Financebotresearch desk研究台

OLED › analyze

OLED · Analyze

ACCUMULATE Technology

Price at analysis: $84.85 | 52-week range: $76.42 – $153.38 | All-time high: $262.77 (2021)

0. Knowledge Check

Knowledge/INDEX.md and python .mcp/kb.py find OLED returned no live note — this is the first pass on the name in the knowledge base. Pitfalls checked and applied below: [[pitfall-vendor-forward-eps-is-the-wrong-fiscal-year]] (fired — see Valuation), [[pitfall-dyt-inverts-when-price-caused-the-yield]] (applied — see DYT), [[principle-down-a-lot-is-not-cheap]] (applied — see Valuation band test). OLED is a US-domiciled Delaware company (Ewing, NJ), not an ADR — the ADR-specific pitfalls (share-count, book-value, currency-mixed EV) do not apply. roic.ai's FCF figures matched fin.py's to the thousand ($144.36M FY25, $211.10M FY24) — no vendor FCF divergence found this pass.

1. Fundamentals

Universal Display licenses and sells phosphorescent OLED (PHOLED) materials and IP to panel makers (Samsung Display, LG Display, BOE, CSOT and others) for use in smartphone, TV, IT/PC and automotive displays. Revenue is two lines: material sales (emitters sold into production) and royalty & license fees (IP licensing, which includes cumulative true-up adjustments that make the line lumpy quarter to quarter).

Metric 2022 2023 2024 2025 TTM (thru Q2'26)
Revenue $616.6M $576.4M $647.7M $650.6M $606.9M
Gross margin 79.3% 76.5% 77.1% 76.3% 73.8%
Operating margin 43.3% 37.7% 36.9% 38.2% 35.3%
Net income $210.1M $203.0M $222.1M $242.1M $195.6M
FCF $79.6M $28.4M $211.1M $144.4M ~$170.0M*
Diluted shares 47.47M 47.62M 47.65M 47.66M 46.7M

TTM FCF summed from quarterly cash flow: Q3'25 $84.3M + Q4'25 $5.1M + Q1'26 $60.3M + Q2'26 $20.3M. The quarter-to-quarter swing is large — Q4'25 and Q2'26 both show working-capital drags (inventory build, receivables build) and Q1'26 includes a ~$40M one-time payment tied to the Merck emissive-OLED patent-portfolio acquisition, which is strategic IP spend, not maintenance capex. Adjusting it out, normalized TTM FCF is closer to $210M (~$4.50/share)*.

Revenue is decelerating, and the deceleration is intra-year, not a stale annual read. TTM revenue ($606.9M) already sits below FY2025's full-year print ($650.6M). Q2'26 revenue was $152.2M vs. $171.8M a year earlier (−11.4% YoY); material sales fell ~26% YoY on weak smartphone volumes and rising component costs. Royalty/license revenue is the volatile line: Q1'26 was $54.2M vs. $73.6M in Q1'25 (−26%), but Q2'26 rebounded to $81M vs. $76M (+6.6%), helped by a ~$9M cumulative catch-up adjustment — read that quarter's "growth" with the true-up stripped out. Full year 2026 guidance was lowered to the low end of $630–670M, essentially flat-to-down against 2025 — consistent with fin.py's 3-year revenue CAGR of only 1.8%. Net income 3yr CAGR: 4.8%.

Capital allocation. Debt is negligible ($19.2M against $1.96B assets — Debt/Assets ~1%, functionally debt-free). The company sits on ~$471M of cash and short-term investments and carries negative net debt (EV $3.45B < Market Cap $3.90B) — no balance-sheet risk of any kind. Capital returns are split between a 9-year unbroken dividend-growth streak and a buyback program that resumed in 2025 ($32.9M) and accelerated through H1'26 ($48–67M/quarter) alongside the ~$40M strategic patent purchase. Shares outstanding are essentially flat over 3 years (CAGR +0.1%) — the buyback is offsetting dilution, not meaningfully shrinking the float.

Per-share: Revenue/share $13.65 (fin.py) / $13.68 (roic.ai), FCF/share $3.03 (reported) to ~$4.50 (normalized), Book value/share $36.42–37.26.

Dividend Growers overlay (applies — OLED is a dividend payer with a 9-year raise streak): Dividend run-rate is now $2.00/yr ($0.50/quarter, raised March 2026), yield 2.3%, payout ratio 46% of TTM EPS. The 9-year dividend CAGR since the 2017 initiation ($0.03/qtr → $0.50/qtr) is arithmetically large (~37%) but that is a near-zero-base artifact; the more honest read is the annual raise rate, which has decelerated every year: +100% (2018) → +67% → +50% → +33% → +50% → +17% → +14% → +12.5% → +11% (2026). This is the normal decay curve of a maturing payer, not a red flag — coverage is comfortable (46% payout against FCF that, even in the weak 2023 year, covered the dividend $28.4M FCF vs ~$57M dividends only barely — the one year worth flagging; every other year FCF covered dividends 2–4x over).

2. Moat & Competitive Advantage

A portfolio-specific passage was removed from the public build.

Adversarial stress-test — "how would a well-funded rival attack this?" The honest answer is: not easily, and not soon, but the one open wound is real. Universal Display holds the dominant IP position in phosphorescent OLED (PHOLED) emitters — the technology that lets OLED panels hit commercial power-efficiency targets. It reinforced that position in 2023/2025 by buying Merck KGaA's entire emissive-OLED patent portfolio (550+ patents, 172 families) rather than competing against it, which took one credible rival off the board entirely. Remaining named competitors — Idemitsu Kosan, Sumitomo, Kyulux, Cynora — sell or sample alternative emitter chemistries (including TADF/hyperfluorescence) but lack PHOLED's commercial maturity today.

The genuine vulnerability is blue. The industry-wide, decade-plus unsolved problem is a commercially viable blue phosphorescent emitter — UDC still does not ship one at scale. Blue is roughly a third of every OLED pixel, and every panel currently uses a lower-efficiency fluorescent-blue subpixel alongside UDC's phosphorescent red/green/yellow. This is not an existential gap today (UDC's royalty structure is tied to the broader device architecture and patent portfolio, not solely to material sales, so panel makers keep paying even without a UDC blue emitter), but whoever cracks commercial blue PHOLED first — UDC itself, or a rival selling directly to Samsung/LG — resets the competitive map. UDC's own 2026-08-19 "OLEDX" architecture announcement (a transparent-cathode/reflective-anode/nanoparticle outcoupling structure aimed at near-100% light-extraction efficiency) is management's answer to this pressure and is worth tracking as a catalyst, not yet a proof point — it is a lab-stage architecture announcement, not a shipped product.

Customer concentration is the other named risk, and it is structural, not cyclical. UDC does not disclose an exact customer-revenue breakdown, but Samsung Display and LG Display together represent roughly 61.5% of global OLED material market volume (Samsung ~41%, LG ~20.5%) — a reasonable proxy for UDC's own revenue concentration given it is the near-sole phosphorescent material and IP supplier to both. A demand air-pocket at either customer (exactly what happened in H1'26 on weak smartphone volumes) flows straight through UDC's income statement with no diversification to cushion it.

China exposure — a genuine tailwind with an open question attached. BOE and CSOT are both long-term licensees (since 2020 and 2019/2020 respectively) and both are gaining OLED panel share/volume, which is a structural growth vector as China's panel makers scale. The open question is enforceability and pricing power in that market over a 10-year horizon as domestic Chinese material suppliers mature — Samsung's own 2025 patent-infringement win against BOE (a different case, Samsung's IP not UDC's) is at least a signal that OLED-related patent enforcement in China is active and can succeed, which is a mild positive read-through for UDC's own licensing durability, not proof of it.

Disruption forecast (5–10yr): Base case is continued but slowing volume growth as OLED penetrates IT/PC and automotive displays beyond its smartphone/TV base, offset by a genuine technology risk (blue) that UDC is racing to close with its own R&D (OLEDX) rather than acquiring away (as it did with Merck). Evergreen assessment: near-evergreen, not fully evergreen — the IP moat is real, defensible, and reinforced by acquisition, but it depends on UDC continuing to win the blue-emitter race rather than a rival (or a customer's in-house program) winning it first.

3. Valuation

⚠️ Forward P/E pitfall fired again — [[pitfall-vendor-forward-eps-is-the-wrong-fiscal-year]]. fin.py's reported "PE(fwd) 17.82x" prices forwardEps $4.7612 — but Yahoo's own epsCurrentYear/priceEpsCurrentYear fields show the true current-year (FY2026) figure is EPS $4.19 at a P/E of 20.23x. The $4.76 "forward" figure is FY2027 consensus, one fiscal year further out, and it always makes the stock look cheaper than it is. Use 20.2x as the current-year multiple, not 17.8x — a ~2.4-point understatement in the direction that manufactures a false bargain, exactly as the pitfall predicts.

Model Output Weight Rationale
Graham IV √(22.5×EPS×BVPS) √(22.5 × 4.14 × 36.42) = $58.25 vs. price $84.85 (price 46% above IV) Low Graham penalizes an IP-licensing business — BVPS ($36.42) captures almost none of the patent portfolio's economic value, so the model understates fair value for this business type. Reported as a reference floor, not a decision input.
Bogle Expected Return Div yield 2.3% + sustainable growth ~9.3% (roic.ai sustain_growth_rt, i.e. ROE × retention) ± no assumed P/E re-rating = ~11.6%/yr base case. Partial multiple mean-reversion (current 20.2x vs. FY2025's own 20.4–32.3x range and FY2024's 31.1–50.7x range) would add further upside but is not the base case. Highest Best-fit model for a moderate-growth, moderate-yield industrial-IP compounder.
DYT Current yield 2.3% vs. 5yr avg yield 0.96% — screens "cheap" on yield history Partial Decomposition per [[pitfall-dyt-inverts-when-price-caused-the-yield]]: the dividend roughly grew 2.5x over 5 years while price fell ~47% over the same span — the elevated yield is genuinely about half dividend growth and half price decline, not a pure price-caused inversion. Carries real but partial weight.
DDM (2-stage) D1 $2.20 growing 10%/yr for 5yrs, terminal growth 4%, discount rate ~12% (CAPM: 4.2% rf + 1.56β × 5% ERP) → ~$33 Very low / sanity check only DDM structurally undervalues a company with a 35–46% payout ratio — it prices only the dividend stream and ignores the ~55–65% of earnings retained into buybacks and balance-sheet compounding. Not decision-driving; included per framework for completeness.
FCF/owner-earnings multiple Normalized FCF/share ~$4.50 (excludes the one-time Merck patent payment) × 20–24x = $90–108 High Cross-check against a reasonable multiple for a debt-free, ~75%-gross-margin, cyclical-but-structurally-growing licensor.

Band test ([[principle-down-a-lot-is-not-cheap]]): Current TTM P/E (20.5x) sits at the low end of FY2025's own multiple range (20.4x–32.3x) and below FY2024's own low (31.1x). This passes the "cheap against its own band, not just off the high" test — a genuine signal, not a manufactured one. The companion clause (is the earnings base permanently reset?) reads as no: the deceleration is a cyclical smartphone-volume/component-cost air pocket layered on a still- structurally-growing OLED penetration story (IT/PC, automotive), not a share-loss or technology-displacement story. That said, full-year 2026 guidance at the low end of $630–670M means the base has not yet demonstrably re-accelerated — this is a thesis to monitor at Q3, not one to declare resolved.

Fair-value range: $85–108, weighted toward Bogle and the normalized FCF multiple, with Graham and DDM carried as low-weight floors given the model mismatches described above. At $84.85 the stock sits at the bottom of that range — priced for the cyclical trough, not for a permanent impairment, but not yet pricing in the OLEDX/blue-emitter optionality either.

4. Sentiment — why is it near the 52-week low right now?

The drawdown is real and well-explained, not mysterious. Three compounding factors:

  1. A genuine Q2'26 miss-then-mixed-beat cycle. Q1'26 EPS was $0.76 against a ~$1.28 estimate, revenue $142.2M against ~$168.4M expected — weak smartphone demand and rising component costs were named directly. Q2'26 beat on EPS but the "beat" leaned on a one-time $9M royalty catch-up; material sales continued to lag (down to $66M from $89M YoY).
  2. Guidance cut to the low end of the full-year range, which is what actually re-rates a stock — Susquehanna and Citigroup both cut price targets to $85–90 in the days following; Oppenheimer and Needham held Buy/Outperform but trimmed targets to $100–120. Consensus remains net-positive (average rating 1.78/"Buy", 9 analysts, mean target $115.37) — the sell side has not turned bearish on the name, it has recalibrated the near-term number.
  3. Insider signal is the most useful tell here, and it is constructive. In May 2026, with the stock trading $92–94 (above today's $84.85), CEO Steven Abramson bought 11,000 shares (~$1.03M) and Chief Legal Officer Mauro Premutico bought 3,694 shares (~$345K) — open-market purchases, not option exercises or scheduled awards. That is a genuine golden flag: management put personal capital to work at a higher price than today's, which reads as conviction in the cyclical-not-structural framing above. The only insider sale on record recently was a small, scheduled-looking director sale (Lacerte, 4,967 shares, Aug 18) — nothing that reads as a pattern.
  4. The August 19, 2026 OLEDX architecture announcement (light-extraction efficiency platform) drew a modest positive share reaction and gives the bull case a forward catalyst — but it is a lab-stage unveiling at a display-industry conference (IMID, Busan), not a shipped product or a disclosed customer commitment, so it should not be over-weighted yet.

Read: this is a sentiment/cyclical drawdown (weak smartphone-cycle demand + component-cost pressure flowing through a concentrated customer base), not a structural break in the OLED penetration thesis. The insider buying at a higher price than today is the strongest piece of corroborating evidence for that read; the customer-concentration and blue-emitter risks are the reasons this isn't a higher-conviction call.

5. Synthesis — Weighted Verdict

Verdict: ACCUMULATE. Conviction: 6.5/10.

Weighting rationale: OLED is neither a pure mature-dividend name (Fundamentals + Valuation dominant) nor a pure high-growth story (Moat + Sentiment dominant) — it's a moderate-growth, moderate-yield industrial-IP licensor mid-cycle, so Fundamentals and Valuation carry the most weight, with Moat and Sentiment as corroborating context, per the framework's company-type weighting rule.

  • Fundamentals says: balance sheet is bulletproof (functionally debt-free, net cash positive), the dividend is safe and still growing (if decelerating), and the earnings/FCF decline is real but explainable by a cyclical smartphone-demand air pocket rather than margin or balance-sheet deterioration — gross margin held inside its guided band and operating margin actually improved YoY despite the revenue miss.
  • Valuation says: the stock passes the "cheap against its own band" test (not just off its high), the corrected current-year P/E (20.2x, not the vendor's 17.8x) is still reasonable for the quality on offer, and the FCF-multiple and Bogle approaches both land the fair-value range above today's price. Graham and DDM disagree sharply lower, but both are known-weak models for this business type and are down-weighted accordingly rather than ignored.
  • Moat says: the IP position is real and was just reinforced (Merck acquisition, OLEDX), but customer concentration (Samsung + LG ≈ 61.5% of the addressable volume) and the unresolved blue- emitter race are genuine, named vulnerabilities — this keeps conviction at 6.5 rather than higher.
  • Sentiment says: the drawdown reads as cyclical, and CEO/CLO open-market buying at a higher price than today is the strongest single piece of evidence for that read — but the buying was three months ago at $92–94, not confirmation of a bottom at $85 today.

Named tensions, not buried: 1. The royalty line's true-up volatility makes quarter-to-quarter "beats" and "misses" unreliable — Q2'26's headline beat leaned on a one-time catch-up; the real trend line is full-year guidance at the low end of range, which is a soft signal, not a strong one. 2. This is a close call on timing, not on quality. The bull case (cyclical trough, insider buying, reinforced moat, OLEDX optionality) and the bear case (concentrated customers, unsolved blue emitter, guidance still being cut) are both live. Nothing here demands urgency — Q3 earnings (Nov 5, 2026) is the next real data point on whether the deceleration has troughed.

Position sizing note: a quality-on-sale industrial-IP compounder with a clean balance sheet and a growing dividend fits well in either the Evergreen Compounders or Income-Yield Tomorrow sleeve — not a high-conviction concentrated bet, but a reasonable accumulate-in-tranches candidate into the $76–88 zone this analysis defines as entry.