Skip to content
primarysourced Photonics sector Tower
TSEM
~7 min read · 1,622 words ·updated 2026-04-29 · confidence 52%

Short interest history — Tower Semiconductor

Period: 2025-04-15 to 2026-04-15 ✓ (verified primary, internal companies/tsem/data/STOCK_SHORT_INTEREST.json)

As of: 2026-04-29 (latest data point: 2026-04-15)

Data source: FINRA short-interest settlement data (bi-monthly cycle); auto-refreshed by the update_finra data-pipeline job. ✓ verified-primary

Confidence legend: ✓ verified-primary FINRA · ◐ partial / aggregator · ⚠ inferred / estimate

Executive summary

TSEM short interest is structurally low by US-listed-mid-cap-semis standards — consistent with the dispersed-ownership profile (no controlling shareholder) and the short-locate-difficulty that often accompanies Israeli FPI ordinary shares (vs. ADR equivalents that have deeper borrow pools).

Short interest oscillated between 1.2% and 2.7% of float through most of the 12-month observation window, then stepped up materially in the final two settlements to 3.45% (2026-03-31) → 4.01% (2026-04-15) — corresponding precisely to the AI-photonics-premium blow-off into the all-time high $228.73 (2026-04-20). The interpretation: discretionary short sellers fading the parabolic move into the ATH, not structural bear conviction.

PhaseWindowAvg shares shortAvg % floatRead
Pre-AI-narrative-emergence2025-04 → 2025-08~1.55M1.4-1.7%Quiet baseline; no incremental short pressure
AI-photonics premium build2025-09 → 2026-02~2.10M1.8-2.7%Modest build coinciding with $74 → $130 rally
ATH-window short build2026-03-13 → 2026-04-15~3.44M3.0-4.0%Discretionary fade of the parabolic move

Comparison to specialty-foundry peers (analyst characterization ⚠): TSEM’s 1-4% short-interest range is substantially below GFS (which oscillated 11-21% during the comparable period — see ../../gfs/kb/06_market_data/short_interest_history). UMC and TSM short interest typically <2% in normal conditions. Tower-as-FPI ordinary-shares is structurally less-shortable than US-listed common-stock equivalents.

12-month short interest trajectory

Settlement DateShares Short% of FloatDays to CoverCatalyst Context (analyst-traced ⚠)
2025-04-151,925,6181.75%2.24Pre-recovery; cycle-trough sentiment
2025-04-301,924,1631.75%3.82Stable
2025-05-151,763,7241.60%2.91Modest covering
2025-05-301,906,7141.73%4.81Q1 2025 results re-build
2025-06-131,630,7691.48%3.44
2025-06-301,441,1961.31%1.98Pre-Q2 print covering
2025-07-151,361,4401.24%1.71Multi-month low
2025-07-311,465,2131.33%1.88
2025-08-151,364,7191.24%1.3112-month trough — ahead of Q2 2025 print
2025-08-291,776,8231.62%1.00Post-Q2 print partial re-build
2025-09-151,635,1761.49%1.11
2025-09-301,445,2611.31%1.41Pre-Q3 print
2025-10-151,394,3111.27%1.48Quiet
2025-10-311,582,7681.44%1.56
2025-11-141,979,0141.80%1.00Step-up: post 2025-11-12 CPO-foundry-technology press; AI-photonics narrative ignites
2025-11-282,093,1031.90%1.29Sustained build
2025-12-152,940,9332.67%1.62Acceleration: continued AI-photonics rally fading by shorts
2025-12-312,790,0122.54%2.40Year-end stable; partial covering
2026-01-152,441,0122.22%1.63Modest covering
2026-01-301,999,3461.82%1.24Pre-Q4 2025 print
2026-02-132,356,2792.14%1.00Around 2026-02-11 6-K Q4 2025 print
2026-02-271,763,5031.60%1.22Post-print covering
2026-03-132,110,4991.92%1.00Pre-LWLG-agreement build (LWLG announced 2026-03-11)
2026-03-313,792,7183.45%1.00Surge: parabolic-move fade; coincides with +17% 2026-03-19 single-day move
2026-04-154,411,8354.01%1.75Peak — highest reading in 12 months ahead of 2026-04-17 ATH close $226.45

Thematic analysis

1. The two-phase regime change

A structural break is visible in the series at 2025-11-14: pre-November short interest averaged ~1.5% of float; post-November averaged ~2.5%; the final two settlements jumped to 3.45% → 4.01%.

The 2025-11-14 inflection coincides with the 2025-11-12 Tower CPO-foundry-technology press release (GlobeNewswire ✓). This is the single most plausible AI-photonics-premium ignition event for TSEM — it formally pitched Tower into the CPO-foundry conversation. Short sellers responded by initiating fade positions immediately, but the rally accelerated faster than shorts could build.

2. The terminal blow-off short-build

The 2026-03-31 surge to 3.79M shares (3.45%) — a +1.68M share / +80% increase from the prior 2026-03-13 settlement (2.11M) — coincides with:

  • The 2026-03-11 LWLG-Tower PH18 development agreement (LWLG press ✓)
  • The 2026-03-19 +17% single-day move on 7.94M shares (the largest gap since the 2022-02-15 Intel-deal-announcement)
  • OFC 2026 conference week (March 16-20, 2026)
  • The 2026-03-31 +10.7% move (Tower price closed $175.48)

The next settlement (2026-04-15: 4.41M / 4.01%) extended the build by another +619K shares / +16% as TSEM continued to rally toward the 2026-04-17 ATH close $226.45 and 2026-04-20 intraday $228.73.

Read. This is not structural bear-thesis short interest. The build is concentrated in the parabolic-move window, not during cycle-trough or fundamental-narrative-degradation periods. The most plausible interpretation is discretionary momentum-fade trading by hedge funds that view the 6.6× peak-to-trough move and 60-70 PE multiple as overshooting reasonable forward earnings power. Days-to-cover 1.75 (current) is benign — no squeeze risk.

3. Days-to-cover trajectory

Days-to-cover oscillated in the 1.0 - 4.8 range through the period. The lows around 1.0 reflect the elevated daily volume during catalyst windows (when shorts could exit easily); the highs around 4.8 reflect quiet-period structural cover difficulty. Current 1.75 days indicates a complete unwind of the 4.41M short position would require approximately 2 trading days at average volume — squeeze potential is low.

4. Year-end pattern

The 2025-12-31 settlement showed only modest covering (2.94M → 2.79M = -5%) — consistent with non-tax-loss-driven positioning. Hedge funds were not closing the position for year-end book-management; they were sustaining the trade. This contrasts with GFS’s much larger ~32% year-end covering pattern over the same window.

Cross-checks for short-interest signal

SignalDirectionConfidence
Absolute short shares 4.41M (current)Modest bearish✓
Days-to-cover 1.75 daysLow squeeze potential✓
Short interest expanded 2.3× over 12 months (1.93M → 4.41M)Net-bearish position over the period✓
Surge concentrated in final 2 settlements (Mar 31 + Apr 15)Discretionary fade of parabolic move — not structural bear conviction✓
Borrow rate / cost-to-borrowNot in JSON; aggregator cross-check open ⚠⚠
Failure-to-deliver (FTD) dataNot collected ⚠⚠

Comparison to peers

PeerRecent short interest (% of float)Days-to-coverComment
TSEM4.01% ✓1.75Structural low; ATH-fade trade emergent
GFS14.89% ✓4.93High due to Mubadala-overhang + small-float math
TSM<2% ⚠<1Mega-cap; minimal short interest
UMC~3% ⚠~2Liquid Taiwan-listed
SMIC~5% ⚠~3China-listed; moderate short interest
VIS (Vanguard International Semi)~2% ⚠~2Mature-node Taiwan
Marvell~3-4% ◐~2Liquid large-cap AI semis

TSEM short interest is structurally below GFS by a factor of 3-4× — driven primarily by:

  1. No Mubadala-equivalent overhang. Tower has no controlling shareholder filing F-3ASR shelf registrations; institutional ownership is dispersed across Israeli pension funds (Menora 6.17%, Migdal 6.5%, Phoenix 5.77%, Harel 7.8%, Clal 4.8%) plus US institutional (./institutional_holders). No structural-supply-overhang short trade.
  2. No related-party-secondary cadence. GFS short builds time to Mubadala F-3ASR filings; Tower has no parallel structural-seller dynamic.
  3. FPI-ordinary-share borrow constraints. Tower’s NIS-15.00-par-value ordinary shares (CUSIP M87915274) trade as ordinary shares (not ADRs), creating slightly more friction in the share-locate process for institutional shorts ⚠.
  4. No material-weakness ICFR overhang. Unlike GFS which has carried a multi-year ICFR remediation flag, TSEM’s auditor relationship (Brightman Almagor Zohar / Deloitte ⚠ verify) is clean.

Forward catalysts that would force short cover

  1. Q1 2026 6-K (May 2026) delivers above-consensus revenue or PH18 traction commentary — the post-LWLG-agreement quarter would be a key data point.
  2. Named PH18 + LWLG customer disclosure — would re-rate the AI-photonics-premium thesis and force discretionary shorts to cover.
  3. CPO design-win disclosure with named hyperscaler — extension of the 2025-11-12 platform-positioning press into commercial wins.
  4. OFC / ECOC 2026 conference momentum — H2 2026 conference cycle could deliver additional named-customer demos.

Forward catalysts that would force additional shorting

  1. Q1 2026 6-K delivers below-consensus or guides cautiously on AI-photonics revenue near-term — would crystallize the “premium too high” bear case.
  2. Macro-driven AI-photonics-trade unwind — sector rotation away from photonics names would lift TSEM short interest broadly.
  3. PH18 commercial slip — public delays in PH18-LWLG productization.
  4. Customer-concentration risk realization — any named customer (e.g., RF SOI customer in Smart Mobile chain) reducing wafer commitments.

Sources

  • companies/tsem/data/STOCK_SHORT_INTEREST.json — 25 FINRA bi-monthly settlements 2025-04-15 to 2026-04-15. ✓ verified-primary internal artifact.
  • FINRA short-interest dissemination — bi-weekly settlement cycle; primary source for all NASDAQ-listed names.
  • companies/tsem/data/STOCK_PRICE_DATA.json — float and average-volume context.
  • TSEM CPO-foundry-technology press release (2025-11-12) — ✓ GlobeNewswire — narrative-ignition catalyst.
  • LWLG-Tower PH18 development agreement (2026-03-11) — ✓ LWLG press — terminal-build catalyst.

Cross-references