Phase Space AI

Valuation

Coherent [COHR]

Coherent Corp [COHR] — Valuation

Spot $222.19 (2026-07-29). Two outputs are produced, as required: a 12-month target and the 5-year implied-path test. Neither replaces the other.

---

1. COMPANY STATE — declared first

STATE B — cyclical / commodity-sensitive

Evidence for B, not A:

1. Three exogenous cycles drive the top line: AI datacenter capex (Datacenter segment), semiconductor-capital and industrial-laser demand (Industrial), and carrier spend (Communications). Management explicitly frames industrial as cycle-dependent: "we are now seeing demand signals that indicate a pickup in the growth of our industrial business over the course of this calendar year, led by strong orders from our semi-cap equipment customers."
2. Margin variance disqualifies A outright. GAAP quarterly operating margin over the last 16 quarters ranged −1.7% (Q1FY24) to +11.8% (Q2FY26) — a 13.5pp band. State A requires low-variance margin across ≥5 years.
3. A revenue cycle is visible in the record: $1,370.3m (Dec-2022) → $1,053.1m (Sep-2023), −23.1% peak to trough, then $1,805.6m (Mar-2026), +71.5% off the trough.
4. A structural regime break: the II-VI / Coherent combination closed 2022-07-01, and the Series B preferred converted in Q2FY26 taking the share count +26%. State A requires no transformative acquisition and no accounting-basis break.

Not C: COHR is profitable and has been for nine straight quarters; the terminal period does not carry essentially all the value (TTM EBITDA is $1.29bn). Not D: the reverse DCF returns a solution across the whole plausible exit-multiple range.

Consequence of State B, per valuation.md

• Terminal margin must be mid-cycle, normalised on BOTH revenue and margin. A peak margin against a cycle-average growth forecast is internally inconsistent and always flatters.
• The cycle position must be DATED, not merely normalised.

Dating the cycle — the single most important observation on this name

COHR's earnings are early-cycle and its multiple is late-cycle.

• Earnings, early-cycle: the datacenter business is rationing output to its own laser supply (CEO: growth "limited by the supply of indium phosphide lasers"), holds record bookings with orders placed "over a year" forward, and guides FY2027 growth above FY2026's +21.4%. A business turning away revenue is not at a cycle peak.
• Multiple, late-cycle: COHR trades at the 91st percentile of its own five-year price-to-TTM-revenue history (n=262 observations from 2021-05-17; current 0.0337 vs median 0.0165, p90 0.0331). And that percentile is understated, because the metric holds share count constant while COHR's count rose 26% in Q2FY26 — on a market-cap basis the current multiple is roughly 26% higher relative to history than the figure shown.

The forecast therefore begins early in the earnings cycle and at the top of the valuation cycle. That asymmetry is the whole COHR question, and it is why the exit multiple decides the answer.

evidence_grade: B. Filings are current and tie to AV; guidance is explicit and dated; the weaknesses are the absent AR-concentration disclosure, the absent design-win disclosure, and the unavailable consensus (§6 of the research file).

---

2. TERMINAL MARGIN — set from the company's own economics

Underwritten terminal EBIT margin: 20.0%.

Basis (not a peer median, not a cap)

The opex bridge — m_EBIT,T = m_gross,T − R&D − SG&A − other

Hard constraint m_EBIT,T ≤ m_gross,T: 20.0% ≤ 42.0% ✓. The bridge leaves 22.0pp of revenue for all R&D, selling and administrative cost against 26.0pp today — a tightening, not a fantasy.

Why 20.0% is above the trailing GAAP actual (10.28%) — the required causal argument

valuation.md requires that a terminal margin below trailing be justified; this one is above, and the bridge is mechanical rather than aspirational:

1. Acquired-intangible amortisation runs off. $2,958.2m of acquired intangibles sat on the balance sheet at 2026-03-31, amortising through the GAAP operating line. That is the bulk of the ~9.6pp GAAP-to-non-GAAP gap and it is contractually finite.
2. Restructuring ends. $160m in FY2025 alone, against site consolidations that are completing.
3. SBC remains and is retained inside the R&D and SG&A lines above — this is why 20.0% sits below the 21.3% guided non-GAAP figure rather than above it.

What the screen did, and why it is wrong

The screen carried terminal_margin: 0.144, terminal_margin_basis: "max(own, sector peer median)".

That is the fifth recorded failure of this one parameter (known-silent-failures.md §9.8) and it is a banned construction: max(own, industry_median) can only ever flatter a low-margin business and can only ever cap a high-margin one. Applied here it took a business guiding to a 21.3% non-GAAP operating margin and assigned it 14.4% — 5.6pp too low — because it was averaging COHR against a "sector" median it did not derive from COHR's own economics. Correcting it raises the terminal margin and therefore lowers the required growth: the correction moves the answer toward PASS, which is worth stating because it removes any suspicion that the correction was chosen for its direction.

---

3. THE EXIT MULTIPLE — derived, not asserted

valuation.md rule 4: the exit multiple is not an independent parameter.

EV_T / EBIT_T  =  (1 − t)(1 − g/ROIC) / (WACC − g)

Base exit multiple adopted: 15.0x, at the upper edge of the base parameterisation.

Sanity check against the shared comparator set (traded, 2026-07-29)

Recorded in full in COHR_analysis.json. EV/EBIT is distorted across this entire cluster because GAAP EBIT is depressed by acquired-intangible amortisation everywhere, so EV/EBITDA is shown alongside.

Reference-class assertion. valuation.md rule 6 requires n≥5 mature profitable firms, a matching operating model, and dispersion in the matched dimension. Only APH matches COHR's operating model — a vertically integrated interconnect components manufacturer earning a mid-to-high-30s gross margin. ANET, AVGO and CRDO are fabless or systems businesses at 63–68% gross margin; LITE and CIEN carry GAAP EBIT so depressed that their EV/EBIT figures (245x, 78x) carry no information. n=1 on a matched operating model, and the class therefore fails the n≥5 test: the peer-derived exit multiple is UNIDENTIFIED.

Consequently the derived multiple is used as the base and APH's traded 24.1x is carried only as the upper bound of the sensitivity. The base 15.0x sits below the only stated anchor (24.1x), which valuation.md prohibits absent a separately argued reason. The reason: APH's demonstrated 3-year revenue CAGR (32.2%) is 3x COHR's (10.9%) and its operating margin (27.2%) is 2.6x COHR's (10.3%), so anchoring COHR at APH's multiple imports growth and profitability COHR has not demonstrated — and APH carries no leverage against COHR's 1.7x. The argument is recorded rather than applied silently.

---

4. IMPLIED-PATH TEST — the Valuation Criteria

Terminal value is >90% of EV in every parameterisation, so the reverse DCF is the mandatory primary long-horizon output and the forward DCF is not run.

reverse_dcf.py --spot 222.19 --shares 195.639 --net-cash -776.1 --revenue 6602.4
               --years 5 --wacc 0.10 --terminal-margin 0.200 --solve cagr

Held fixed and named: terminal EBIT margin 20.0%; WACC 10.0%; horizon 5 years; TTM revenue $6,602.4m; diluted shares 195.639m (10-Q cover, verified); net debt $776.1m (10-Q, verified); EV $44,245m.

The required parameter

Flip point: exit multiple 23.6x. Above it the name passes; below it, it fails.

The margin the strategy ranks on

Demonstrated − required = 18.0% − 29.18% = −11.18pp.

Demonstrated is the TTM year-over-year rate (+18.0%). The 3-year CAGR is 10.9%, on which the margin is −18.3pp; both windows are reported. The 3-year window contains the FY2024 telecom inventory correction and understates the current regime, so TTM YoY is used as the primary measure and the 3-year figure is disclosed rather than discarded.

The implied compression, as a number

Current EV/EBIT is 65.2x. The base path exits at 15.0x — a compression of 50.2 turns, 77.0%. On EV/EBITDA: 34.2x today to an implied 15.0x × (20.0% ÷ (20.0% + ~6% D&A)) ≈ 11.5x, a 66% compression. The required path is not only 29.2% revenue compounding — it is 29.2% compounding into a multiple three-quarters below today's.

Cross-check: solve for terminal margin instead

At the demonstrated 18.0% CAGR and a 15.0x exit, the required terminal margin is 31.45% against the underwritten 20.0% — a shortfall of −11.45pp. Both axes of the surface fail by almost exactly the same amount, which is the signature of a price that is simply above the underwriting rather than of a single mis-set parameter.

Verdict: FAIL

The price requires 29.18% five-year revenue compounding against 18.0% demonstrated. Is there a "specific, evidenced reason" that reaches 29.2%, which would make this PASS WITH ARGUMENT?

The evidence reaches roughly 21–25%, not 29%: - FY2026 tracking +21.4% (9M actual $5,073.0m + Q4 guide midpoint $1,980m = $7,053m). - FY2027 guided above that rate — the guide sets a floor of 21.4%, not a level of 29%. - Q4FY26 guided +29.5% YoY — the single quarter does reach the required rate, but a one-quarter print is not a five-year path. - The named mechanisms (1.6T ramp, 6-inch InP, OCS, CPO) are real and specific, and they support acceleration above 21%. None of them is quantified to a level that supports 29% for five consecutive years, and design wins — the disclosure that would quantify it — are not published.

FAIL, by −11.18pp, with the exit multiple named as the decisive parameter and the flip point stated at 23.6x. A reader who believes COHR warrants APH's traded 24.1x should read this as a marginal PASS; that disagreement is located precisely, which is the point of reporting the flip point.

Sensitivity — run on the exit multiple, never on scenario probabilities

The table above is the sensitivity. A 1.0x change in the exit multiple moves the required CAGR by approximately 1.3pp. Nothing else in the model has comparable leverage: a ±2pp change in the terminal margin moves the required CAGR by ~1.9pp, and a ±100bp change in WACC by ~1.5pp.

---

5. TWELVE-MONTH TARGET

Target $271. +21.9% to spot $222.19.

valuation.md step 1 is "start from near-term consensus." That input does not exist for this name: AV EARNINGS_ESTIMATES returns an empty array for COHR, FN and LITE while returning 22–41 records for CRDO, CIEN and AAOI on the same key in the same minute. The substitution is declared here rather than hidden: company guidance replaces consensus, which is a legitimate substitute only because COHR's guidance is quantified, dated and filed (8-K 2026-05-06).

Build

Named product-cycle events inside 12 months

Each appears dated in COHR_Catalyst_Calendar.md: FY2026 Q4 results and the first FY2027 guide (mid-August 2026); the 1.6T VCSEL ramp management placed in "the second half of this year"; the CPO large-order ramp at "end of the year into next year"; OCS capacity post-bottleneck; and the industrial recovery led by semi-cap orders.

The honest caveat on this target

It is a multiple statement as much as an estimate statement. At the median of COHR's own five-year history the multiple roughly halves and the target falls far below spot. The +21.9% depends on COHR holding the 85th–91st percentile of its own valuation range for another year. That is the "multiple late-cycle" half of §1 restated in target form, and a reader should treat the revenue path as the well-evidenced part and the multiple as the assumption.

No external professional target was available for a sanity band on this name. Recorded as absent rather than substituted.

---

6. Liquidity Criteria and vehicle — PASS, equity

Vehicle: EQUITY. Ladder rung 1. No argument required, and none is offered.

Why not a LEAP. A 12.5–14.0 point implied-over-realised premium, paid for 17 months, on a name whose 12-month target is +21.9%. The $220 December-2027 strike carries 18 contracts of open interest and a 6.1-point bid-ask spread (6.3% of premium) — the framework's own record notes HCA at 18 contracts and CRDO at 78 as uninvestable. The January-2028 strike is deeper (149×147) but extends the premium to 18 months. Equity.

---

7. MEASURED Criteria

---

8. Summary card