CNXC — Concentrix Corporation · 2026-09-05 · Verdict: WATCH · Conviction 3
Price $32.16 (screen row from the 2026-09-05 universe build; no live prices in this sandbox) · Mkt cap $1.96B · EV $6.28B (screen said $5.64B) · EV/EBIT 5.2x guided non-GAAP, 12.0x guided GAAP · Adjusted FCF yield 32.6% guided, 26% on owner earnings · Net debt $(4.32)B · ADV $32.7M Sources read: 10-K filed 2026-01-28 for FY ended 2025-11-30 (Items 1, 1A, 7), 10-Q filed 2026-07-02 (Q2 ended 2026-05-31), DEF 14A 2026-02-13, 8-Ks 2026-03-23/03-24/03-27/05-01/06-29/07-24, Form 4s (12m). No transcript in the bundle.
1. What the business actually does
Concentrix sells outsourced customer experience and back-office work to more than 2,000 clients across five verticals, using about 455,000 staff in 74 countries (10-K 2025, Item 1). Roughly 99% of revenue is recognised as services are performed, priced on staffing hours or transactions at flat unit rates; contracts run one to five years and are cancellable by the client for any reason on 30 to 90 days' notice (10-K 2025, Item 7). No client exceeds 10% of revenue, top-30 tenure averages 16 years, and the company doubled itself by buying Webhelp for $3,774.8M in September 2023.
2. Why it is mispriced — the edge case
The edge case is real and dated. Groupe Bruxelles Lambert took 8,773,667 shares, 14.3% of the company, as Webhelp merger consideration and still held them in January 2026 (DEF 14A 2026-02-13). On 2026-04-29 GBL sold 6,000,000 shares at $22.25, cutting the stake to 2,773,667 (Form 4, 2026-05-01). The company confirmed the same day that this dropped GBL below the Investor Rights Agreement board threshold and that its designated director resigned effective immediately (8-K 2026-05-01, Item 5.02). That is a price-indifferent seller disposing of stock it received rather than bought, exiting near the low.
The problem is timing. That sale was four months ago at $22.25 and the screen price is $32.16, 45% higher; the CEO's last open-market buy was 1,000 shares at $21.25 on 2026-07-08 (Form 4, 2026-07-09). The overhang discount has largely been paid away. A residual 2.77M shares plus registration rights for two demand offerings of $100M-plus per rolling twelve months remain (DEF 14A 2026-02-13). Buying here is not buying the forced seller; it is buying a levered, slowly shrinking outsourcer at 4.4x EBITDA after a 50% rally, on $32.7M of daily volume that leaves no coverage gap.
3. Unit economics and growth
Revenue was $9,825.8M in FY25 against $9,618.9M in FY24 (10-K 2025, Item 7). Profit, not revenue, is what moves: non-GAAP operating income went $1,317.9M (FY24) to $1,253.5M (FY25) to a guided $1,200-1,230M (FY26), and adjusted EBITDA margin 16.2% to 15.0% to about 14.4% (10-K 2025, Item 7; 8-K 2026-06-29). Gross margin fell from 35.6% to 33.7% in H1 FY26; roughly 100bps is severance booked in cost of revenue, so underlying erosion is about 90bps (10-Q 2026-07-02). Constant-currency growth was 1.9% in Q1 and 0.6% in Q2, with technology down 5.8% and healthcare down 13.9% against banking up 12.6%.
The screen's headline loss is an artifact: FY25's $918.2M operating loss is entirely a $1,527.7M impairment, of which $1,523.3M is goodwill. Note how it arose. The September 1, 2025 annual test passed; the company then ran an interim test on November 1 only because "we experienced a sustained decrease in the market price of our common stock" (10-K 2025, Item 7). The write-off is the share price marking the balance sheet, not a change in cash flows.
Against that, the company's own risk factor says "some lower complexity services currently performed by our game-changers have been and may continue to be replaced by tools deployed by clients" (10-K 2025, Item 1A). AI displacement acknowledged in the past tense is why flat revenue is not obviously a floor.
4. Balance sheet and capital allocation
The screen understates the enterprise: it used long-term debt of $3,934.9M and ignored $650.0M of current maturities, so gross debt is $4,584.9M against $267.1M of cash including cash held for sale, giving $4,317.8M net debt and a $6.28B EV (10-Q 2026-07-02). Debt is unsecured, $2,150M of notes at 6.50%-6.85% maturing 2026-2033 plus $1,837.5M of term loans. Covenants require leverage under 3.75x and interest cover over 3.00x; the company is compliant, and on guided FY26 adjusted EBITDA of about $1,432M net leverage is roughly 3.0x against $1,504.6M of liquidity.
Capital allocation is the ugliest fact. Concentrix repurchased 3,556,736 shares at about $47 in FY25 and 2,200,819 at about $62 in FY24 (10-K 2025, Item 7), roughly 1 million at an average $40.06 in Q1 FY26 (8-K 2026-03-24), then nothing at all in Q2 while the stock traded at $22-27, with $396.6M of authorization unused (8-K 2026-06-29). Shareholders approved 3,700,000 new plan shares in March, 6.1% of the count (8-K 2026-03-27). Insider open-market buying over twelve months totals $118,089 across three purchases, against CEO compensation of $13,347,449 in FY25 (form4_last12m.csv; DEF 14A 2026-02-13). The 2025 TSR PRSUs were struck at $52.54 and need 75% TSR to reach threshold, so they are effectively worthless, and 65% of the annual bonus runs off non-GAAP operating income, the measure that excludes the $175M of restructuring being incurred.
5. Management: what they said vs what they did
The 2026-03-24 release headline was "Maintains consistent full year outlook," with FY26 revenue of $10.035-10.180B, constant-currency growth of 1.5-3.0% and non-GAAP EPS of $11.48-12.07. Thirteen weeks later revenue was cut to $9.925-10.025B, growth to 0.25-1.25%, EPS to $10.83-11.18 and GAAP operating income from $636-686M to $509-539M (8-K 2026-06-29). Adjusted free cash flow guidance of $630-650M was held both times, but H1 delivered $97.7M, so H2 must produce about $542M against $465.9M in H2 FY25, up 16% while operating profit falls. On 2026-07-23, twelve weeks after GBL's exit, the board doubled change-of-control severance to 2x salary plus bonus and added a 280G cut-back (8-K 2026-07-24); the filing calls it market alignment.
6. Valuation
Owner earnings are guided adjusted FCF of $640M less $105M of share-based compensation and about $25M of factoring benefit, roughly $510M, a 26% yield on the equity. Base (45%): constant-currency revenue flat, margin drifts to 13.5% by FY29, EBITDA $1,350M, net debt paid to $3.6B, 58M shares, 4.75x exit. $48. Bear (35%): displacement accelerates, revenue -4%/yr to $8.8B, margin 12.0%, EBITDA $1,056M, deleveraging stalls at $3.7B, 4.0x. $9. Bull (20%): iX Suite converts, 3% growth, 14.5% margin, EBITDA $1,580M, net debt $3.3B, 5.5x. $96. Probability-weighted $44, about 37% above price, with a 35% branch losing three-quarters of the capital. Reverse DCF: at $32.16 the enterprise is priced for a 3% perpetual decline in unlevered cash flow at a 9% WACC while the equity is priced for a 14% perpetual decline at a 12% cost of equity; the gap between those two numbers is the leverage, and it is the entire investment question.
7. Catalysts and timeline
Q3 results in late September test the 0.0-1.0% constant-currency guide and the H2 cash load; Q4 in January shows whether $630-650M was real. The term loan maturing 2026-12-27 is the refinancing marker. Resumed buybacks, or disposal of GBL's residual 2.77M shares, resolve the two open signals.
8. Risks and pre-registered kill criteria
- Constant-currency revenue growth below 0% for two consecutive quarters.
- Adjusted EBITDA margin below 13.5% in any reported quarter (Q1 13.9%, Q2 14.1%).
- FY2026 adjusted free cash flow reported below $560M against the $630-650M guide.
- No shares repurchased in fiscal Q3 and Q4 2026 despite $396.6M authorised, or a cut to the $0.36 quarterly dividend. Other risks: contracts cancellable on 30-90 days' notice, delivery-geography wage inflation, and 3.0x net leverage that turns a 25% EBITDA decline into roughly a 75% equity loss.
9. Verdict and one-paragraph summary
WATCH, conviction 3. Concentrix had a genuine, dated forced seller: GBL took 14.3% of the company as Webhelp merger consideration, sold 6,000,000 shares at $22.25 on 2026-04-29, and lost its board seat the same day. The screen is wrong twice in the company's favour, since the FY25 loss is a $1,523.3M goodwill write-off triggered by the share price itself rather than by cash flows, and the screen missed $650M of current debt, so real EV is $6.28B, or 4.4x guided FY26 adjusted EBITDA on a 26% owner-earnings yield. What stops it being an IDEA is that the dislocation has already been paid away: the stock is 45% above GBL's exit and 51% above the CEO's July purchase, on $32.7M of daily volume, so nothing identifiable is selling now. Underneath, adjusted operating profit has fallen two years running and is guided down again, gross margin is off 90bps ex-severance, guidance was cut thirteen weeks after being called consistent, H2 must deliver $542M of free cash flow against $97.7M in H1, and the 10-K itself says lower-complexity work "has been" replaced by client-deployed AI tools. Management bought stock at $40.06 in Q1 and none at $22 in Q2 with $396.6M authorised, which is their own verdict on the discount, and insiders bought $118,089 all year. At 3.0x net leverage the equity is a levered option on whether flat revenue holds, and with no transcript in the bundle the AI-displacement question cannot be tested here.
Research for discussion, not investment advice. Positions and sizing are the reader's decision.