Sivers Semiconductors

He announced a long position when Sivers was worth only around $130-140M, framing it as the next Lumentum that the market and institutions had overlooked. He explained that Sivers makes InP CW DFB lasers feeding future CPO and silicon-photonics architectures (via POET, Ayar and others) and argued the valuation made no sense given that the packaging companies buying its lasers were worth many times more. He pointed to a recent debt refinancing, a sizable revenue pipeline, Win Semiconductor capacity qualification, and a Lidar segment as supporting factors, while flagging execution, dilution and CPO-delay risks.

After a sharp same-day jump, he argued the market was starting to price in Sivers as the next photonics breakout. He drew a distinction between Lumentum, which benefits from today's pluggable-transceiver EML bottleneck, and Sivers, whose CW DFB lasers serve the upcoming co-packaged-optics shift. He positioned Sivers as the upstream light source feeding packagers like POET and Ayar, noting multi-source competition as the main risk but contending Sivers got in early with custom laser designs.

In a portfolio recap touting large year-to-date gains, he grouped Sivers with other silicon-photonics and CPO names he expected capital to rotate into. He noted Sivers was already up roughly 70-100% but argued it still had far to run as the market recognized the photonics supercycle. He framed his overall method as spotting structural AI supply-chain bottlenecks before institutions discover them.

In a weekly performance roundup of a follower-selected basket, he highlighted Sivers as one of the top three gainers with a roughly 35% one-week return. The mention was mainly a performance tally rather than new analysis, but it reinforced his continued endorsement of the name among his favored picks.

He argued that if Sivers were US-listed it should already trade above $2B versus its then ~$520M valuation, citing comparisons to peers and packagers trading at much richer multiples. He emphasized that Sivers sits in hyperscaler supply chains through Marvell, Jabil, O-Net and others rather than depending on a single customer, making it either an attractive acquisition target or a candidate to expand downstream like Lumentum did. He projected a path toward a $10B+ valuation over coming years once a US listing arrived and institutions completed due diligence.

Pushing back on criticism, he defended his practice of surfacing names to retail before institutions buy them in. He attributed Sivers' rise to new deals with Jabil and O-Net plus its prior, under-the-radar status as the laser supplier to Marvell's CPO program. He argued that without his attention, US institutions or strategic acquirers would eventually have bought the company and Swedish retail would have missed the upside.

In a post recapping his year-to-date winners and the criticism he received, he listed Sivers (as SIVEF) among names dismissed by skeptics as a pump-and-dump or meme stock. He framed the recurring abuse as evidence that his fundamentals-based ideas keep being vindicated, positioning himself as sharing information synthesis before institutions price it in.

Describing CPO and 1.6T as his most compelling theme, he listed Sivers first in a five-name supply-chain portfolio. He argued its laser revenue scales aggressively with Jabil, Marvell, Ayar and O-Net, and that there are very few laser suppliers worldwide, making this sub-$1B-cap company a CHIPS-Act-like chokepoint. He acknowledged multi-sourcing as the downside risk but contended there was a reason Jabil chose Sivers over larger rivals.

He laid out a market-mapping scenario in which Nvidia bought out Lumentum/Coherent allocation, leaving Sivers as one of the last remaining pure-play merchant CW laser suppliers for AMD, Marvell and Jabil CPO programs. He cited supporting clues such as Ayar dropping competitors from its site in favor of Sivers and GlobalFoundries listing only two public laser players. He stressed that hyperscaler suppliers would not pick a small Swedish laser company without reason and called it his high-conviction long.

Among three names he expected to triple, he ranked Sivers first, arguing it could reach $10B+ within a year from a ~$1.3B valuation as the bleeding edge for CPO lasers. He listed an expanding roster of potential customers spanning photonics, space/defense and consumer silicon photonics, calling the customer count remarkable and still growing. He argued Sivers could expand its addressable market downstream through IP acquisitions or vertical integration to eventually replicate Lumentum's much larger valuation.

He presented a detailed breakdown of Sivers' ownership, arguing the West now controlled the majority of shares as US retail (via Fidelity, Schwab, IBKR) and institutions (JP Morgan, Goldman, Morgan Stanley) built positions. He sarcastically thanked Swedish media for scaring off local investors, claiming this handed the float to US holders ahead of the CPO supercycle. He framed the progression as Swedish retail to US retail to gradual US institutional ownership.

Analyzing the 2025 annual report, he called it extremely bullish, highlighting that Sivers expanded beyond CPO into pluggable transceivers and was qualifying with multiple manufacturers. He emphasized statements about a coming CW laser shortage and locked-in Win Semiconductor capacity, plus signals of H2 2026 volume production and a Lidar customer ramping in Q4. He inferred Nokia as a likely customer and concluded the report pointed to broad TAM expansion and a strong CW-laser chokepoint position.

Responding to retail comments he viewed as lacking technical depth, he argued the photonics market is growing roughly tenfold over two years and is not a fleeting trade. He described Sivers as the core CW-laser chokepoint and IP holder for next-gen photonics, tethered to CPO and pluggable players, and a comfortable multi-year hold. He noted the main risk is how much overall TAM it captures and stressed that downstream expansion into full optical engines is what makes laser companies valuable, citing Lumentum's huge re-rating.

He argued that AI tools cause analysts to miss nuance, contending Sivers is both a chokepoint (leading CPO players route through it) and a bottleneck (it controls finished-laser output from allocated Win Semi capacity, analogous to Sandisk in NAND). He pushed back on the claim that CW lasers are interchangeable commodities, asserting Sivers' specific DFB architecture is uniquely suited to high-power, low-thermal laser arrays chosen by Ayar and Jabil. He also argued investors using trailing revenue metrics are wrong because volume ramps in 2027 and Sivers will pursue M&A to expand TAM like Lumentum.

Calling Sivers his most compelling CPO and photonics exposure, he stated he had not sold and would not sell a single share, planned to acquire more, and intended to support the company's M&A. He compiled a long list of recently discovered catalysts including Ayar/Amazon links, Jabil's faster-than-expected ramp, POET volume orders, Apple and defense ties, CHIPS Act funding, and a coming Nasdaq listing with index inclusion. He projected it could become the next $80B+ Lumentum from around $2.1B and said earnings would merely confirm his supply-chain mapping.

He restated that Sivers is both a chokepoint and bottleneck for CPO, explaining that CW lasers are bottlenecked because fabs are tied up with EMLs, and Sivers controls scarce finished-laser output. He argued the entire ASIC and merchant CPO ecosystem (Marvell, Ayar, Lightmatter and others) is designed around Sivers as primary or sole source, with no easy alternative for early generations, and that it serves as the GlobalFoundries reference laser design. He concluded Sivers could become the next ~$75B Lumentum as CPO scales in H2 2027 amid a TAM exploding toward $80-90B.

In a reflective post on his discretionary, inference-driven investing style, he grouped Sivers with names like AXTI, RPI and IQE as bets based on unstructured relationships the market had not yet recognized. He described his edge as guessing industry trends and supply-chain links from fragmented evidence before validation arrives, acknowledging it is high-conviction inference that could be wrong. Sivers appeared as an example of this method rather than receiving fresh thesis detail.

Recapping his core European holdings, he listed Sivers first and summarized it as the laser chokepoint over next-generation photonics, embedded across hyperscaler suppliers from Jabil to Ayar, expected to ramp in 2027 with qualifications priced in now. He stated Sivers had the highest upside of all his European names because of laser companies' ability to vertically integrate and acquire downstream to make their lasers more valuable, like Coherent and Lumentum.

In a candid review acknowledging some of his picks were red, he singled out Sivers as one of the top performers within his core photonics theme. He contrasted it with losing Taiwanese and Korean CPO-related positions hit by a CPO-delay report that Nvidia refuted, while noting his blended results stayed strongly positive. Sivers served here as a flagship winner amid a more balanced accounting of hits and misses.

Responding to a newcomer, he recounted that many of his ideas drew intense early backlash and that Sivers was widely dismissed as a meme stock with heavy criticism from Swedish media. He called it probably his most successful idea, now validated by institutional buying from the likes of Fidelity and JP Morgan and by formally announced partnerships with Jabil and GlobalFoundries. He framed the market's outcome, not the critics, as the final arbiter.

In a research note on a small Korean optical company (OE Solutions), he used Sivers as the contrasting, lower-risk benchmark. He argued OE is a far higher-risk, catch-up player with unanswered questions on customers and volume ramp, whereas Sivers is a major CPO player embedded in Ayar, Jabil, GlobalFoundries and other hyperscaler suppliers, with Win Semiconductor and others de-risking its volume ramp. The post reaffirmed his view of Sivers as a comparatively secure, well-positioned holding.