FLEX — Flex Ltd.

As of 2026-09-13 the sell-side is firmly bullish: 19 Buy vs 7 Hold ratings, a $147.50 consensus (recent-average $152.50) target against the $115.78 price, and reported institutional accumulation by Third Point and Soros. The 2026-09-13 session's +7.2% pop followed the early-September announcement of the $4.4B EPC Power acquisition and the reaffirmed early-2027 CPI spin-off, which reframes Flex from a ~15x-P/E contract assembler into an AI data-center power play. The caveat: the market is bullish on the story but already violently re-pricing it — the stock is still down 22.7% over three months and 7.0% over one month after running +82.5% in six months.

This is a genuine inflection, not a blip. Revenue YoY accelerated for four straight quarters: +4.0% (Sep 2025 vs Sep 2024), +7.7% (Dec 2025), +16.9% (Mar 2026), +20.6% (Jun 2026), with absolute revenue climbing $6.58B → $6.80B → $7.06B → $7.48B → $7.93B. Gross margin expanded from ~7.5% in late 2023 to 9.4% in Jun 2026 (9.7-9.8% in the two prior quarters), though operating margin has stayed stubbornly flat at ~5% and actually dipped to 4.9% in Jun 2026 despite record revenue — consistent with new-program ramp costs offsetting mix. Diluted EPS moved $0.51 → $0.45 → $0.34 (Jun 2024) → $0.54 → $0.67 → $0.57 → $0.50 → $0.52 → $0.64 → $0.67 → $0.76, a clear upward stair-step over the last six quarters. Note the Mar 2024 quarter is distorted — a -$162M operating loss against +$395M net income ($0.93 EPS) from a one-off gain — so trailing comparisons through mid-2024 overstate the base. Free cash flow is steady, not accelerating: ~$1.07B TTM across the last four quarters (283/211/272/303M), with capex rising (236M in Jun 2026 vs 133M a year earlier) to fund growth.

Expensive on trailing, defensible only if forward estimates land. As of 2026-09-13 the trailing P/E of 44.7x (EPS TTM $2.602) sits roughly 3x Flex's own 5-year average of 15.53x, and P/S of 1.46x is nearly 3x its 5-year average of 0.51x; EV/EBITDA is 24.4x and PEG (TTM) is 3.41. Forward, the picture softens: consensus next-FY EPS of $4.698 puts the forward P/E at ~24.6x, which is a premium but not absurd for a business now compounding revenue at +20%. The honest framing is that Flex has genuinely re-rated from an EMS multiple to an AI-infrastructure multiple — the question is whether ~25x forward is right for a company still earning 5.2% operating and 3.3% net margins, versus 20x (a multiple that better matches the margin structure). Valuation is a headwind here, not a tailwind.

The market is paying today for margin expansion that hasn't shown up yet. Consensus next-FY EPS of $4.698 implies ~80% growth over TTM $2.602 and requires net margin to expand from 3.3% to roughly 5.0% on ~$34.6B of revenue — a ~170bp jump, well above anything visible in the reported numbers (operating margin has been flat at ~5%, and actually slipped to 4.9% in Jun 2026). Revenue is arguably ahead of earnings — the company is growing the top line faster than it is converting it to profit. The practical read: as long as CPI organic growth (guided 65-75% for FY27) and gross margin keep moving the right way, the forward multiple is defensible and estimates can keep grinding up. If either stalls, both the estimate and the multiple compress at once — that is exactly what the -22.7% three-month move already demonstrated, and it argues for weighting the actual quarterly CPI numbers over the narrative.

As of 2026-09-13, RSI stood at 51.23 — dead neutral, neither oversold nor overbought. Price was +2.73% vs the 20-day SMA but -3.2% vs the 50-day, while sitting +20.65% above the 200-day SMA. Read: a strong long-term uptrend (6-month return +82.47%) that has just been through a sharp 3-month correction (-22.66%), with the single-day +7.19% bounce on 2026-09-13 starting to reclaim short-term trend. Entry-timing implication: this is a 'prove it' setup rather than a clean oversold entry — RSI in the mid-50s means the correction has neither flushed out nor fully repaired, and the stock remains extended versus the 200-day. A reclaim and hold of the 50-day SMA would be the constructive signal; a failure that puts the 200-day back in play (~$96 area) would test the thesis harder.

• AI/hyperscaler data-center capex cycle — the demand engine behind the Cloud & Power Infrastructure (CPI) segment and the single biggest swing factor in both the +20.6% YoY revenue growth and the planned spin-off valuation.
• Interest rates and cost of debt — the $4.4B EPC Power deal is part-debt funded and Fitch projects ~3.4x peak consolidated EBITDA leverage in FY27, so a higher-for-longer rate path directly pressures the equity case.
• Grid power constraints and electrification — data-center power management, conversion and cooling demand is precisely what re-rated Flex from a ~15x P/E assembler to a ~25x forward AI-infrastructure name.
• Tariffs, trade policy and supply-chain regionalization — Flex's distributed global manufacturing footprint (Mexico, Asia, US) is exposed to tariff shifts and customer re-shoring mandates that can move program economics quarter to quarter.
• Sector rotation and multiple compression across high-multiple AI infrastructure names — the 22.7% three-month drawdown shows how fast this cohort de-rates when the AI trade cools, regardless of company-specific execution.

• Closing of the $4.4B EPC Power acquisition (expected toward end of calendar 2026) and disclosure of the final debt/equity funding mix — the funding split directly determines post-deal leverage and share count.
• Form 10 filing and structural separation of the CPI business for the early-2027 spin-off — the clearest potential sum-of-parts value unlock in the story.
• Quarterly CPI organic growth against management's guided 65-75% for FY27, plus evidence gross margin (9.4% in Jun 2026, up from ~7.5% in late 2023) keeps expanding rather than plateauing.
• Deleveraging back below 3.0x EBITDA after the spin-off (per Fitch) — would defend the BBB- investment-grade rating and remove a key overhang.
• Analyst estimate revisions and target resets — the gap between the $115.78 price and the $147.50-152.50 consensus target (as of 2026-09-13) leaves room for either upgrade cycles or a downward target reconciliation.

• Leverage and integration risk: the $4.4B debt-funded EPC Power deal carries Fitch-projected peak leverage of ~3.4x in FY27, and the balance sheet is already moving — long-term debt jumped from $3.75B (Mar 2026) to $5.22B (Jun 2026) against only $5.5B of stockholders' equity.
• Valuation risk: 44.7x trailing P/E and 24.4x EV/EBITDA against a 15.5x five-year average P/E leaves almost no margin of safety; the -22.7% three-month drawdown is a live demonstration of how fast the multiple can reset.
• Structurally thin margins: 5.2% operating and 3.3% net TTM leave very little cushion, and Jun 2026 operating margin (4.9%) actually fell sequentially despite record revenue — mix or ramp costs can erase operating leverage quickly.
• Forward estimate credibility: the $4.698 consensus next-FY EPS assumes ~80% growth over TTM $2.602 and net margin expansion to roughly 5.0%; if AI capex slows or margins stall, both the estimate and the multiple applied to it fall together.
• Spin-off execution risk: separations are complex, and if the stub is left holding the acquired debt while the growth unit is priced separately, the blended multiple can end up lower than the pre-spin whole.

Flex's operating story is genuinely inflecting — revenue growth accelerated four quarters running (+4.0% → +7.7% → +16.9% → +20.6% YoY), gross margin climbed from ~7.5% to 9.4%, and the AI data-center power buildout plus the planned early-2027 CPI spin-off are real, dated catalysts. But at $115.78 on 2026-09-13 you are paying 44.7x trailing and ~25x forward earnings for a business still earning ~5% operating margins, while a $4.4B debt-funded acquisition pushes leverage toward 3.4x and the June quarter already showed debt jumping $1.47B. That is a good business at a full-to-rich price, not a bargain — and the 22.7% three-month drawdown suggests the market agrees it ran ahead of itself. I'd want either a pullback toward the high-$90s/low-$100s or hard evidence that CPI margins are landing before getting constructive; confidence is capped by the pending acquisition and spin-off, which make forward estimates unusually hard to pin down.

Blended forward P/E approach on fiscal-year-end-March estimates: I take the $4.698 consensus next-FY EPS, trim it ~4% to $4.50 for the leverage and integration risk of the $4.4B debt-funded EPC Power deal, and apply a 25x multiple — a justified re-rating above Flex's 15.5x five-year average given +20.6% YoY revenue growth and AI-infrastructure exposure, but well below the 44.7x trailing P/E the market currently pays. That yields $112.50. A simplified DCF cross-check (TTM FCF $1.07B, ~25% five-year FCF growth, 9% discount rate, 3% terminal growth, net debt ~$4B post-deal) lands near $110, so I anchor at $112.50 — roughly a 3% discount to the 2026-09-13 price of $115.78, i.e. the stock is approximately fairly valued.