY2025, compounding near 12% a year across the five years, while free cash flow held around C$2 billion annually and return on invested capital ran in the mid-to-high teens [10].

Source: FY2025 Annual Report, MD&A 2.1 Selected Yearly Information & Key Performance Measures [11].

Source: FY2025 Annual Report, MD&A 2.1 Selected Yearly Information [12].

Two features stand out. Cash generation is exceptional and consistent: free cash flow was C$2.12 billion in FY2025, or 1.28 times reported net earnings, a conversion rate above one that CGI has sustained for years [13]. And the order book is large and visible: backlog stood at C$31.5 billion — roughly two years of revenue — with a book-to-bill ratio of 110.4% in FY2025, meaning new bookings outpaced revenue recognized [14]. For a cold reader, this is the shape of the business: unglamorous, cash-rich, defensively positioned, and run for per-share compounding.

What changed in FY2025

The chart above hints at the tension. Revenue grew 8.4% in FY2025, but net earnings fell C$34.4 million, or about 2%, and reported diluted EPS was essentially flat — C$7.35 against C$7.31 [15]. Two things drove the gap between top-line growth and per-share progress.

First, the growth was flattered by currency and acquisitions rather than underlying demand. Of the 8.4% reported growth, 3.8 percentage points came from a weaker Canadian dollar; constant-currency growth was 4.6%, and much of that came from acquired revenue [16]. The organic reality is starker one year back: in FY2024, constant-currency growth was just 0.9% [17]. Western and Southern Europe, CGI's largest segment, actually shrank 1.7% before currency in FY2025 [18].

Source: FY2025 Annual Report, MD&A 2.1 Selected Yearly Information [19].

Second, profitability slipped as CGI absorbed newly bought businesses. The pre-tax margin fell to 14.1% from 15.6%, and return on invested capital dropped to 13.6% from 16.0% as freshly deployed acquisition capital had not yet earned its keep [20].

Management prefers an adjusted lens: on that basis, adjusted diluted EPS rose 8.9% to C$8.30, stripping out acquisition and integration costs [21]. The gap between reported EPS of C$7.35 and adjusted EPS of C$8.30 — about C$0.95, or roughly 11% of the adjusted figure — is itself worth watching, because the add-backs are largely the cost of the acquisition strategy that is now carrying growth. The chart below shows both lines, and the deceleration underneath them.

Source: FY2025 Annual Report, MD&A 2.1 Selected Yearly Information; adjusted EPS disclosed from FY2023 [22].

The market has noticed. The shares traded at C$131.51 in mid-January 2026 and sit at C$94.85 in early July — near the low end of a 52-week range of roughly C$83 to C$132 — which puts CGI at about 12.9 times reported trailing earnings, 11.4 times adjusted earnings, and 10.7 times the FY2026 consensus EPS of C$8.90. Sell-side analysts carry a consensus target near C$126.

Source: market data as of July 2026 and consensus estimates, as reported.

Capital and control

Two facts frame how value reaches — and stays with — shareholders. First, capital allocation. CGI has historically returned cash through buybacks, repurchasing 8.86 million Class A shares in FY2025 for C$1,258.5 million at an average price of C$142.01 [23]. That average sits well above today's C$94.85, a reminder that even a disciplined buyer can repurchase into a de-rating. In a notable shift, CGI also initiated a quarterly dividend in FY2025 — nil the prior year — and raised it to C$0.17 per share in November 2025 [24].

Second, control. CGI has a dual-class structure: about 195 million Class A subordinate voting shares carry one vote each, while 24.1 million Class B multiple voting shares carry ten votes each [25]. Founder Serge Godin holds 100% of the Class B shares, giving him about 55.6% of the votes on roughly 11.25% of the equity [26]. Public Class A holders own the economics but do not control the outcome; the founder does. That has coincided with a long record of disciplined execution, and it concentrates key decisions — succession, capital allocation, the pace of acquisition — in one family's hands.

The through-line

CGI's appeal to long-term owners is a decade of compounding per-share earnings at a near-double-digit rate, built by pairing steady organic growth with disciplined, self-funded acquisitions. The question this report exists to answer is whether that engine keeps running now that organic growth has slowed to low-single digits, reported growth leans increasingly on acquisitions and currency, margins and returns on capital have dipped as deals are digested, and the shares have de-rated to roughly 11 times forward earnings. The chapters that follow test the pieces of that question — the durability of the moat and the order book, the true economics of the acquisition machine, the gap between reported and adjusted earnings, what artificial intelligence does to a labor-based services model, and what the current price implies.

---

Acquisition Math

CGI spent about C$1.9 billion on five acquisitions in FY2025, its busiest year of dealmaking in more than a decade. Almost the entire price — C$1.67 billion of goodwill (87%) plus C$247 million of client-relationship and backlog intangibles — was intangible; the acquired firms carried almost no net tangible assets. On the numbers the businesses came with, CGI paid roughly 40 times their earnings. Whether that clears its cost of capital depends on a margin lift management has promised but not yet delivered.

The spine of this report is a compounding engine now leaning more heavily on the "Buy" half of Build-and-Buy (Build and Buy). This chapter looks at what that half actually costs and what has to happen for it to pay.

The machine reaccelerated

For three years the acquisition machine idled. Goodwill on the balance sheet — the running tally of what CGI has paid above the fair value of the net assets it bought — barely moved from FY2021 to FY2023, sitting near C$8.5 billion. FY2024 nudged it up. Then FY2025 added C$2.27 billion in a single year, a 24% jump, the largest step in the company's recent history.

Source: Consolidated Balance Sheets, FY2021–FY2025 Annual Reports [1]; [2]; [3].

The five deals span the geography and mix of CGI's own footprint. BJSS, a U.K. technology and engineering consultancy of roughly 2,400 people, was the anchor at C$1,255.6 million. Daugherty, a St. Louis AI and data-analytics firm of about 1,100, went for C$343.0 million; Apside, a French digital-engineering house of about 2,500 people, for C$229.9 million; Novatec (Germany) and Momentum (Québec City) were smaller and disclosed only in aggregate [4].

Sources: FY2025 Annual Report, MD and A, Investment in Subsidiaries [5]; Note 27 Investments in Subsidiaries [6]. Novatec and Momentum prices grouped under "Other" per the filing.

The C$1.83 billion of cash that left the business for acquisitions (net of cash acquired) was covered by the year's C$2.12 billion of free cash flow, but not with much room to spare once buybacks and the new dividend are added. Cash on hand fell from C$1.46 billion to C$864 million, the company issued US$650 million of five-year notes in March 2025, and interest on long-term debt rose to C$80.9 million from C$48.0 million [7] [8]. The self-funding flywheel still turns; the buffer thinned.

What the price actually buys

An IT services firm has little to sell but its people and its client contracts, so its acquisitions convert almost entirely into goodwill and intangibles rather than plant or inventory. CGI's FY2025 deals are a clean illustration. Of the C$1.93 billion in total consideration, C$1.67 billion (87%) landed as goodwill and a further C$247 million as intangible assets — described in the filing as mainly client relationships and backlog. Together they are 99% of the price. Net tangible assets acquired, including the cash that came with the targets, were roughly break-even [9].

Source: Note 27 Investments in Subsidiaries, FY2025 Annual Report [10]. Net tangible assets derived as consideration less goodwill and intangibles.

That is not unusual for the model, but the accumulation is worth sizing. Goodwill now stands at C$11.74 billion — 60% of total assets [11] and larger than the company's entire C$10.28 billion of shareholders' equity [12]. CGI has never recorded a goodwill impairment in its history [13].

Source: Consolidated Balance Sheets, FY2025 Annual Report [14].

A clean impairment record across thirty years of dealmaking — including through 2020 — is a genuine mark of discipline: CGI has historically bought small, integrated fast, and avoided the transformational deal that blows up. It also means C$11.7 billion of acquired value has never been tested against a write-down, so the strength of the record and the size of the untested balance are the same fact seen from two sides.

The margin that has to arrive

The revealing disclosure sits in Note 27. CGI tells you what its three largest FY2025 targets — Daugherty, BJSS and Apside — would have earned over a full year: on roughly C$1,120 million of combined revenue, about C$46 million of net earnings, excluding acquisition and integration costs. That is a 4.1% net margin, against CGI's own 10.4% [15].

Set that against the C$1,828.5 million CGI paid for those three. On the earnings they came with, the price is roughly 40 times net income. Lift the acquired revenue to CGI's group margin and the same price falls to about 16 times. CGI's own shares trade near 11 times forward earnings (Build and Buy) — so even a fully margin-normalized deal was bought at a premium to CGI's own multiple. The value case rests on two things landing: the margin lift, and revenue synergies on top.

Source: derived from Note 27 pro-forma revenue and net earnings [16] and combined purchase prices [17]; CGI forward multiple per Chapter valuation snapshot.

Management is explicit that the lift is the plan, and that the mechanism is CGI's own operating system. On the Q4 FY2025 call, CFO Steve Perron said savings "are coming a lot faster when we are using our system, when we are using our processes," and that as the most recent deal, Apside, is folded in, "margin will improve. It's part of the plan." CEO François Boulanger described cross-selling CGI's managed-services and offshore delivery into the acquired client bases — pitching India-based delivery to BJSS and Daugherty clients — as the source of "1 plus 1 equal 3" [18].

There is early corroboration. In the same quarter CGI delivered a 16.6% operating margin even while integrating several deals, and the U.K. and Australia segment grew 28% with BJSS folded in [19]. Daugherty, BJSS and Apside contributed C$551 million of revenue and C$46 million of net earnings from their acquisition dates to year-end, so the businesses are performing broadly in line with the pro-forma [20].

Two costs of digestion show up in the meantime. Acquisition and integration costs jumped to C$88.2 million from C$5.9 million a year earlier, and a separate C$196.8 million restructuring of the Continental European operations ran through FY2025 [21]. The integration costs are the same add-backs that widened the gap between reported and adjusted earnings (Build and Buy); they are also the reason return on invested capital dipped as the goodwill entered the capital base before the margin lift arrived.

What would change the read

The read here is measured, not adverse: CGI paid full prices for lower-margin consulting firms, and the deals earn their cost of capital only if it lifts their margins toward its own and cross-sells managed services into them. That is exactly what management is executing, with early evidence it is working — so the question is delivery, not concept.

Three checks over the next several quarters are falsifiable. First, the segment margins where the deals sit — U.S. Commercial and State Government (Daugherty), U.K. and Australia (BJSS), and Western and Southern Europe (Apside) — should trend toward the group's mid-teens; stalling there is the warning sign. Second, the pro-forma net earnings on these three, C$46 million, should visibly rise as a normalized run-rate in FY2026 disclosure. Third, the first goodwill impairment in company history, on a base that now exceeds equity, would signal that a specific bet — most likely BJSS, which alone carries C$1.14 billion of goodwill — has not delivered.

---

AI and the Labor Model

CGI sells roughly 94,000 people's time, so the question generative AI poses is direct: does it hollow out the revenue or lift the margin? The evidence points to margin. A managed-services majority priced on outcomes — where CGI has kept its productivity gains for decades — turns AI into a cost lever rather than a revenue hole, and headcount has stayed flat while revenue grew. The catch is that AI does nothing for the organic-growth stall, and the one lever that could turn it into growth — CGI's IP-revenue ambition — is moving backwards.

What AI puts at risk

CGI is a labor business. It employs approximately 94,000 consultants and professionals [1], and its revenue is, at root, billed expertise. The company says so itself in its risk disclosures: developments in AI — "including agentic AI, generative AI, as well as automation and machine learning" — could pressure "our revenue, net earnings and resulting cash flow from operations" if it fails to adapt [2]. AI also sits on the standing list of competitive risks in every recent filing [3].

The exposure is not evenly spread. About 45% of revenue is business and strategic IT consulting and systems integration (SI&C), and 55% is managed IT and business process services [4]. The SI&C half is the part a skeptic worries about first: consulting is largely time-and-material, and systems integration is the kind of implementation work AI is meant to automate. Management concedes the market shares that fear — one analyst noted that "the market apparently thinks there's going to be an impact from AI… automating a lot of the implementation processes that you're exposed to" [5].

The industry backdrop makes the concern concrete. The offshore-heavy players built on billable headcount are already shrinking: Infosys cut its workforce by about 8,000 sequentially in its most recent quarter [6], and Cognizant carries roughly 336,800 employees, the overwhelming majority in India [7]. For a pure labor-arbitrage model, fewer billable hours is fewer dollars. The question for CGI is whether its own model works the same way.

The buffer: outcome-based pricing

It mostly does not, and the reason is how CGI prices. Its default managed-services model is outcome-based — "we commit to cost predictability and delivering results, not just inputs" — and those engagements have, "for decades, included commitments to deliver ongoing productivity improvements" [8]. CGI has been here before: it absorbed the shift from onshore to offshore, then to cloud, each time promising the client savings and keeping a share of the efficiency as margin. Advanced AI, in management's framing, is the next lever in that same mechanism — one that "has contributed to improving profit and reinvesting in capability building" [9].

When work is priced on outcomes, a productivity gain does not shrink the invoice; it widens the spread. CFO François Boulanger put the arithmetic plainly: with AI in delivery, "I don't need necessarily the same head count number or same number of people to deliver the services," so revenue per employee should keep rising; and on fixed-price work, "every way of reducing the cost would go directly in our margin improvement" [10]. Even the SI&C side is more insulated than the "time-and-material" label suggests: management estimates 40% to 50% of the SI&C book is fixed price, where AI tools "increase the profit or the margin on our projects while hitting the right price tag" [11].

The early numbers fit the story. Over FY2023 to FY2025, revenue rose about 11% while headcount barely moved — and the FY2025 increase to 94,000 was itself acquisition-driven, so organic headcount is flatter still.

Source: revenue from reported financials FY2023–FY2025; headcount from annual reports (approximately 91,500, 90,250 and 94,000 consultants and professionals) [12]. Indexed to FY2023 = 100; ratios are unitless.

That gap is revenue per employee, and it is widening — up roughly 4% a year, to about C$169,000 in FY2025. Management pegged the FY2025 lift near 5% and credited AI alongside offshore delivery [13]. The honest qualifier: that rise blends three forces, not one — AI in delivery, but also the acquisitions and a growing India-and-Poland offshore mix that Boulanger named in the same breath [14]. AI is a contributor to the decoupling of revenue from bodies, not yet a proven driver of it.

CGI is also putting operational substance behind the claim rather than only slideware. Its DigiOps suite runs in production for many clients with over 165 AI agents and 2,000 automation workflows, delivering up to 30% productivity gains and up to 40% faster resolution of IT requests; in software development, AI code generation — about a quarter of the development life cycle — is showing efficiency gains near 30% [15]. And CGI is buying the capability as well as building it: its FY2025 acquisition of Daugherty is described as a firm "specializing in artificial intelligence, data analytics, strategic IT consulting" [16].

The catch: deflation, and a stalling IP bet

The same mechanism that defends margin has a second edge. If AI lets CGI deliver a given outcome with fewer people, then on outcome and fixed-price contracts the value delivered — and eventually the contract's price at renewal — can compress even as margins hold. Boulanger himself expects the input-based billing model to "continue to reduce" in favour of fixed and outcome-based work [17]. That is good for margin and neutral-to-negative for the top line — which is precisely the organic-growth problem the report opened on (Build and Buy). AI helps CGI earn more on flat volume; it does not obviously help CGI sell more.

The intended answer to that is IP. CGI's stated "IP30" ambition is to earn 30% of revenue from its own intellectual property — software and platforms sold repeatedly, where AI would expand the model rather than deflate it, and where each sale is not a fresh block of billed hours [18]. It is the part of the strategy where the AI story turns from defense to offense. The problem is that the number is going the wrong way.

Source: quarterly earnings-call remarks, IP as a percentage of total revenue [19][20][21]; target from May 2026 Corporate Overview [22].

IP was 22.6% of revenue in late FY2023 and 22.9% a year later; through FY2025 it fell to 21.6%, then 21.5%, then 20.6% [23][24][25][26][27]. Nearly ten points below the target, and drifting away from it. Management's own explanation ties the two halves of the strategy together: the decline is due to "the dilutive impact of recent business acquisitions" — the labor-and-services-heavy consultancies like BJSS, Daugherty and Apside carry little IP, so buying them mechanically shrinks the IP share [28]. The Buy engine that sustains reported growth (Acquisition Math) is, on this axis, pulling CGI away from the IP-led model it says AI will reward.

What would change the read

Two markers would move this from defensive to genuinely offensive. First, the IP share breaking back above the low-20s and trending toward 30% — evidence that CGI can build and sell AI-era software faster than its acquisitions dilute the mix. Second, a sustained pickup in organic growth (stripping out both currency and acquired revenue) alongside flat or falling headcount — the signature of AI expanding the business rather than merely making the existing book cheaper to run. Absent those, the fair read is that AI protects CGI's economics without solving its growth problem, and CGI has not yet published data proving implementation work resists automation as well as it argues it does [29].

---

Order Book

CGI's order book is where the durability of its compounding is tested. Contracted backlog stands at C$31.5 billion — about two years of revenue — and has grown every year since FY2021, with a book-to-bill ratio above 100% throughout [1]. The durability is narrower than the headline. A recurring managed-services base books at 117–122% of revenue, while consulting sits at or below 100% [2]. The main vulnerability is concentration: government is 38% of revenue, the U.S. federal government alone 14.1% [3].

The size and shape of the order book

Backlog is the single most useful number for judging whether CGI's revenue base is durable. At C$31.5 billion against FY2025 revenue of C$15.9 billion, the order book is roughly two times annual sales, and it has risen from C$23.1 billion in FY2021 — a compound rate near 8% a year, slightly ahead of revenue [4] [5].

Source: FY2025 Annual Report, MD&A 2.1 Selected Yearly Information & Key Performance Measures [6].

Sources: FY2025 Annual Report, MD&A 2.1 (FY2023–FY2025) [7]; FY2023 Annual Report, MD&A 2.1 (FY2021–FY2022) [8].

The order book is also long-dated, which matters more than the size. CGI discloses when its backlog is expected to convert into revenue: about C$12.0 billion within twelve months, C$10.5 billion in one to three years, C$4.6 billion in three to five years, and C$4.4 billion beyond five years [9]. Roughly 62% of the book is contracted to convert after the next twelve months, and C$9.0 billion — close to 30% — sits beyond three years. That is a revenue base with visibility few consulting-led peers can match.

Source: FY2025 Annual Report, MD&A 2.1 Selected Yearly Information & Key Performance Measures [10].

Two caveats belong next to that visibility. First, CGI's own definition of backlog includes commitments "acquired through business acquisitions" and management estimates of demand-driven usage, and it is reduced when clients cancel or scale back — so it is an estimate of contracted revenue, not a firm receivable [11]. Second, part of the backlog's growth was bought rather than won: acquired backlog enters the balance directly and does not pass through the bookings figure, as management confirmed on the FY2025 year-end call [12]. The consequence is useful: because bookings exclude acquired backlog, the book-to-bill ratio is a cleaner organic signal than the backlog balance, which — like reported revenue (Build and Buy) — is flattered by acquisitions.

Where the durability comes from

The book-to-bill ratio held above 100% in each of the last three years — 113.7%, 109.3% and 110.4% — meaning new bookings outpaced revenue burned, so the order book kept filling faster than it emptied [13]. That headline conceals two very different businesses underneath it.

Split by service line, the recurring half carries the order book. On a trailing-twelve-month basis, managed IT and business-process services booked at 117% of revenue in FY2024, 120% in FY2025 and 122% by the first quarter of FY2026. Consulting and systems integration booked at 100%, then 99%, then 96% over the same span [14] [15] [16]. In plain terms, the managed-services base is adding net new recurring revenue each year, while consulting is barely replacing what it delivers.

Source: Q4 FY2024, Q4 FY2025 and Q1 FY2026 earnings-call transcripts; 100% marks bookings equal to revenue [17] [18] [19].

This is the location of CGI's moat, and it is a specific one. Managed services is roughly 55% of revenue, and these are multi-year outsourcing engagements — running a client's systems, applications or business processes — with real switching costs: the incumbent holds the operational knowledge, the transition risk of moving is high, and CGI's local-delivery "proximity" model keeps decision-makers close to the client [20]. The retention shows in the composition of bookings: roughly 60% of the year's new bookings were extensions, renewals and add-ons rather than new business, with new business the remaining 40% [21]. A book weighted toward renewals is a book that clients keep choosing to re-sign.

That switching-cost base is also what makes the outcome-based pricing described in AI and the Labor Model credible: a client locked into a multi-year managed-services contract is buying an outcome, not hours, which is why CGI can keep AI-driven productivity gains as margin rather than passing them back as lower billings. The consulting and systems-integration half, by contrast, has neither the switching costs nor the recurring revenue — its sub-100% book-to-bill is the discretionary, cyclical part of the story, and the softest link in the order book.

A pattern the sector leader shares

The shape of CGI's order book is not unique; it mirrors the industry's largest player, which is a useful check that the split is structural rather than a CGI quirk. In its most recent quarter Accenture booked new work at an overall 1.2 times revenue, with consulting at 1.0 and managed services at 1.4 [22]. The same division holds: managed services carries the order book, consulting merely holds its own.

Sources: CGI trailing-twelve-month figures, Q4 FY2025 call [23]; Accenture Q4 FY2025 call [24]. Accenture's figures are single-quarter bookings; CGI's are trailing twelve months, so levels are indicative rather than exactly comparable.

The comparison also frames a limit. CGI's total book-to-bill of 110% trails Accenture's 120%, and its managed-services 120% trails Accenture's 140%. CGI's order book is durable, but it compounds more slowly than the leader's — consistent with the organic-growth stall that opened this report. The moat protects the base; it does not, on this evidence, restore the pace of growth.

Where the order book is exposed

The clearest risk in the order book is concentration. Government is CGI's largest end-market at 38% of revenue, and under the accounting definition of a single customer, the U.S. federal government and its agencies alone were 14.1% of FY2025 revenue — up from 13.6% a year earlier [25]. Government work cuts both ways: it is sticky and defensive in a downturn, but it is exposed to budget cycles, procurement freezes and shutdowns in a way commercial work is not.

Source: FY2025 Annual Report, MD&A 3.3 Revenue Distribution [26].

That exposure is not theoretical. CGI's own risk disclosure names "government shutdowns or the threat of government shutdowns," curtailed use of consulting firms, and budget cuts as factors that could cause agencies to reduce, terminate or decline to renew contracts [27]. The U.S. federal segment is also the lumpiest part of the book: its full-year FY2025 book-to-bill was 92.5%, below replacement, even though the fourth quarter spiked to 185% on contract timing [28]. By the first quarter of FY2026 management confirmed U.S. operations had been hit by the federal shutdown, with a sequential recovery expected but the segment still "operating in a very dynamic environment" [29].

A second, quieter erosion sits inside the sticky base itself. Many managed-services contracts carry benchmarking provisions — clauses that let a client force pricing back toward market rates during the contract term. CGI flags that a downturn, by pushing competitors to cut rates, "may trigger pricing adjustments related to the benchmarking obligations within our contracts" [30]. Switching costs keep the client; they do not fully protect the price. A backlog that renews at lower margins is durable in revenue but not in profit.

What the order book says

On the evidence, CGI's moat is real but narrow. It rests on a recurring managed-services base that books at 117–122% of revenue, renews rather than churns, and gives two years of contracted visibility — a genuine switching-cost franchise, and the reason the outcome-based pricing model holds. It does not extend to the consulting and systems-integration half, whose sub-100% book-to-bill is the discretionary drag on the whole, nor does it neutralize a government concentration — 38% of revenue, 14.1% in one federal client — that is exposed to shutdowns and benchmarking repricing.

The read that fits the numbers: the order book is durable enough to defend the base but not to restore the pace of compounding, which keeps the burden on acquisitions (Acquisition Math) to supply growth. The single figure that would change it is the managed-services book-to-bill: as long as it holds above roughly 110%, the recurring engine is still net-adding; a sustained fall toward 100%, or a step-down in U.S. federal work, would signal the moat is eroding faster than consulting can recover.

---

The De-Rating

CGI's shares have fallen about a third in a year, from roughly C$132 to C$94.85, and now trade near 10 times forward earnings against a five-year average close to 20. The instinct is to read this as a company-specific discount. The evidence points the other way: the multiple compressed while earnings kept rising, and it compressed alongside the entire labor-based IT-services sector, not below it. On forward earnings CGI is priced at a modest premium to Accenture, Capgemini and Cognizant — not a bargain within its group, but a quality business repriced with the pack.

What the multiple did, and what earnings did

The de-rating is almost entirely a change in the multiple, not the earnings. Reported diluted EPS rose from C$6.86 in FY2023 to C$7.35 in FY2025, and adjusted diluted EPS from C$7.07 to C$8.30 [1]. The most recent quarter continued the pattern: in Q2 FY2026, revenue grew 3.3% (1.6% in constant currency) yet diluted EPS climbed 10.6% to C$2.09, carried by margins and buybacks rather than volume [2]. Over the same window the share price nearly halved its valuation. The earnings engine slowed but did not stall; the price did the damage.

Reported and adjusted diluted EPS, FY2023–FY2025, per the FY2025 Annual Report [3]. Consensus pencils in C$8.90 for FY2026 and C$9.69 for FY2027, on an adjusted-type basis.

The earnings trajectory runs against a falling valuation. At C$94.85, CGI trades at 12.4 times trailing earnings and about 10 times forward — 10.7 times the C$8.90 the Street pencils in for FY2026 and 9.8 times the C$9.69 for FY2027 — against a five-year average price/earnings ratio near 20. Earnings-per-share have risen every year of that span; the multiple is what moved.

Market data as of July 10, 2026; FCF yield derived from FY2025 free cash flow of ~C$2.1 billion against market capitalization of ~C$19.9 billion [4].

The price slide was a sector event

The fall was not gradual disappointment with CGI. It tracked a repricing of the whole labor-based IT-services complex through 2026, as the market began to doubt whether firms that sell people-hours can hold revenue and margin against generative AI. Accenture cut its outlook and its shares fell to the mid-US$130s, less than half their 2024 peak; Capgemini dropped sharply on the same fear; Infosys reached a five-year low and Tata Consultancy Services a multi-year low. CGI's chart is a member of that cohort, not an outlier within it.

Market data, Toronto Stock Exchange closing prices, January–July 2026.

That context changes how the valuation should be read. CGI did not de-rate because the market singled it out; it de-rated because the market marked down the business model it shares with its peers. Whether that mark-down is an opportunity depends on how CGI is priced relative to that peer group — and there, the discount narrative breaks down.

Not the cheap one in the room

On forward earnings, CGI is the most expensive of the three Western labor-based peers it most resembles. It trades at 10.1 times forward earnings against Accenture's 9.4, Capgemini's 6.8 and Cognizant's roughly 9.1; on EV/EBITDA the gap is wider, 7.7 times versus 6.3 for both Accenture and Cognizant. CGI screens cheap only against IBM, which has re-rated toward a software-and-AI story near 22 times, and against the Indian majors — Infosys near 14 times, TCS near 13 — which carry a structural growth-and-return-of-capital premium.

Forward P/E from market data as of July 10, 2026. Peer set is CGI's closest labor-based IT-services competitors — Accenture, Capgemini, Cognizant, Infosys, TCS and IBM.

Market data as of July 10, 2026; blanks where a clean current figure was not available. Peer group is the analyst's labor-based IT-services comparison set.

The premium to Accenture, Capgemini and Cognizant is small, and there is a defensible case for it. CGI runs a higher share of recurring managed services — roughly 55% of revenue — and books that work at 117–122% of revenue, a mix the market appears to treat as more defensible against AI than discretionary consulting (Order Book). Its adjusted EBIT margin held at 16.4% in FY2025, and it converts well over 100% of earnings to free cash [5]. What the premium is not is a margin of safety relative to the peer group. An investor buying CGI here is paying for quality and durability at a small markup to cheaper, faster-shrinking peers — not buying the sector's discount name.

What ten times earnings is paying for

Reduced to what it implies, the multiple is undemanding. At C$94.85 on FY2026 consensus of C$8.90, the forward earnings yield is about 9.4%, and free-cash-flow yield sits near 11% on FY2025 output of roughly C$2.1 billion. Consensus has EPS reaching C$9.69 in FY2027 — high-single-digit growth. A high-single-digit compounder at roughly a 10% earnings yield is not demanding; the price embeds low expectations for growth and a real fear that AI erodes the labor model.

Two facts complicate the bull reading. First, the earnings quality has softened where it matters to a valuation: return on invested capital fell to 13.6% in FY2025 from 16.0% in each of the two prior years, as acquisitions were digested [6]. A lower return on capital deserves a lower multiple, all else equal. Second, capital allocation has not been vindicated by the tape: during FY2025 CGI repurchased C$1,258.5 million of stock at a weighted-average price of C$142.01 — about 50% above today's price — while paying a dividend that yields just 0.7% [7]. The same buyback program, run at C$95, is far more accretive than it was at C$142; the balance sheet, at net debt near C$3.5 billion, has the room [8].

What would change the read

The evidence supports a measured view: CGI is reasonably valued rather than obviously cheap. The de-rating restored a sensible price for a high-quality, slower-growing compounder, but it did not open a wide discount to comparable businesses. The strongest fact against a constructive read is the return-on-capital slippage to 13.6% paired with constant-currency growth of only 1.6% in the most recent quarter — a business that is both growing more slowly and earning less on incremental capital than it was two years ago [9][10].

Three things would decide which way the multiple travels from here. Constant-currency organic growth is the first: a durable move back toward mid-single digits would argue the AI fear is overdone and support a re-rating; a slide toward flat would validate the discount. ROIC is the second: a recovery toward 16% as the FY2025 acquisitions season would confirm the deals earned their cost of capital, while a further drift would mark the buy half of Build-and-Buy as multiple-dilutive. Capital allocation is the third: sustained repurchases at today's price, rather than at C$142, are the clearest signal management sees the shares as undervalued — and the most direct route to per-share compounding while the multiple stays low [11].

---

Earnings Quality

CGI's free cash flow has exceeded net earnings every year for a decade — C$2.12 billion in FY2025, roughly 128% conversion on capital spending under 1% of revenue — so the free-cash-flow yield behind the de-rated valuation (The De-Rating) rests on cash that is genuinely there. Two things soften the picture: quarterly conversion swings on collection timing, and a widening gap between reported and "adjusted" earnings built from add-backs that recur every year.

Cash that converts

The strongest fact in CGI's favour is boring and durable: the business turns accounting profit into cash, and then some. In FY2025 it generated C$2,234.2 million of operating cash flow and C$2,117.6 million of free cash flow after C$116.6 million of capital spending [1]. Free cash flow ran to about 128% of net earnings — and it has cleared 100% in each of the last ten years. This is what an asset-light services model looks like when it works: capital spending is under 1% of revenue, so almost all operating cash is free.

Source: FY2025 Annual Report, Selected Yearly Information [2]; free-cash-flow conversion derived from reported cash-flow and income statements.

Source: consolidated cash-flow and income statements, FY2021–FY2025; FY2023–FY2025 figures per FY2025 Annual Report [3].

One detail confirms the quality rather than assumes it. CGI runs payroll and payment services, which means client money flows across its balance sheet — the kind of float a company can quietly lean on to flatter operating cash. CGI does not: the C$468.6 million FY2025 movement in clients' funds obligations sits in financing activities, not operating [4]. The operating-cash line is therefore a clean read on the business, not a working-capital construction.

Timing, not trend

The caution begins one level down. CGI's headline cash-conversion rate — operating cash as a percentage of revenue — is heavily driven by the timing of collections, and it swings hard from quarter to quarter. Over the last ten quarters it has ranged from 10.9% to 21.4% of revenue, with days-sales-outstanding moving between 37 and 45 days.

Source: CGI quarterly press-release financial highlights, Q1 FY2024–Q2 FY2026; anchor prints Q1 FY2025 [5] and Q1 FY2026 [6].

The pattern is seasonal, and management says so plainly. First quarters print the strongest cash — the 21.4% in Q1 FY2026 came "due to the strength of our collection efforts," and, as the company reminds investors, "our first quarter has the lowest DSO due mainly to higher levels of client prepayments or annual IT maintenance fees" [7]. A year earlier, asked whether the 17.1% Q1 FY2025 print signalled a higher full-year rate, CFO François Boulanger declined the invitation: "on a long-term basis, 15% makes sense. Obviously, in the quarter, the cash from ops was 17%, but it was with some improvement of the DSO" [8]. A single strong quarter is not a run-rate, and CGI is candid about that.

The same timing works in reverse across full years. Operating cash as a share of revenue slipped from 15.0% in FY2024 to 14.0% in FY2025 [9], which reads like deterioration until the working-capital line is separated out. Cash flow before working-capital changes actually rose C$213.3 million; the decline came from receivables, work-in-progress and deferred revenue swinging from a C$147.8 million tailwind in FY2024 to near-neutral, while payables and provisions turned into a C$105.2 million use of cash [10]. Management's own summary: "the timing of our working capital inflows and outflows will always have an impact on the cash flow from operations" [11]. Year-end DSO did drift up — 44 days in FY2023, 41 in FY2024, 45 in FY2025 [12] — partly the receivables that come with newly acquired businesses (Acquisition Math). It is worth watching, but a four-day move inside a business that still converts at 128% is a wrinkle, not a warning.

Source: CGI quarterly press-release financial highlights, Q1 FY2024–Q2 FY2026 [13][14].

The adjusted-earnings wedge

Where the cash line is honest, the profit line is dressed. CGI reports both IFRS earnings and an "adjusted" measure that strips out restructuring, acquisition and related integration costs, and the gap between the two has widened sharply. Reported diluted EPS was essentially flat in FY2025 — C$7.35 against C$7.31 — while adjusted diluted EPS rose 8.9%, from C$7.62 to C$8.30 [15]. The wedge between them grew from about 3% of reported EPS in FY2023 to nearly 13% in FY2025.

Source: FY2025 Annual Report, Selected Yearly Information [16].

The engine of that wedge is a C$285.0 million pre-tax add-back in FY2025 — C$196.8 million of restructuring (largely the European cost realignment) and C$88.2 million of acquisition and integration costs [17]. Excluded, they turn a 2.1% decline in earnings before tax into 5.8% growth, and a 2% fall in net earnings into a 6% rise [18]. The whole of FY2025's earnings "growth," on the adjusted view, lives in the items added back.

The question a skeptic raises is whether these are one-time costs or a permanent feature of the model. The record points to the latter. Under rotating labels — acquisition and integration costs, a "Cost Optimization Program," now "Restructuring" — CGI has excluded specific items every single year, and the amounts are escalating: about C$62 million in FY2023, C$97 million in FY2024, C$285 million in FY2025 [19][20]. For a company whose strategy is to buy and integrate consultancies every year (Build and Buy), acquisition and integration cost is not an interruption of the business — it is the business. Calling it "specific" understates the ongoing cost of the compounding machine.

Source: FY2024 Annual Report, Excluding Specific Items [21]; FY2025 Annual Report, Adjusted Net Earnings [22].

A fair counter runs the other way on the reported number. FY2024 net earnings carried a C$146.1 million deferred income-tax recovery that did not repeat in FY2025 [23], so FY2025's flat reported EPS is measured against a tax-flattered base and understates operating progress. At the operating level, adjusted EBIT margin held at 16.4% and adjusted EBIT itself grew 8% [24]. Both statements are true at once: the underlying business grew, and the adjusted headline flatters that growth by treating a recurring cost as exceptional.

The read

The two halves of this chapter point in opposite directions, and the reconciliation is the useful part. An investor who anchors on cash is on firm ground: the restructuring, acquisition and integration charges that inflate adjusted EPS are real cash costs, and they are already inside the free-cash-flow figure — which still converts above 100%. The cash-based valuation lens from The De-Rating is not fooled by the adjusted-earnings optics; it captures exactly the costs that optics remove. An investor who anchors on the adjusted-EPS growth rate, and pays a multiple on it, is on softer ground: a meaningful share of that "growth" is the exclusion of costs the model incurs every year.

The evidence points to earnings that are cash-backed but flattered at the reported-profit line — high quality where it counts for a valuation built on free-cash-flow yield, lower quality where it counts for anyone extrapolating adjusted EPS. The main fact against the benign read is the escalating, ever-present add-back; the fact against the bearish read is that free cash flow, which already bears those costs, keeps converting. What would change the assessment: full-year free-cash-flow conversion falling below 100% for two consecutive years, a structural ratchet in DSO as acquired receivables fail to collect, or the annual "specific items" add-back continuing to climb faster than revenue. Those three lines — conversion, DSO, and the size of the add-back — are the ones to watch.

---

Funding the Buy

For a decade, CGI paid for its acquisitions and its buybacks out of its own cash flow. In fiscal 2025 it did not. Acquisitions, share repurchases and a first dividend came to C$3.24 billion against C$2.12 billion of free cash flow, and the roughly C$1.1 billion gap was bridged with a C$923.9 million note issue and a drawdown of cash [1]. Net debt jumped to C$3.45 billion and net-debt-to-capitalization to 25.1% [2]. Leverage is still modest; the self-funding pillar of the thesis is what bent.

Source: net debt and net-debt-to-capitalization, FY2025 Annual Report, MD&A Section 4.5 [3]; net-debt-to-EBITDA per company disclosure [4]; coverage derived from adjusted EBIT and net finance costs [5].

The year the cash stopped covering it

The engine described in Build and Buy rests on a simple rhythm: operating cash funds the deals, the buybacks and the debt paydown, all at once, with room to spare. That rhythm held through fiscal 2024. It did not hold in fiscal 2025.

Cash generation itself was fine. Operating cash flow was C$2.23 billion and free cash flow C$2.12 billion, in line with the prior year and consistent with the conversion record set out in Earnings Quality [6]. What changed was the size of the claims on that cash. CGI spent C$1.83 billion on acquisitions net of cash acquired, C$1.27 billion buying back and cancelling Class A shares, and C$135 million on its newly initiated dividend [7]. Those three uses total C$3.24 billion — about C$1.1 billion more than the year produced.

Source: FY2025 Annual Report, Consolidated Statement of Cash Flows [8].

The gap was closed two ways. In March 2025 CGI issued US$650 million of five-year senior notes at 4.95%, swapped back to a 3.71% Canadian-dollar cost [9], raising the C$923.9 million shown as an increase in long-term debt [10]. And the cash balance was allowed to fall, from C$1.46 billion to C$864 million over the year [11]. Neither is remarkable on its own. Together they mark the first year in this cycle where CGI could not do everything it wanted with the cash the business threw off — the "self-funded acquisitions" language in the through-line met its first real test.

How much of a stretch this is

Not much, on the numbers that matter for solvency. Net-debt-to-capitalization of 25.1% looks like a jump only against fiscal 2024's 16.2%, which was itself a trough after years of deleveraging. Set against a longer run, 25.1% sits below where CGI operated as recently as fiscal 2022, when the ratio was 28.8% [12].

Source: FY2025 net debt to capitalization of 25.1%, FY2025 Annual Report, MD&A Section 4.5 [13]; net debt to capitalization of 28.8% in FY2022 and 20.4% in FY2023, FY2023 Annual Report, MD&A Section 4.5 [14].

The cleaner leverage lens tells the same story. On the company's own measure, net debt closed fiscal 2025 at roughly 1.0 times trailing EBITDA, up from a 0.5-times low earlier in the year but well inside CGI's historical band — the ratio ran near 1.2 times through fiscal 2022, reached about 1.5 times during 2020, and touched 2.7 times after the 2012 Logica deal, deleveraging quickly each time [15]. Interest is barely felt: net finance costs were C$83.7 million against adjusted EBIT of C$2.61 billion, coverage of roughly 31 times, even after interest on long-term debt rose to C$80.9 million from C$48.0 million on the new notes [16] [17]. And liquidity is ample: C$2.39 billion of capital resources available at year-end, most of it an undrawn C$1.5 billion revolving facility extended to 2030, with all debt covenants in compliance [18].

Against C$2 billion-plus of annual free cash flow, a net debt of C$3.45 billion is something CGI could clear in under two years if it chose to. The balance sheet did not weaken. It lost its slack.

The choices that slack used to hide

Losing slack matters because it forces a ranking. Through fiscal 2024, free cash flow paid for acquisitions, buybacks and debt reduction together; nothing had to be traded off. In fiscal 2025 those uses began to compete, and the calendar sharpens the point. C$835.9 million of 2021 U.S. senior notes matures in September 2026 — a maturity large enough that it pushed year-end working capital slightly negative, to minus C$49.0 million [19]. The maturity wall beyond it is manageable but no longer trivial.

Source: Q2 FY2026 Investor Presentation, Capital Structure [20].

The reasonable read is that CGI now has to choose among deleveraging, buying back stock and funding the next large acquisition, where before it could do all three. That is a milder problem than it sounds: at roughly one times EBITDA, 31 times interest coverage and C$2.4 billion of liquidity, the ranking is a decision made from strength, not a constraint imposed by lenders. Management frames the same tension as a "balanced approach towards share buybacks and reducing leverage," which is candid about the trade-off now in front of it [21].

The counter-case to any worry here is the coverage and the deleveraging history; the case for watching it is that the fiscal 2025 buyback ran at an average of about C$142 a share, above today's price (see The De-Rating), and was part-funded by borrowing rather than cash. What would change the read is a larger debt-funded acquisition that carried leverage toward the 2-to-2.7-times range while organic growth stays in the low single digits — the combination that would turn a balance-sheet choice into a balance-sheet limit. Until then, the funding shift is best read as CGI spending its accumulated slack to keep the Build-and-Buy machine and the buyback running in the same year, not as a company reaching for leverage it cannot carry.

---

A controlled company

CGI is not a widely-held company that happens to have a large shareholder. It is controlled. Serge Godin, who founded the firm in 1976, holds 55.6% of the votes on 11.25% of the equity through Class B shares that carry ten votes each to the Class A share's one [1] [2]. Whoever runs CGI, and however the shares trade, the Godin family decides the outcome of any shareholder vote. That fact conditions everything else in this report: the patient acquisition machine, the buyback pace, the willingness to add leverage in FY2025 — all of it sits downstream of an owner who cannot be outvoted and does not have to answer a quarterly market.

Source: 2025 Management Proxy Circular, control table and share-class voting rights [3] [4]; public-float figures derived as the residual.

The 24,122,758 Class B shares represent 100% of that class and are held entirely by Godin [5]. The structure reaches beyond the equity: the change-of-control clause in CGI's senior notes defines a "Permitted Holder" as Serge Godin, his spouse, children and lineal descendants, and family trusts — so a handover of control within the family does not trip the noteholders' repurchase right, while a transfer of control outside it would [6]. Control is engineered to pass down a bloodline, not to change hands.

Class A holders are not without protection. Each Class A share is a "restricted security," and CGI's constating documents grant subordinate holders coattail rights in the event of a take-over bid for the multiple-voting shares [7]. What they do not have is the votes to direct the company.

What the control has underwritten

For a professional investor, dual-class control is not automatically a mark against a company; it is a variable whose sign depends on what the controller does with it. In CGI's case, the record it has bought is the compounding engine the rest of this report examines: the self-funded Acquisition Math, the decade of free cash flow above net earnings (Earnings Quality), and buybacks run through the cycle. Serge Godin led CGI as CEO from 1976 to 2006 and, as Executive Chairman thereafter, presided over the "Build and Buy" strategy that grew the firm from two consultants to 94,000 [8]. The long-termism the thesis depends on and the concentration of control are, historically, the same fact.

The alignment is real in the one way that matters most: the family's roughly 24.4 million shares are the same instrument public holders own, exposed to the same per-share value — a stake worth on the order of C$2.3 billion at the current price (The De-Rating) [9]. An owner with that much of their wealth in the stock has a direct interest in the per-share compounding the thesis is built on, not merely in the size of the enterprise.

The handover now in motion

The reason governance is a live question now, rather than a standing footnote, is that the people who built the machine are handing it over — and in a compressed window. CGI has changed chief executives twice in nineteen months, and reorganised the chair at the same time.

Sources: 2025 Management Proxy Circular, director profiles [10] [11] [12]; CGI leadership page for the May 2026 appointment [13].

George Schindler, the architect of Build-and-Buy, stepped down as CEO in October 2024 and remains on the board; François Boulanger, a 25-year insider who had been CFO and then COO, succeeded him [14]. Nineteen months later, in May 2026, Tim Hurlebaus became President and CEO — the third CEO in the span, again promoted from within [15]. The Q2 FY2026 call, on which management would ordinarily frame such a change, is the one recent transcript this corpus could not recover, so the primary-source detail on the rationale is thinner than the rest of this report; the appointment itself is confirmed on CGI's own leadership disclosure.

The chair change is the more telling structural move. Effective January 29, 2025, Serge Godin, age 76, moved from Executive Chairman to Founder and Co-Chair, with a defined remit to "oversee transformational acquisitions and large-scale engagements" — narrowing his role to the part of the strategy he cares most about [16] [17]. His daughter Julie Godin, who joined in 2009 and had run HR, strategic planning and M&A, became Executive Chair, charged with setting the company's strategic direction and its rolling three-year plan [18]. The shape of the succession is now visible: a professional, long-tenured CEO runs operations; the founder retains the acquisition file and control; and a second-generation family member holds the strategic chair. It is a stewardship handover, not a break — every incoming principal is an insider, which argues for continuity of the capital-allocation discipline the thesis rests on.

Guardrails and flags

The board around the family is genuinely independent and unusually deep in finance. Nine of the thirteen directors (69%) are independent [19], the chair roles being non-independent are balanced by an independent Lead Director in George Cope, former CEO of BCE [20], and the Audit and Risk Management Committee is stocked with former CFOs and a former central-bank governor — Kathy Waller (ex-Coca-Cola CFO) as chair, Alison Reed (ex-Marks & Spencer and Standard Life CFO), Frank Witter (ex-Volkswagen CFO), and Stephen Poloz, former Governor of the Bank of Canada [21]. This is not a captive board.

Two flags sit against that. The first is pay. In FY2025 the highest-paid named executive was not the CEO but Serge Godin, in the Co-Chair role, at C$12.08 million against Boulanger's C$8.49 million — the bulk of it a share-based award of roughly C$10.7 million that has been near-identical for three straight years [22]. He forwent his short-term incentive for the year, which narrowed the gap, but a controlling founder in a chair role out-earning the sitting CEO is a structure that rewards control as much as execution [23].

Source: 2025 Management Proxy Circular, Summary Compensation Table [24].

The second flag is the treatment of minority holders' voice. A shareholder placed a proposal on the 2025 ballot asking CGI to disclose voting results by share class and to open a structured dialogue on executive pay in the multiple-voting-share context. The board recommended a vote against it, noting that CGI holds no say-on-pay advisory vote at all and arguing that direct engagement — including a Shareholder Satisfaction Assessment Program that scored 9 out of 10 in FY2025 among Class A holders — serves the purpose better [25]. The proposal itself is the more useful signal: some minority holders want more transparency than the structure gives them, and the controller is not obliged to provide it.

One quieter dynamic runs beneath the buyback. Repurchases only retire Class A shares — the public float, which fell to 192.2 million by December 2025 from more than 202 million a year earlier — which, on its own, concentrates the family's voting share [26]. Working the other way, a Godin holding company converted 1,422,948 Class B shares into Class A in FY2024, surrendering votes [27]. The net of the two forces has been mild dilution of family voting power — from 56.2% of votes in late 2023 to 55.55% in late 2025 [28]. Control is not tightening; it is holding comfortably above the majority line.

The read

The evidence points to founder control being, on balance, a support for the thesis rather than a threat to it — but a support whose durability is now being tested in real time. The case for it: the same concentrated, patient ownership produced the disciplined self-funding, the through-cycle buybacks and the high-return compounding the report documents, and the succession is being executed through insiders with the family retaining both control and its C$2.3 billion economic stake. The strongest fact against it is that minority holders have no binding lever and one just asked, publicly, for more voice and was declined — in a structure where the controlling founder already out-earns the CEO. What would change the read is behaviour after the handover: the FY2025 funding shift (Funding the Buy) showed the capital-allocation trade-offs getting harder, and whether Hurlebaus and Julie Godin hold the same line on price-disciplined buybacks and self-funded M&A — rather than reaching for growth or diluting the return test — is the thing to watch, and it will show up in the numbers before it shows up in any proxy.

---

Bull, Base, Bear

The eight chapters before this one each isolated one part of CGI: the acquisition machine, the AI question, the order book, the de-rating, earnings quality, the funding shift, the founder's control. This chapter puts them back together. At roughly 10 times forward earnings, the shares already carry a large share of the pessimism that repriced the whole IT-services sector in 2026. Over a two-year horizon the gap between a good and a bad outcome is driven mostly by whether the multiple recovers — the low starting point cushions the downside. A falsifiable watch-list follows.

Sources: share price C$94.85 at July 10, 2026 and consensus estimates (FY2026E EPS C$8.90, target C$126.31), as reported; earnings and FCF yields derived from consensus EPS and FY2025 free cash flow [1].

How the pieces fit

CGI is a business that compounds per-share earnings by pairing slow organic growth with self-funded acquisitions, and the report's through-line asks whether that engine still runs now that growth has stalled. The eight chapters resolve into a consistent picture rather than a contradictory one.

The engine still turns, more slowly. Reported revenue grew 8.4% in FY2025, but only 4.6% in constant currency and 1.6% in the most recent quarter — the growth is increasingly acquired and currency-aided, not organic [1][2]. Per-share earnings kept climbing anyway: adjusted diluted EPS rose to C$8.30 in FY2025 from C$7.07 in FY2023, and Q2 FY2026 diluted EPS grew 10.6% on 1.6% constant-currency revenue — the arithmetic of (Acquisition Math) and buybacks doing the work that organic demand no longer does [1][3].

The cash is real, the returns have dipped, the balance sheet took on leverage, and control sits with a family in the middle of a handover. Free cash flow again exceeded net earnings in FY2025 (C$2.12 billion, ~128% conversion), which is what makes the ~10% cash yield meaningful (Earnings Quality) [1]. But return on invested capital fell to 13.6% from 16.0%, net debt jumped to C$3.45 billion as the capital program outran cash for the first time in the cycle (Funding the Buy), and the Godin family still decides every vote through a leadership succession it is executing now (Control and Succession) [1]. None of these is fatal on its own; together they explain why a business that still grows earnings trades where a stalling one would.

Sources: reported and adjusted diluted EPS FY2021–FY2025 [1]; FY2026E/FY2027E are consensus estimates, as reported.

The gap between the two EPS lines is itself a finding the report has already made. The adjusted figure sits above the reported one because CGI adds back "specific items" — acquisition, integration and restructuring costs — every year under rotating labels; the wedge widened to about 13% in FY2025 on a C$285 million add-back, and how fast it narrows is one of the watch items below (Earnings Quality).

The scenarios

The scenarios below are illustrative arithmetic, not forecasts: each pairs an EPS path two years out with an exit multiple, starting from consensus FY2026E EPS of C$8.90 and today's C$94.85 share price. They are meant to bound the outcome, not predict it. The multiple does most of the work — which is the point of buying a business after a de-rating rather than before one.

Source: illustrative scenarios derived from consensus FY2026E EPS (C$8.90) and reported financials; exit multiples chosen against CGI's ~10x current and ~20x five-year-average forward P/E (The De-Rating).

Source: illustrative scenarios (above); current price C$94.85 at July 10, 2026, as reported.

The asymmetry is the substance here. The bear case requires two things at once — EPS growth slowing to about 2% a year and the multiple compressing further to 8.5 times, below where any large Western peer trades — and even then the price falls about 17%. The base case, roughly consensus, needs only mid-single-digit EPS growth and a partial re-rating to 12 times to reach about C$125, close to the C$126 sell-side target. The bull case does not need heroics either: high-single to low-double-digit EPS growth (organic recovery plus a large acquisition redeployed at today's depressed multiple, plus buybacks now struck below the C$142 average CGI paid in FY2025) and a re-rating to 15 times — still a quarter below the five-year average — roughly doubles the shares [4]. Dividends add perhaps another 0.7% a year to each path; they are immaterial to the picture.

That shape — limited downside from a compressed multiple, upside that depends on the multiple healing — is why the low starting valuation matters more than any single operating number. It is also the honest limit of the analysis: the bear scenario is not a tail. If AI genuinely erodes people-hour revenue, the 1.6% constant-currency growth already printed is the leading edge of it, not a trough.

Where bulls and bears actually disagree

The two cases do not argue over different facts; they read the same numbers in opposite directions. Each row below is a figure the report has established, with what would settle it.

Sources: FY2025 figures [1]; Q2 FY2026 constant-currency growth [2]; buyback average price [4]; peer valuation and five-year multiple (The De-Rating).

Two of these rows are more load-bearing than the others. The valuation debate is settled by the growth debate: at 10 times, the multiple re-rates only if organic growth convinces the market that AI is a margin tool for CGI rather than a revenue solvent — the exact question (AI and the Labor Model) leaves open. And the earnings-quality row conditions everything downstream: if the specific-items add-backs prove recurring rather than one-off, the "real" earnings base is nearer the reported C$7.35 than the adjusted C$8.30, and every multiple in the scenario table applies to a smaller number.

What to watch

The watch-list is the practical output of the report. Each item is a specific line in a specific filing, with a threshold that would move the read toward the bull or bear column. They are checkable at each quarterly result.

Sources: FY2025 metrics and net-debt/capitalization [1]; constant-currency growth and TTM book-to-bill [2]; buyback detail [4].

The single most informative pairing on this list is constant-currency growth against the specific-items add-back. The first tells the reader whether AI is a demand headwind or a passing air-pocket; the second tells them which earnings number the market is actually paying for. A quarter or two of both moving the right way would validate the base case without any re-rating heroics; both moving the wrong way is the bear scenario arriving.

The balance of evidence

On the evidence assembled across these chapters, CGI looks like a durable, cash-generative compounder priced for a growth stall that may prove cyclical — a business whose downside is cushioned by a multiple that has already compressed to roughly 10 times, and whose upside depends on the market coming to believe the engine still runs. The strongest fact against that read is on the same page as the case for it: constant-currency growth of 1.6% is not obviously a trough, and if AI is structurally deflating people-hour revenue, the low multiple is a fair price rather than a cheap one. What would change the read is not the valuation but the operating line — two or three quarters of organic growth and a shrinking add-back would move CGI from the bear column to the base one, and the watch-list above is where that shows up first.