TSEM Valuation Ranges — Bull / Base / Bear Scenarios

This file maps three explicit valuation scenarios for Tower Semiconductor (NASDAQ: TSEM) with the assumptions per scenario. All price targets are analyst estimates labeled as such — they do not represent guidance from Tower, sell-side firm-specific consensus (which is aggregator-only without firm-by-firm sourcing as of 2026-04-29), or any externally-disclosed forecast. The framework is intentionally simplified to separate the SiPh-narrative-multiple drivers from the specialty-foundry-cyclical drivers.

Spot context (2026-04-29 close):

Price: $191.54 ✓ (per STOCK_PRICE_DATA.json ✓)
Market cap: ~$21.41B (111.76M shares × $191.54)
52-week range: $34.73 - $228.73
FY 2025 revenue: $1.57B ✓ (+9% YoY); net income $220M; diluted EPS $1.94 (per Tower Q4 2025 release ✓)
FY 2025 gross margin: 23.2% ✓ (per quarterly_trend.md ✓)
FY 2025 operating margin: 12.4% ✓
Net cash: ~$1.04B ✓ (FY 2024 YE: cash + ST deposits $1.22B, total debt $0.18B)
Implied trailing EV/Revenue: 13.0× ✓ (per comps_valuation.md ✓)
Implied trailing P/E: ~98.7×
Implied trailing P/B: ~8.0×

Anchor points:

Current spot $191.54 — base for valuation analysis
52-week high $228.73 — recent peak; within bull-case range
52-week low $34.73 — extreme floor reference
Intel-deal-canceled $53/share floor reference — August 2023 (Intel press release ✓); Tower’s stock collapsed to ~$24/share in weeks after termination establishing $24-53 double-bottom range through 2023-2024; broke above $53 mid-2024 as SiPh narrative began compounding
FY 2028 financial-model target — $2.84B revenue, 39-40% GM, 31.7% Op margin (per Tower Q4 2025 release ✓, TipRanks ✓). Note: Tower’s CEO Russell Ellwanger has been in role since 2005 with multi-cycle execution credibility.
Sell-side aggregator consensus ⚠ aggregator-only without firm-by-firm sourcing as of 2026-04-29 (per comps_valuation.md ✓ Section 7): mean target ~$190-220, high target ~$260-280, low target ~$120-140; coverage believed to include Citi, Needham, Stifel, JPMorgan, Susquehanna, Wells Fargo ⚠. Specific firm price targets and report dates are not publicly verified at this date. The 24/7 Wall St $230 column (24/7 Wall St ◐) is aggregator/opinion source, not sell-side primary

Valuation framework

Tower is a hybrid: ~85% of FY 2025 revenue from a diversified specialty-analog foundry business with cyclical mature-node-foundry-peer multiples (typical 2-4× EV/Revenue for VIS, DBHiTek per comps_valuation.md ✓); ~15% from AI-photonics-narrative-growth segment that warrants premium multiples (10-15× EV/Revenue at peer photonics pure-plays). The blended-multiple approach:

Sum-of-parts valuation framework:

Specialty-foundry segment: revenue × specialty-foundry-peer-multiple range (VIS, DBHiTek, GFS specialty comp set)
Silicon-photonics + SiGe segment: revenue × photonics-pure-play multiple
Adjusted for: net cash, customer-prepayment NPV, Israel-risk-premium discount

The simpler single-multiple framework treats the consolidated entity at a blended multiple appropriate to the AI-photonics-narrative weighting. This file uses the simpler approach for the headline scenarios and discusses sum-of-parts as a sensitivity check.

The structural relationship (per comps_valuation.md ✓ Section 4) is: at the FY 2028 model (revenue $2.84B, 39-40% GM, 31.7% Op margin → ~$900M Op profit + ~$500M D&A = ~$1.4B EBITDA, ~$6.70 EPS at 114M shares), the current EV $20.4B / FY 2028 EBITDA $1.4B = 14.6× FY 2028 EV/EBITDA — in line with TSMC and modestly above specialty peers. The current multiple is only justified if the FY 2028 model materializes mostly intact.

Core thesis assumptions:

All three bull pillars compound through 2027-2028 (per bull case)
AI-photonics premium sustains at current 13× EV/Revenue or modestly expands toward 15×
SiPh foundry duopoly with GFS Fotonix holds — Tower captures 40-50% of merchant SiPh share through volume + LWLG-EO-polymer differentiation
FY 2028 financial-model target achieved: $2.84B revenue, mid-30% GM (target 39-40%; bull achieves 35-37%), 31.7% Op margin
FY 2026 revenue: $1.85-1.95B (+18-24% YoY)
FY 2027 revenue: $2.30-2.50B (+24-28% YoY)
FY 2028 revenue: $2.75-2.90B (~$2.84B target +/- $50M)
FY 2026 gross margin: 26-28%; FY 2027: 31-34%; FY 2028: 35-37%
FY 2026 operating margin: 15-17%; FY 2027: 22-26%; FY 2028: 30-32%
SiPh segment trajectory: $228M FY 2025 → $400-500M FY 2026 → $600-800M FY 2027 → $900M-$1.0B FY 2028 (35% of mix at $1B / $2.84B)

Multiple assumptions:

Blended EV/Revenue: 9-11× (premium to 13× current sustained on validated execution; mild compression from current as revenue base expands)
SiPh-segment implicit multiple: 12-15× (premium peer photonics multiple)
Specialty-foundry-segment implicit multiple: 5-7× (above VIS / DBHiTek baseline due to NASDAQ-listed liquidity premium + customer-prepayment-funded capex)
Israel-risk-premium discount: stays at ~5-8% (Tower already trades at “Israel-tested” multiple)

Price-target derivation:

FY 2028 revenue $2.84B × 9× EV/Revenue = $25.6B EV
Less zero net debt to mild net cash $0-1B (analyst estimate ⚠ inferred — net cash compresses during capex peak FY 2026-2027) = $25.6-26.6B equity value
÷ 114M shares (post incremental modest dilution) = $224-233 per share at 9× multiple
Multiple expansion to 11× reflecting narrative strength: $31.2B EV / 114M = $274 per share
Range bull case: $250-300 per share (analyst estimate ⚠)

What needs to happen for the bull case:

Q1 2026 6-K (May 13 2026) reaffirms SiPh trajectory + RF Infrastructure mix expansion + GM holds >24% sequentially ✓ catalyst-tested
LWLG-PH18 first tapeout hits 110 GHz+ target during 2026 ✓ catalyst-tested
December 2026 capacity milestone hit on time (>5× Q4 2025 SiPh/SiGe wafer starts) ✓ catalyst-tested
Tower-Intel Fab 11X H2 2026 production ramp on schedule ✓ catalyst-tested
Additional named SiPh customer announcement (beyond Coherent + LWLG) in 2026-2027 ✓ catalyst-tested
Israel-geopolitical environment remains contained (no material Migdal Haemek operational disruption)
AI-capex through 2026-2028 sustains; no significant digestion cycle
First CPO Foundry Program customer disclosed 2026-2027

Bull-case high-end ($300): all eight conditions hold + multiple expansion + first CPO customer disclosed at hyperscaler-tier name + US CHIPS Act award materializes for Fab 11X.

Triggers/falsifiers — what makes the bull case happen or fail:

Trigger: Multiple LWLG-PH18 tapeouts at 110 GHz+ during 2026 + first CPO customer = “Tower-as-AI-photonics-foundry-of-choice” narrative crystallizes
Trigger: FY 2026 SiPh revenue prints at $450-500M = trajectory clearly inflecting to $1B FY 2028
Falsifier: Single tapeout failure at 110 GHz+ + Coherent shifts material volume to GFS Fotonix = bull thesis decays toward base case

Core thesis assumptions:

5-6 of the bull-pillar conditions hold; 2-3 underperform expectations
Specialty-analog cycle normalizes 2026-2027 with mild RF Mobile recovery (RF Mobile mix stabilizes 22-24%; modest GM-improvement)
SiPh ramps to $300-500M FY 2027 (slight underperform on bull trajectory) and $500-800M FY 2028 (well below $1B but meaningful)
Tower-LWLG integration delivers 110 GHz+ but not all tapeouts succeed; commercial production timeline extends to 2027-2028
Tower-Intel Fab 11X production ramps but slips ~6 months
December 2026 capacity milestone partially hit (3-4× rather than >5×)
AI-capex digestion cycle is mild but real
Valuation compresses from 13× EV/Rev to 7-9× EV/Rev as revenue base expands and SiPh-narrative-credibility settles to “credible but not transformative” tier
FY 2026 revenue: $1.70-1.80B (+8-15% YoY)
FY 2027 revenue: $2.00-2.15B (+15-20%)
FY 2028 revenue: $2.40-2.60B (15-20% miss vs $2.84B target)
FY 2026 gross margin: 25-27%; FY 2027: 28-31%; FY 2028: 31-34%
FY 2026 operating margin: 13-15%; FY 2027: 19-22%; FY 2028: 25-28%

Multiple assumptions:

Blended EV/Revenue: 7-9× (compression from current 13× as the bull narrative cools)
Israel-risk-premium discount: maintained at ~8-12%

Price-target derivation:

FY 2028 revenue $2.50B × 8× EV/Revenue = $20.0B EV
Less mild net cash $0.5B = $20.5B equity value
÷ 115M shares = $178 per share
Range with multiple-spread: at 7× → $153/sh; at 9× → $197/sh
Median base case: $150-200 per share

What characterizes the base case:

The market gives Tower partial credit for the AI-photonics narrative but discounts the FY 2028 model by 15-25%
Multiple compresses 4-6 turns from current as revenue base expands and growth-rate normalizes
AI-capex digestion happens but is mild; the FY 2028 model materializes ~12-18 months late
Tower’s premium vs VIS / DBHiTek narrows from +9-11 turns to +4-6 turns

Most-likely 12-month base case price: $170-190 — incorporating Q1 + Q2 + Q3 2026 6-Ks + FY 2025 20-F + LWLG-PH18 first-tapeout outcomes, with each catalyst delivering modestly favorable but non-explosive outcomes.

Triggers/falsifiers:

Base-case trigger: SiPh segment grows 30-50% YoY in 2026 (vs +115% YoY in 2025); GM expands +200-300 bps annually; RF Mobile cyclical recovery is mild
Falsifier upward: Multiple consecutive 6-K beats + LWLG-PH18 tapeout success accelerates trajectory toward bull case
Falsifier downward: FY 2026 SiPh growth slows below 30% YoY + RF Mobile cycle worsens = trajectory decays toward bear case

Core thesis assumptions:

2-3 of the bear-pillar conditions materialize simultaneously (per bear case)
AI-photonics adoption slows: hyperscaler 2026-2027 capex digestion compresses optical interconnect demand
SiPh customer wins decelerate; Coherent shifts material volume to GFS Fotonix (multi-source qualification incentive)
Tower-Intel Fab 11X ramp slips by 12-18 months; capacity comes online too late for FY 2028 model
Israel-geopolitical disruption produces meaningful operational impact (Migdal Haemek event-driven disruption flows to 6-K)
Valuation reverts to specialty-foundry comp set: 3-5× EV/Rev (toward VIS 3.2× / DBHiTek 2.0× / GFS 4.5× range with Tower at premium-to-pure-cyclical but well-below-current)
FY 2026 revenue: $1.55-1.65B (flat to modest growth)
FY 2027 revenue: $1.65-1.85B (modest 6-12% growth)
FY 2028 revenue: $1.80-2.10B (well below $2.84B target — meaningful FY 2028 model reset)
FY 2026 gross margin: 22-24% (compression on RF Mobile + ASP pressure)
FY 2026 operating margin: 10-13%
FY 2028 SiPh revenue: $400-600M (meaningful underperform vs bull-case $900M-$1B)

Multiple assumptions:

Blended EV/Revenue: 3-5× (multiple compression on AI-photonics-narrative-break)
Israel-risk-premium discount: widens to 15-25% on regional-conflict spike or general risk-aversion
Multiple compression accelerates if AI-capex-digestion narrative compounds with Israel-risk-spike

Price-target derivation:

FY 2028 revenue $2.00B × 4× EV/Revenue = $8.0B EV
Less zero net debt to mild net cash $0.3B = $8.3B equity value
÷ 115M shares = $72 per share
Range with multiple-spread: at 3× → $54/sh (revisits Intel-deal-anchor $53/sh region); at 5× → $96/sh
Median bear case: $60-100 per share

What characterizes the bear case:

AI-photonics narrative re-rating reverses (Pillar 3 bear case confirmed)
Multiple compresses to peer-foundry baseline (specialty-foundry comp set)
AI-capex-digestion cycle compounds with mature-node ASP recompression + Israel-risk-spike
FY 2028 model reset: Tower publicly resets target downward to $2.3-2.5B (from $2.84B)
Customer-prepayment quantum disclosed in 20-F at $200-300M (well below implied $644M from “70% reserved” framing)

Bear-case low-end ($60): AI-photonics narrative break + AI digestion + Israel regional-conflict event-driven disruption + Tower-Intel Fab 11X major slippage + customer-prepayment quantum disappointment + GFS Fotonix wins multiple hyperscaler shifts = revisits Intel-deal-anchor $53/share region (Intel press release ✓).

Triggers/falsifiers:

Bear-case trigger: SiPh revenue growth in any 2026 quarter compresses below 50% YoY + GFS Fotonix announces hyperscaler-tier customer that previously was Tower-PH18 = thesis-confirming events stack
Bear-case trigger: Israel regional-conflict spike + Tower 6-K disclosure of operational disruption = Israel-risk premium expansion
Falsifier upward: Any one bear pillar fails to materialize while 2-3 bull catalysts hit = path back to base case

Sensitivity grid — single-variable changes from base case

Variable	Base case	+1 standard deviation	-1 standard deviation
FY 2027 SiPh revenue	$400M	$550M (+$15 share)	$300M (-$15 share)
FY 2028 SiPh revenue	$650M	$850M (+$25 share)	$450M (-$25 share)
FY 2027 RF Mobile mix recovery	flat to +5%	+10% (+$5 share)	-5% (-$5 share)
Blended EV/Revenue multiple	8×	10× (+$45 share)	6× (-$45 share)
Tower-Intel Fab 11X production timing	H2 2026 productionized	H1 2026 (+$8 share, faster ramp)	H2 2027 (-$15 share, FY 2028 model reset risk)
Customer-prepayment quantum	$400-500M (analyst estimate)	$600M+ (+$8 share, capacity-funding de-risked)	$200-300M (-$15 share, capacity-funding pressure)
Israeli-risk premium	8-12% discount	5% (+$10 share, regional normalization)	20% (-$25 share, regional-conflict spike)
AI-capex digestion timing	mild 2026-2027	none (+$15 share, sustained tailwind)	severe (-$30 share, CY 2026 deceleration)
GFS Fotonix share-shift	flat	Tower captures share (+$20 share)	Tower loses share (-$25 share)

The dominant sensitivities are:

Blended EV/Revenue multiple — a 1-turn swing translates to ~$22 per share. The AI-photonics-narrative-credibility is the single most important variable for the equity.
Israeli-risk premium — asymmetric: regional-conflict-spike downside is materially larger than regional-normalization upside.
AI-capex digestion timing — broad asymmetric: severe digestion has $30+ per share downside potential, while sustained tailwind has $15 per share upside.

The Q1 2026 6-K (May 13 2026) is the most material near-term multiple-setting event — management has the opportunity to (a) reaffirm SiPh trajectory, (b) reaffirm December 2026 capacity milestone, (c) provide LWLG-PH18 tapeout-progress commentary. Any one of these positively-skewing modestly extends the multiple. Any disappointment compresses multiple by 1-2 turns.

Sum-of-parts sensitivity check

A sum-of-parts valuation provides a useful sensitivity check on the blended-multiple approach, calibrated to FY 2028 base-case revenue mix:

Segment	FY 2028 base case revenue	Comp multiple	Implied EV
Specialty-foundry (RF Mobile + Power + Sensors + Discretes)	$1.85B (74% of mix)	4.0× EV/Revenue (VIS / DBHiTek blended + Tower NASDAQ premium)	$7.4B
SiPh + SiGe	$0.65B (26% of mix)	12-15× EV/Revenue (photonics pure-play premium)	$7.8-9.8B
Total enterprise value	$2.50B	(blended ~6.1×)	$15.2-17.2B
Add net cash (analyst estimate)			+$0.5B
Implied equity value			$15.7-17.7B
Implied per-share			$137-154

The sum-of-parts gives a lower base-case range than the blended-multiple approach. This reflects the structural reality that SiPh is still a minority of consolidated revenue (26%) that cannot fully justify the blended-multiple premium if SiPh-segment-only revenue is the basis for the premium.

The reconciliation: the blended-multiple approach implicitly credits future SiPh growth beyond FY 2028 plus operating-leverage-driven margin expansion at the FY 2028 model — i.e., the implicit post-2028 SiPh runway plus the 31.7% Op margin target. The sum-of-parts approach takes a snapshot of FY 2028 revenue mix without future-growth credit. The gap between the two approaches ($137-154 sum-of-parts vs $150-200 blended-multiple base case) is a measure of how much the equity is implicitly being rewarded for SiPh-narrative growth optionality + margin-expansion execution.

Key implication: if the SiPh-narrative breaks (Pillar 1 bull case fails), the equity drifts toward the sum-of-parts valuation in the $130-150 range — between the base and bear-case midpoints. If the narrative compounds AND the FY 2028 margin target hits, the blended-multiple approach extends toward the bull-case high end. The current spot ($191.54) is approximately at the midpoint of the blended-multiple base case — implying the market gives partial credit for the SiPh-narrative growth optionality and partial credit for the FY 2028 margin trajectory but is not fully pricing in the bull case.

Cycle-low / cycle-high reference points

Reference	Price	Implication
52-week low	$34.73	Extreme floor reference; below Intel-deal anchor; would require severe combined-stress scenario
Intel-deal anchor	$53.00	August 2023 reference; Tower’s $5.4B all-cash acquisition collapsed on Chinese antitrust (Intel press release ✓). Hard-bear-case floor; reached in worst-case AI-photonics narrative-break + Israel-risk-spike
52-week high	$228.73	Within bull-case range; suggests the market has pushed close to bull-case mid in past 12 months
Spot 2026-04-29	$191.54	Mid-base-case — most-likely-12-month range $170-200
24/7 Wall St $230 column (24/7 Wall St ◐)	$230	Bull-case-mid aggregator opinion; consistent with sell-side aggregator high-target $260-280 range

Risk-adjusted expected value calculation

Combining the three scenarios with subjective probability weights:

Scenario	Price-target midpoint	Probability (analyst estimate ⚠)	Contribution to expected value
Bull	$275	25%	$68.75
Base	$175	50%	$87.50
Bear	$80	25%	$20.00
Probability-weighted expected value			$176.25

The probability-weighted expected value of ~$176 sits modestly below the current spot ($191.54) — implying the equity is modestly overvalued on a probability-weighted basis if the analyst probability estimates are correct. The asymmetry slightly favors the downside: bull-case multiple-expansion is +44% from spot; bear-case multiple-compression is -58% from spot. The 0.76:1 upside/downside ratio at the analyst-estimated probabilities supports a more measured view at current levels relative to the constructive view supported by GFS at $59.49 (where the GFS valuation_ranges ✓ probability-weighted EV is $65 vs spot $59.49).

Important caveat: the probability estimates are analyst priors, not consensus or formal risk-management probabilities. A more bullish view on AI-photonics tailwind and LWLG-PH18 tapeout success could shift the bull-case probability above 30% and the bear-case below 20%, which would push the probability-weighted expected value above the current spot. Conversely, a more bearish view on Israel-risk + AI-digestion would push the bear-case probability above 30% and the expected value below $160.

Comparison to consensus / peer multiples

Sell-side consensus (per comps_valuation.md ✓ Section 7, ⚠ aggregator-only without firm-by-firm sourcing):

12-month target range: ~$120-280 ⚠
Median target: ~$190-220 ⚠
High target: ~$260-280 ⚠
Low target: ~$120-140 ⚠
Believed coverage: Citi, Needham, Stifel, JPMorgan, Susquehanna, Wells Fargo ⚠ (firm-specific reports + dates not publicly verified at this date)
Rating distribution: ⚠ skews to Buy (AI-photonics narrative)

Specific firm price targets and dates are not separately verified from primary sources as of 2026-04-29. The sell-side coverage list is aggregator-curated; individual firm-by-firm extraction is queued for analyst coverage (not yet written). The 24/7 Wall St $230 column (24/7 Wall St ◐) is opinion / aggregator, not sell-side primary.

Peer-multiple context (per comps_valuation.md ✓):

Tower 13.0× EV/Revenue
VIS 3.2× — closest pure-specialty-foundry comp
DBHiTek 2.0× — Korean specialty equivalent
GFS 4.5× — direct AI-photonics merchant-foundry competitor
UMC 2.3×; SMIC 6.0×; X-FAB 0.8×; Hua Hong 2.7×; TSMC 10.0×

Tower’s premium of +9.8 turns vs VIS reflects the AI-photonics + SiPh growth-narrative re-rating. The bear-case interpretation: Tower is materially over-priced relative to specialty-foundry baseline, with $147 per share of “AI-photonics premium” above the cyclical baseline. The bull-case interpretation: Tower’s smaller revenue base means SiPh ramp is more concentrated and more accretive than at GFS — the same $1B SiPh revenue is 64% of Tower’s FY 2025 base vs 15% of GFS’s — so Tower has higher beta to the SiPh thesis.

What this analysis explicitly does not include

A detailed DCF model with line-by-line operating-cash-flow projections (a future-refresh-cycle expansion)
Sensitivity to interest rate environment (capex financing cost) — assumed neutral; relevant if Tower issues $300-500M debt to bridge $920M envelope per risks K2
M&A optionality (Tower as buyer or as target — note the Intel deal precedent at $5.4B / $53/sh August 2023 ✓)
Customer-prepayment-NPV explicit treatment — the $920M envelope at 70%+ reserved with prepayments has NPV value that is implicit in the bull-case multiple
Detailed FX-translation sensitivity beyond risks F1-F4 framework
US CHIPS Act award outcome — modeled as upside-only ⚠ probability-flagged at 25-40% per open questions Q6

These are not the dominant valuation drivers and are deferred to future-refresh-cycle expansions.

TSEM Valuation Ranges — Bull / Base / Bear Scenarios

Valuation framework

Bull case scenario — $250-300 / share (12-24 month horizon)

Base case scenario — $150-200 / share (12-month horizon)

Bear case scenario — $60-100 / share (12-24 month horizon)

Sensitivity grid — single-variable changes from base case

Sum-of-parts sensitivity check

Cycle-low / cycle-high reference points

Risk-adjusted expected value calculation

Comparison to consensus / peer multiples

What this analysis explicitly does not include

Cross-references