Serenity's letter to Sivers Photonics isn't about strategy. It's about narrative mismatch. A UK-based III-V compound semiconductor IDM holding two wafer fab allocations, six new pluggable customers, and an ELS partnership with O-Net — yet the market prices it like a Swedish turnaround story. The data doesn't support the discount.
The activist's core argument is simple: Sivers is telling the wrong story to the wrong investors. The company's communication strategy remains anchored to its Swedish listing heritage, focused on near-term cost containment. Meanwhile, the AI-driven optical communication cycle is the most significant demand inflection the industry has seen since the 2000s telecom boom. The market is rewarding companies that articulate their AI exposure clearly. Sivers is not one of them.
Here's what the technical teardown reveals.
The Capacity Signal
Two wafer fab allocations. The letter mentions this almost in passing, but it's the most consequential data point in the entire document. For a company of Sivers' scale — estimated revenue between $50-100 million — securing external foundry capacity is a structural shift. The implication is that Sivers is moving from a purely captive IDM model to a hybrid model that includes external manufacturing partners. This is consistent with the light-asset strategy that smaller compound semiconductor players typically adopt when demand outpaces internal capacity.
The supply bottleneck referenced in the letter confirms the market context: optical communication laser chips are in a seller's market. Industry-wide capacity utilization for III-V photonics fabs is running above 90%. ASPs are rising. The letter explicitly notes this. When an activist investor highlights ASP increases, it's not a courtesy — it's a signal that the fundamental supply-demand equation has shifted in the company's favor.
The CPO/ELS Position
The real story is CPO and ELS. Co-packaged optics is the most significant packaging architecture shift in optical communications since the advent of pluggable modules. The market is projected to reach $50-100 billion by 2028. ELS — external laser sources — is the critical enabling component for silicon photonics CPO routes. Sivers' partnership with O-Net positions the company in this supply chain.
The technology gap with Lumentum and Coherent narrows significantly in these emerging areas. In traditional pluggable laser chips, Sivers trails the leaders by 1-2 product cycles, roughly 2-4 years. In CPO lasers and ELS, the gap compresses to 1-1.5 years. This is the window. The company's DFB laser design and manufacturing IP — its core asset — is directly applicable to the ELS use case.
The Yield Question
III-V photonics yields typically run 60-85%, depending on device type. This is structurally lower than silicon CMOS, and it directly impacts unit economics. Laser chips account for 30-50% of optical module BOM cost. Yield improvement is the single most important operational variable for Sivers' margin trajectory. The company's current gross margin is estimated at 35-45%, with headroom to reach 40-50% in a supply-constrained environment. But this assumes the new capacity ramps without significant yield degradation.
The Competitive Landscape
Sivers sits in the second tier of optical laser chip suppliers, with less than 2% global market share. Lumentum holds roughly 25%, Coherent about 20%. The gap is substantial. But the competitive dynamics are shifting. Chinese manufacturers like Yuanjie Technology and Yunling Optoelectronics are rapidly closing the gap in mid-speed DFB/EML products. The threat is real — but it's concentrated in the 100G-400G segment. At 800G and above, and in CPO-specific laser solutions, the Chinese players still lack the design and manufacturing maturity required for high-volume deployment.
Customer concentration is another variable. The top five customers likely account for 60-80% of Sivers' revenue. This is typical for a company of its size, but it's a structural risk. The six new pluggable customers mentioned in the letter are a positive signal — they suggest the customer base is diversifying. But customer qualification cycles in optical communications run 12-18 months. The revenue impact of these new customers won't materialize until 2026.
The Geopolitical Arbitrage
Here's the contrarian angle. The bulls are right about the US market pivot — but not for the reasons they state. The US demand for non-Chinese laser chip suppliers isn't just about AI data center buildout. It's about supply chain diversification as a strategic imperative. US cloud providers and CPO manufacturers like Ayar Labs are actively seeking alternative suppliers to reduce concentration risk. Sivers, as a UK company, occupies a geographically neutral position. It can serve both US and Asian markets without triggering export control concerns.
The export control analysis is clean. III-V compound semiconductor photonics are not on the US BIS entity list. The company doesn't require EUV or advanced DUV lithography. MOCVD equipment from Veeco, AIXTRON, or AMEC is available through multiple channels. InP substrates from Sumitomo Electric and JX Metals are not subject to export restrictions. The supply chain risk profile is fundamentally different from advanced logic semiconductors.
The Execution Risk
But here's the problem. The gap between narrative and execution is where value gets destroyed. Two wafer fab allocations mean nothing if the capacity doesn't ramp on time. Six new pluggable customers mean nothing if they don't convert to volume orders. The estimated timeline for new capacity to reach full production is 2026. That's 12-18 months of execution risk.
The depreciation drag is real. New capacity will suppress gross margins by 3-5 percentage points during the ramp period. Break-even capacity utilization is 70-75%. If the company can't reach that threshold within two quarters of production start, the margin compression will be more severe than the market anticipates.
The Valuation Tension
The current valuation reflects optimism. Estimated PS ratio of 8-15x, EV/EBITDA of 30-50x — these are growth multiples for a company that hasn't yet demonstrated the revenue acceleration. ROIC is estimated at 2-5%, below WACC of 10-12%. The company is not yet creating economic value. The activist's intervention is essentially a bet that the growth story can be converted into financial performance within 12-24 months.
The risk scenario is straightforward. If capacity ramp disappoints, if yield improvement stalls, if customer conversion slows — the valuation has room to correct 30-50%. The opportunity scenario is equally clear. If the company successfully pivots to the US narrative, converts the six new customers, and ramps the new capacity, revenue could double from $50-100 million to $150-200 million within 18 months.
The Accountability Question
Volatility is just liquidity leaving the room. The same applies to narratives. When an activist investor forces a strategic conversation, the market is essentially asking: who is accountable for the gap between what this company could be and what it currently communicates?
The answer lies in the data. The capacity allocations are real. The customer pipeline is real. The CPO/ELS positioning is real. What's not yet proven is execution. Trust is a variable I refuse to define. But the market is pricing in a specific outcome, and the margin for error is thin.
The next 12 months will determine whether Sivers is a case study in activist-driven value creation or a cautionary tale about narrative exceeding substance. The signals to watch are concrete: quarterly revenue growth, gross margin trajectory, capacity ramp progress, and customer announcements. No commentary required. The numbers will do the talking.
