FABRINET (FN): what the price assumes

In the published model solve dated 2026-Q2, anchored at $407.00, FABRINET (FN) is priced for today's economics sustained for ~10.2 years. boothcheck publishes no house fair value, target price, or buy/sell rating; individual model outputs and user-controlled scenarios are analytical inputs, not Boothcheck targets. Narrative composed 2026-07-26.

Generated: 2026-09-07 · Exported: 2026-09-09 · Source: https://boothcheck.com/report/FN

Headline

FieldValue
TickerFN
CompanyFABRINET
Sector / IndustryTechnology
Current price$407.00/sh
CompositionOptical communications 77% / Non-optical communications 23%

What The Price Assumes (Inversion)

The assumption today's price embeds, recovered by inverting the valuation.

FieldValue
Inversion basiswhole-company
Operating margin (value-band context)6.2%
Operating margin today10.0%
Margin compression (value-band)-3.8pp
Must persist for10.2y
Multiple paid30x operating income

The operating-margin figure is value-band context at year 7: derived from the framework's value band, a separate calculation — not part of the priced-in solve.

Solve inputs: computed at a 11.6% cost of capital; growth searched up to the 25% self-funding ceiling.

How unusual the bet is: within-range (limited comparison data)

ReferenceValue
vs own history+0.26σ
cohort percentile (of 190 peers)52

Valuation X-Ray

The price is justified by relative-multiple and growth-DCF; asset-based/earnings-power land below the price.

How the valuation models price the stock relative to the market price. Price/FV above 1.0 means the market pays more than that lens defends (expensive); at or below 1.0 the lens can defend the price.

FamilyMedian price/FVModelsReads
Asset2.85x5expensive
Earnings2.88x3expensive
Relative0.89x5justifies
Growth0.94x3justifies

Families that justify the price: Relative, Growth Families that call it expensive: Asset, Earnings

The models below discount at their own flat-beta convention rates (cost of equity 9.3%, WACC 9.2%); the inversion above states its own rate.

Per-Model Detail (n=16)

ModelFamilyFVPrice/FVApplicableMethodology
DCF Perpetual GrowthGrowth$281.001.45xyesFCF base $0.3B, growth 25% (input: historical growth), terminal g 4.0%, WACC 9.2%, 7yr projection
DCF Exit MultipleGrowth$432.290.94xyesExit EV/EBITDA: 26.9x / 29.9x / 32.9x (bear / base = today's held flat / bull), 7yr
Relative ValuationRelative$359.111.13xyesP/E 28x (static sector reference · 2026-04), scenarios: 22.4x / 28.0x / 33.6x (bear / base = reference held flat / bull), EV/EBITDA 20x
Simple DDMGrowthno
Two-Stage DDMGrowthno
Simple Excess ReturnAsset$142.702.85xyesBV/sh $68.49, ROE (TTM) 19.3%, ke 9.3%
Two-Stage Excess ReturnAsset$203.652.00xyes5yr excess ROE then converge to ke=9.3%
Discounted Future Market CapGrowth$559.540.73xyesRev $4.6B, growth 30% (input: historical growth; tapered), Terminal P/S: 2.5x / 3.1x / 3.8x (bear / base = today's held flat / bull, cap 12x)
Peter Lynch Fair ValueRelative$456.750.89xyesEPS $13.05, growth 35% (input: historical EPS growth), PEG=0.88 (Undervalued)
Margin TrajectoryGrowthno
Earnings Power ValueEarnings$68.725.92xyesNormalized EBIT (5y avg op income, one-time charges added back) $0.31B × (1−30%) / WACC 9.2% → EPV (no growth)
Residual IncomeAsset$199.022.05xyesBV $68.49 + 5yr PV of (ROE (TTM) 19.3% − Kₑ 9.3%) × BV; BV grows 8.8%/yr
Graham NumberAsset$141.812.87xyes√(22.5 × EPS $13.05 × BVPS $68.49) — Graham's conservative floor
EV/EBITDA RelativeRelative$274.691.48xyesEBITDA $0.48B × sector EV/EBITDA 20.0x
FCF YieldEarnings$9.3343.62xyesFCF $4.2M / Kₑ 9.3% — zero-growth perpetuity (excluded from median)
SBC-Adj FCF YieldEarningsno
Ben Graham FormulaEarnings$421.080.97xyesEPS $13.05 × (8.5 + 2×15.0%) × (4.4 / 5.3%)
ROIC-Justified P/BAsset$32.0812.69xyesBV $68.49 × (ROIC 4.3% / WACC 9.2%)
P/Sales SectorRelative$777.080.52xyesRevenue $4.64B × sector P/S 6.0x
PEG Fair ValueRelative$489.380.83xyesEPS $13.05 × (PEG 1.5 × growth 25.0% (input: historical EPS growth)) → PE 37.5x
Earnings YieldEarnings$141.082.88xyesEPS $13.05 / required return 9.3% (Rf 4.3% + ERP 5.0%)
Funds From Operations MultipleRelativeno
Clinical Phase NPVGrowthno
MertonAssetno
V5 Mechanicalno

Economic-Unit Decomposition (Sum Of The Parts)

Only one reportable economic unit is present in the current topology source. Consolidated operating lenses remain available, but there is no multi-unit decomposition.

UnitRoleValuation basisRevenueReported profitValue evidenceStatus
Fabrinet (single operating segment)operatingenterprise3.4B reported-currencywithheldunresolved no unit value

No total common-equity value is stated. One or more material units lack a supported unit value. The displayed values are an indicative subtotal; consolidated debt and cash cannot be applied to a fraction of the company.

Solvency

FieldValue
Net cash$875.8m
Net debt / NOPAT (after-tax)-2.71x (net cash)
Net debt / operating income (pre-tax)-1.89x (net cash)
Share count CAGR (buyback)-0.8%
Burning cashno

Interest expense is not separately reported in the latest filings, so interest coverage cannot be computed.

Bullet Takeaways

Bull Case

At the end of March the company held 356.6 million dollars in the bank and another 588.7 million in short-term investments, against no funded borrowings at all. More telling is what management is doing with it. Capital spending ran to 160.6 million dollars over the first nine months of fiscal 2026, more than double the 70.7 million a year earlier. Building factory floor ahead of the orders that fill it is not a neutral act. It is the clearest signal a manufacturer can send about what it has been told is coming.

The economics of the model deserve a moment, because they look wrong at first glance. Fabrinet keeps roughly twelve cents of gross profit per revenue dollar, while its own customers keep far more: LITE reports a 37.7 percent gross margin and CIEN 43.0 percent. Yet Fabrinet's operating margin of 9.9 percent for the nine months to March 2026 sits right alongside LITE at 9.5 percent and CIEN at 9.2 percent. It arrives at the same place from the other direction, by simply not carrying the expensive parts. No chip design, no protocol research, no worldwide sales organization. The customer pays for the invention; Fabrinet gets paid for making it work at volume, ten thousand units at a time, without defects. The 10-Q describes the arrangement as having "enabled us to help our OEM customers reduce their manufacturing costs while maintaining or improving the design, quality, reliability, and delivery times for their products."

That model is currently scaling in exactly the way it is supposed to. Operating income for the nine months to March 2026 rose 39.6 percent to 328.6 million dollars, and the operating margin widened from 9.4 percent to 9.9 percent. Widening margins during a volume surge is not the normal outcome in contract manufacturing, where extra scale usually gets handed straight back to the customer in price. The newest driver is not the optical business at all: non-optical products reached 857.1 million dollars, 25.8 percent of nine-month revenue, up 48.0 percent, which the filing attributes to "high demand for high-performance computing."

Capital allocation has been quietly consistent underneath all of this. The diluted share count fell in each of the last four fiscal years, from 37.6 million in fiscal 2021 to 36.3 million in fiscal 2025. That is not a dramatic reduction, and the company's own fourth-quarter guidance assumes the count holds around 36.3 million rather than falling further. The bear will note that a business this dependent on two customers should not be praised for financial conservatism, since conservatism is forced on it. Fair enough. But a manufacturer with no borrowings, nearly a billion dollars of liquid assets and a customer base of engineering organizations that take years to qualify a replacement has more room to lose an argument over price than the concentration figures alone suggest.

Bear Case

The entire price rests on one stream of demand continuing, and it is the newest one. Non-optical products, which the filing ties directly to "high demand for high-performance computing", reached 857.1 million dollars in the nine months to March 2026, a quarter of revenue, growing 48.0 percent year on year. Two years ago that line was a rounding error next to the optical business. Today's price asks the whole company to compound at roughly the fastest rate it can fund out of its own profits, and to keep doing so for about 13 years. Only about 15 percent of comparably fast-growing companies have sustained that pace for even a decade, and the ones that failed did not usually fail because demand vanished. They failed because it arrived somewhere else instead.

Then there is who the demand belongs to. The FY2025 annual report is unambiguous: "During fiscal year 2025, NVIDIA Corporation and Cisco Systems, Inc. contributed 27.6% and 18.2%, respectively, of our revenues." A year earlier the same two were 35.1 percent and 13.4 percent, so the mix shifts but the dependence does not. The company describes what that means for pricing in its own risk section: "Our reliance on a small number of customers gives our customers substantial purchasing power and leverage in negotiating contracts with us." In a business that keeps roughly twelve cents of gross profit on the revenue dollar, there is not much cushion between a hard negotiation and a materially worse year.

Cash is the third crack, and it is the one that has actually widened. Over the trailing four quarters the company earned about 421 million dollars of net income and converted about 45.8 million of it into free cash flow. For the nine months to March 2026, operating cash flow fell to 201.8 million dollars from 273.3 million a year earlier, while capital spending rose to 160.6 million from 70.7 million. Part of that is simply what fast growth looks like from the inside, since inventory and receivables have to be funded before revenue is collected, and the bull is right to say so. But a contract manufacturer carries its customers' components on its own balance sheet, and when the volume goes the other way, that inventory is a bill rather than an asset.

Finally, the tax line. Manufacturing is concentrated in Thailand under preferential treatment that carries obligations: the annual report notes an "agreement not to move our manufacturing facilities out of our current province in Thailand for at least 15 years from the date on which preferential tax treatment was granted", and states that "Currently, the corporate income tax rate for our Thai subsidiary is 20%." The rate the company has actually borne recently is a small fraction of that. A price built on capitalizing today's earnings for well over a decade is implicitly assuming an incentive regime lasts as long as the growth does, and those are two separate bets stacked on the same number.

Valuation

Close to 40 times a year's operating profit is what the shares cost today, on a business that assembles hardware other companies designed. To make that arithmetic work you assume growth stays at the ceiling the company can fund from its own profits, and stays there for something like 13 years. The sensitivity runs the other way from what you might expect: shave a percentage point off the growth rate and the required horizon stretches by a bit over two years, so the assumption is not brittle at the margin. It is simply long. Of companies that have grown comparably fast, only about 15 percent held the pace even a decade.

The methods disagree sharply about this, and the shape of the disagreement is the whole story. Only the forward-growth approaches reach today's quote. The price stands at roughly 3.7 times where the asset-based methods land and roughly 3.8 times where the earnings-power methods land, and even the peer-multiple approaches sit below it. Nothing anchored on today's book value or today's profit stream defends the price. What the market is paying for is durability, and durability is precisely the thing a static frame cannot see.

The single method that does reach the price is worth naming, because of how it gets there. The exit-multiple cash-flow approach projects seven years of growth and then assumes the business is sold at the same multiple of earnings before interest, tax, depreciation and amortization it commands today, held flat. Nothing compresses. That is a defensible convention and it is also the assumption doing most of the work.

The filed inputs underneath are solid enough. The operating margin ran to 9.9 percent for the nine months to March 2026, up from 9.4 percent a year earlier. Trailing revenue of about 4.2 billion dollars is growing near 29 percent, which places it between CIEN, at 5.6 billion dollars growing 30.6 percent, and LITE, at 2.5 billion dollars growing 69.0 percent. Both of those peers earn similar operating margins on gross margins three to four times as wide, which is the structural distinction between designing a product and building one.

Solvency barely enters the argument here. Net cash of 945.2 million dollars and no funded borrowings mean the downside is an earnings question rather than a survival question, and the interest the company pays is not a meaningful line. The item that actually deserves the reader's attention is the gap between reported profit and collected cash. A multiple this far above the static methods is ultimately a claim on cash that arrives later, and for the last three reported quarters that cash has been going into inventory, receivables and new buildings rather than out the other side.

Catalysts

The fourth-quarter print is the next scheduled test, and the bar is already public. Fabrinet guided to revenue of 1.25 billion to 1.29 billion dollars for the fiscal fourth quarter ending June 26, 2026, with GAAP earnings of $3.48 to $3.63 a diluted share on roughly 36.3 million fully diluted shares. That guidance was issued alongside a third quarter the company called a record: revenue of 1,214.3 million dollars against 871.8 million a year earlier, and GAAP earnings of $3.45 a diluted share against $2.25 the year before.

Management pointed forward as well as back. Chief executive Seamus Grady said on May 4, 2026 that the company expects "several new customer agreements, particularly in the datacom market, to further strengthen our growth trajectory as we move into the fourth quarter and beyond." New datacom customers matter more than the revenue they carry in the first year, because a contract manufacturer's dependence problem is solved by adding names to the list, not by growing the existing ones. Watch the concentration disclosure in the next annual report for evidence either way.

The other thing worth tracking through the year is capital deployment, since it is where the company's own forecast is visible before the revenue is. Capital spending of 160.6 million dollars over the first nine months of fiscal 2026, against 70.7 million in the same stretch a year earlier, is capacity being installed for programs that have not yet shipped. If revenue follows the buildings, the operating margin expansion already visible in the nine-month figures has more room. If it does not, the depreciation arrives anyway.

Peer Cohorts (Per Segment, With Filing Citations)

Fabrinet (single operating segment) (reported)

Methodology Note

Fundamentals sourced from SEC EDGAR filings. Current price from Databento. The priced-in inversion and valuation x-ray are computed by the boothcheck engine; narrative composed by AI from the structured data.

Sources

Q3 FY2026 report on Form 10-Q · FY2025 Form 10-K · Q3 FY2026 earnings release, May 4, 2026

View the full interactive FN report on boothcheck