Wolfspeed Deep Dive: A vertically integrated silicon carbide pure play with scarce assets, but margins and per-share value are still underwater

Company Development History

Wolfspeed began with a group of North Carolina State University researchers in 1987 trying to turn silicon carbide from a laboratory material into a commercial semiconductor. It did not begin with the new-energy boom. The company was originally named Cree, and its purpose was direct: commercialize SiC wafers and devices. That origin shaped its later path. It grew out of materials science and crystal growth, not from a systems company reverse-engineering devices or from a distribution channel entering the market. Its early competitors looked more like traditional materials and compound-semiconductor companies than the automotive-grade power semiconductor giants familiar to investors today.

The company entered capital markets early. SEC historical filings show that Cree's Form SB-2 registration statement became effective in February 1993. The SEC News Digest recorded that the offering that year comprised 1.75 million shares and raised about 16 million USD. Early capital markets understood it as a typical materials and device technology company, one pushing SiC from academic breakthrough to manufacturing, not today's "EV/AI power chain" name. The story then was optoelectronics, LEDs, and wide-bandgap materials. It stood a full industrial migration away from today's traction inverters and data centers.

The first stage that truly changed the company's fate was its success in the LED era. Cree spent years scaling blue and green LEDs, lighting, and SiC materials together. That gave the company a base in manufacturing, scale, and capex management, and later gave it the ability to cut legacy businesses and bet on heavier SiC power semiconductors. The second stage came after 2017, when management clearly narrowed resources toward the Wolfspeed business, completed the sale of the LED business in 2021, and formally renamed the company Wolfspeed. This was critical: purity improved, and the capital-market valuation framework changed. The third stage was the all-in bet on 200mm and vertical integration in 2022-2023. The Mohawk Valley 200mm SiC device fab started production, the Siler City materials base launched, the Saarland project in Germany was announced, and the RF business was sold to MACOM in 2023. The company was remade from a diversified semiconductor company into one fully tied to SiC. The fourth stage was the sharp reversal from 2024 to 2025: EV demand slowed, customers destocked, project timing slipped, fixed-cost pressure rose, the CHIPS PMT remained unresolved, and the company entered a prearranged restructuring in 2025. The fifth stage is now: debt has been sharply reduced, old shares have been canceled, new shares trade again, and the story has shifted from "capacity-expansion growth" to "operating repair plus new-application validation."

Several milestones deserve separate attention. In April 2022, the Mohawk Valley fab opened. That was the moment Wolfspeed pushed itself from concept stock into manufacturing stock. It was no longer simply selling materials and talking about the future. It began carrying the depreciation, yield, automotive qualification, and utilization burden of a 200mm fab. In September 2022, the Siler City project was announced, and management pushed the supply logic further: if it wanted to build world-class 200mm SiC, it needed a larger materials foundation alongside the device fab. In October 2024, the preliminary CHIPS Act memorandum appeared, which was supposed to serve as an external funding patch for this capital story. But a memorandum is not a final grant, and by the time the situation deteriorated in 2025, it had not become cash. For the market, the meaning of this sequence was harsh: the company placed its heaviest capital bets first, then ran into a dual shock in demand and financing conditions.

The restructuring itself also changed the company. On September 29, 2025, Wolfspeed completed its Chapter 11 exit. Old shares were canceled, new shares were issued, the board was reorganized, Anthony Abate became chairman, and the company converted from a North Carolina corporation into a Delaware corporation. The 10-Q also explicitly noted that Successor and Predecessor financial statements are not comparable. For researchers, this means two things. First, it is lazy to use "what this company's market cap used to be" or "how many times PS it traded at the historical peak." Second, every per-share metric has to be rechecked, especially the post-restructuring convertible notes, warrants, prepaid warrants, and equity incentive reserves. The hardest part of studying WOLF today is understanding the share count before understanding SiC.

Financial Longitudinal Review

Breaking down the financials from the late 2010s to today, Wolfspeed has gone through a reconstruction in which "revenue logic and cost logic became misaligned." After selling LED and RF, revenue shifted from diversified businesses to a heavy concentration in SiC materials and power products. The benefit was a cleaner story. The cost was the loss of business offsets. When EV demand was strong, this purity deserved a premium. Once EV destocking and project delays arrived together, it became an amplifier. FY2026 Q2 product-line disclosures showed Successor-period revenue of 118.3 million USD from power products and 50.2 million USD from materials products. By FY2026 Q3, total revenue had fallen further to 150.2 million USD, showing that order recovery remained weak.

Profit quality is the clearer problem. FY2026 Q2 GAAP gross margin was -46%, improving to -27% in Q3 but still deeply negative. Adjusted EBITDA improved from -81.8 million USD in Q2 to -62.0 million USD in Q3. The direction was right, but the magnitude was still insufficient. The most important business explanation here is "fixed costs are too high and utilization is too low," rather than simply "losses are large." A 200mm SiC fab can bring cost advantages at high utilization. At low utilization, it first turns depreciation and labor into massive negative gross margin. In March 2025, the company gave a key threshold: after fab simplification, the closure of the North Carolina device line, and further cost cuts, adjusted EBITDA break-even would be roughly 800 million USD of annual revenue. Looking back through this lens, FY2026 Q3 quarterly revenue of 150.2 million USD annualizes to only about 600 million USD, clearly below management's own break-even definition.

Cash flow is more honest than the income statement. During the FY2026 Q2 Successor period, operating cash flow was -42.6 million USD. In FY2026 Q3, the quarterly outflow widened to -84.0 million USD. The good news is that capital spending has been cut sharply. In Q2, the company said annualized operating expenses were down about 200 million USD from Q2 FY2025, capital expenditures were down more than 90% year over year, FY2026 capex had been reduced to 150-200 million USD, and FY2027 capex would fall further to 30-50 million USD. The bad news is that even with capex reduced to this level, operating cash flow is hard to turn positive while gross margin remains negative. In other words, the company's cash problem today is whether existing capacity can be filled, not whether it should keep pursuing another major expansion.

The balance sheet is far safer after restructuring than it was a year ago, but it is not relaxed. As of March 29, 2026, the company had about 1.2 billion USD in cash, cash equivalents, and short-term investments. In the same quarter, it issued 379 million USD of 1.5L convertible notes and about 96.9 million USD of common shares and prepaid warrants to redeem about 475.9 million USD of first-lien secured debt. Total debt fell by about 97 million USD, and annual interest was reduced by another roughly 62 million USD. At the same time, the post-restructuring new first-lien secured notes require minimum liquidity of at least 350 million USD at month-end. Put these three numbers together and the meaning is straightforward: 1.2 billion USD on the balance sheet keeps the company away from the edge of immediate refinancing, but the cash that can truly be consumed freely is what remains after working capital, safety buffer, and covenant requirements, not the full 1.2 billion USD.

My judgment on the cash runway is conservative. Based on the operating cash flow pace in FY2026 Q2 and Q3, plus the current low capex level and still-negative gross margin, Wolfspeed's "comfortable runway" looks more like 18-24 months than the three-to-four-year safety illusion investors may get from hearing "1.2 billion USD of cash." The company will not run out of money immediately. The reason is that its financing tools and minimum-liquidity clauses both remind investors that the timing of a path to positive gross margin matters more than total cash.

Share Price and Valuation History

Wolfspeed's valuation history is almost a lesson in how capital markets price "scarce manufacturing assets" and then quickly reverse when delivery is late. After the company renamed itself Wolfspeed in 2021 and completed the sale of the LED business, the market began treating it as one of the purest U.S. names in the SiC value chain. The Mohawk Valley 200mm fab opened in 2022, the Siler City plan landed in 2022, and the company announced the German project and divested RF in 2023. The narrative upgraded from "pure SiC" to "first mover in the global 200mm SiC ecosystem." During that period, valuation borrowed against many future years of revenue and margin expansion.

The subsequent re-rating was just as astonishing. CEO Gregg Lowe left in November 2024, and the CFO left in April 2025. In March, the stock plunged about 48% on weak demand and financing uncertainty, touching its lowest level since 1998. By June 2025, when the prearranged restructuring plan landed, the market finally acknowledged that this was both a capital-structure imbalance and a commercial-delivery timing imbalance, not a normal cyclical downturn. After the September restructuring exit, the new shares also rallied sharply on float reset, debt reduction, and event-driven flows. For investors, the main risk is confusing the old share price decline with cheapness in the new shares, not volatility itself. Between the two sits a full equity wipeout and redistribution.

As of the latest available quote before the research base date, WOLF's market cap was about 2.2 billion USD. Looking only at basic shares, and combining roughly 2.0 billion USD of post-restructuring debt with about 1.2 billion USD of cash, headline EV is around 3.0 billion USD. For a manufacturing company still posting negative gross margin and negative EBITDA, this already makes PE and EV/EBITDA largely unusable at this stage. More importantly, headline EV may still understate the real dilution pressure at the per-share level. The S-1 filed on June 9, 2026 showed 51.972 million shares outstanding as of May 31, 2026, excluding 2.000 million prepaid warrants, 4.944 million Renesas warrants, about 22.026 million shares underlying non-Renesas 2L convertible notes, 11.096 million shares underlying Renesas 2L convertible notes, 18.822 million shares underlying 1.5L convertible notes, and later RSU, PSU, and incentive reserves. Static addition of these disclosures gives me a potential share count near 123 million shares, about 2.4 times the basic share count at that time. This calculation is not a forecast that all dilution happens in the near term. It is enough to show that researching Wolfspeed cannot stop at the visible 2.2 billion USD market cap.

Business Model and Moat

Wolfspeed's business model is simple, but extremely hard to execute. It has two interlocking revenue lines: materials products sell SiC substrates and epitaxy, while power products sell discrete devices, modules, and die. The materials side determines upstream supply control, and the device side determines the speed of value capture. The real strategic bet is to use vertical integration to fully absorb 200mm economies of scale, letting materials and devices reinforce each other on the same learning curve. In theory, this creates three benefits: more stable supply, faster process iteration, and lower unit cost. In practice, it also creates three costs: higher upfront capex, higher fixed costs, and a bigger penalty for demand forecasting errors.

I see three real moats. The first is technology and process accumulation. Cree/Wolfspeed has not been doing SiC for only the past couple of years. The company itself was one of the early pioneers of this market, and it has built long-running know-how on both the materials and device sides. The second is the scarcity of manufacturing assets. The Mohawk Valley 200mm SiC fab and Siler City materials base are not a slide deck. They have been physically built. The third is supply control from vertical integration. For customers focused on automotive qualification and long-term supply, this still has value even in a downturn.

But I would not readily accept several moats often used in market promotion. First, being a "pure play" is not a natural moat. A pure play has greater upside in an upcycle and greater fragility in a downcycle because it has no other business to offset weakness. Second, "200mm first mover" is not an automatic moat. If utilization is insufficient, first-mover status shows up first in negative gross margin. Only when yield, qualification, and demand ramp together does first mover translate into unit-cost advantage. Finally, the "new AI data center story" cannot be treated as an already delivered moat. In FY2026 Q2 and Q3, the company did disclose sequential growth in AI data center revenue, but it also described it as a "modest but expanding component" of the business, which means the base remains small. By comparison, onsemi's 2025 AI data center revenue had already exceeded 250 million USD, and Infineon even raised its 2026 target for AI power supply revenue to about 1.5 billion euros. Wolfspeed is one of the later entrants on this line, not a lone leader.

On governance, the problems of the past two years are very specific. The CEO changed in November 2024, the CFO left in April 2025, Robert Feurle became CEO in March 2025, and after the September restructuring, the board was reorganized again with Anthony Abate becoming chairman. Management backgrounds are not weak. Feurle came from large semiconductor companies, and CFO Gregor van Issum has experience in complex manufacturing and financing. But investors cannot ignore that before and around the restructuring, the company's capital allocation results were poor: capacity expansion was mismatched with end demand, external subsidies did not arrive on time, the debt structure was too heavy, and technical assets ended up at the courthouse door. Whether the post-restructuring management team is steadier still needs time to prove.

Industry and Cycle Analysis

SiC is in an industry experiencing long-term substitution and short-term adjustment at the same time. It is not a declining industry. The long-term drivers are clear: demand for high-voltage, high-efficiency, compact power conversion keeps rising. EV traction inverters, industrial motors, energy storage, solar, grid upgrades, and parts of the high-voltage data center power chain all give SiC clear application scenarios. In its 2025-2026 external materials, Infineon still lists BEV production growth, rising non-drive semiconductor BOM, and AI data center power as core drivers for power semiconductor growth. ST also continues to treat SiC as one of its key engines before 2030. The industry direction is not wrong. What was wrong was the market's earlier assumption that the intervening cycles and cost swings would be mild.

Wolfspeed sits under three overlapping cycles, not one. The first is the semiconductor inventory cycle: EV destocking and cautious ordering in 2024-2025 directly hit the company's revenue. The second is the capex cycle: the company pushed heavy-asset expansion deep into the cycle at the high point and then faced low utilization when demand weakened. The third is the technology adoption cycle: 200mm SiC and AI data center HVDC are directionally right, but the time lag from "right roadmap" to "financially profitable" is much longer. Wolfspeed's most fragile variable is utilization after these three are multiplied together, not a single ASP or a single customer.

Policy and geopolitics matter as well. In October 2024, Wolfspeed signed a non-binding PMT under the CHIPS Act with the U.S. Department of Commerce, potentially receiving up to 750 million USD of direct support. At the same time, the company also relied on Section 48D tax refunds to improve cash flow. By March 2025, the company confirmed that it had received 192.1 million USD of 48D tax refunds and expected to receive more than 600 million USD of cash refunds in FY2026. Q2 then disclosed that about 700 million USD of tax refunds had already been received early. But direct grants and tax credits are different things: the latter arrived, while the former did not turn into cash during financial deterioration and restructuring. For Wolfspeed, policy is part of the financing equation, not an extra surprise. If it fails to arrive, the impact flows directly into the debt and equity structure.

Peer Comparison

Wolfspeed faces a classic "plenty of competitors" situation, but its real rivals do not come from only one box. For SiC devices and substrates, the most relevant comparisons are onsemi, Infineon, STMicro, and ROHM. For the AI data center 800V power chain, Navitas is an unavoidable reference. They represent four different models.

onsemi is a "power platform company with a profit base." It makes SiC and has broader automotive and industrial product lines. Its 2025 AI data center revenue already exceeded 250 million USD, its 2024 AI data center and aerospace and defense revenue grew by more than 40% and 50% year over year, respectively, and it has already partnered with NVIDIA on 800 VDC data center power solutions. Customers choose onsemi because it can deliver a full power-chain product set with more stable supply and more mature industrialization, beyond any single SiC device. For Wolfspeed, this kind of competitor is the hardest to handle because it will not be dragged down by a single business during an EV trough.

Infineon is a "system-level giant in power semiconductors." Its AI data center power supply business target for 2026 has been raised to about 1.5 billion euros, and its annual revenue scale and cash flow are not comparable to Wolfspeed's. When Infineon talks about SiC, it talks about a combined portfolio: Si, GaN, and SiC across materials, taking share in vehicles while also pursuing structural growth in AI power, industrial, and grid applications. Customers often choose Infineon for the most stable option rather than the purest one. For Wolfspeed, Infineon's presence points to a harsh reality: in AI data centers, the winner is whoever first enters the customer's design, qualification, and volume-production cadence, not whoever says the words first.

ST and ROHM look like two different regional powerhouses. ST has invested in SiC for more than two decades, emphasizes full supply-chain control, and treats SiC as an important growth engine before 2030. ROHM has deep roots in Japan's power semiconductor system, and in 2026 it became involved in Denso's acquisition proposal and subsequent industrial synergy discussions, which shows its strategic position in EV and data center power devices being repriced. Why would customers choose ST or ROHM? A common answer is that beyond the chip itself, they offer greater certainty in regional automotive supply chains and long-term customer relationships. Wolfspeed's strengths are purity and materials. Their strengths are platforms and customer organizations.

Navitas is entirely different. It is much smaller than Wolfspeed in revenue and profitability, but in the AI data center 800V HVDC narrative, it has actually run earlier and louder: in May 2025, it announced joint development with NVIDIA for next-generation 800V HVDC architecture, and in 2026 it released 800V-6V and 10kW DC-DC demonstration platforms in succession. The problem is also clear: Navitas 2025 Q4 revenue was only 7.3 million USD, and 2026 Q1 revenue was only 8.6 million USD. This comparison is valuable for Wolfspeed because it reminds investors that "collaboration," "demo boards," and "platform validation" in the AI data center power chain are still far from scaled revenue. Wolfspeed is also talking about 10kV SiC, TOLT packaging, and a dedicated team today, but revenue delivery on this line cannot be treated as already achieved simply because the story is advanced.

From an ecosystem perspective, Wolfspeed remains a niche but important upstream manufacturing and device player. It is directly trying to capture the profit pool in high-voltage, high-efficiency SiC power semiconductors, especially markets more sensitive to material quality, long-term supply, and high-voltage performance. Its biggest weakness is the absence of a larger cash-cow business as a buffer. If the industry enters price competition, customers delay adoption, or competitors use stronger channels and system solutions to take orders, Wolfspeed's position will be much weaker than that of diversified players. It is difficult to replace in parts of the industry, while still far from irreplaceable.

Current Fundamentals and Long-Short Debate

Looking at the most recent quarters, the conclusion is simple: the company is "repairing," but it has not "turned around." In FY2026 Q2, revenue was 168.5 million USD and AI data center revenue rose about 50% sequentially, but GAAP gross margin remained -46% and operating cash flow was -42.6 million USD. In FY2026 Q3, revenue fell to 150.2 million USD, AI data center revenue grew another roughly 30%, gross margin improved to -27%, and Adjusted EBITDA improved to -62.0 million USD, but operating cash flow still had an 84.0 million USD outflow. Q4 guidance remains 140-160 million USD, with gross margin still negative. If the market only looks at "two consecutive quarters of AI data center growth," the story looks better. If it also looks at negative gross margin and negative operating cash flow, the company is still at the bottom.

The market most wants to trade "asset repair plus a new AI power script." The company is also actively pushing in this direction: in March 2026, it completed strategic refinancing and cut the highest-cost first-lien debt by 43%; in the May earnings release, it highlighted the 10kV SiC MOSFET, TOLT package, 300mm platform, and AI data center revenue growth; in June, it announced a dedicated team for AI infrastructure and signed a memorandum of cooperation with GE Aerospace for high-voltage SiC. Capital markets like this semantic shift from "EV victim" to "AI high-voltage power infrastructure beneficiary" because it naturally creates greater valuation imagination.

But there are at least three gaps between real fundamentals and market narrative. The first is the base. The company has never disclosed the absolute size of AI data center revenue, only emphasizing sequential growth and a "modest but expanding component" of the business. The second is competition. onsemi, Infineon, and Navitas are all talking about the same 800V DC or high-voltage data center power narrative, and some players already have larger revenue scale or stronger customer lock-in. The third is the earnings model. Even if AI orders begin, incremental revenue may not quickly reach the income statement unless Wolfspeed simultaneously raises Mohawk Valley utilization, improves yield, and fully absorbs the materials side after the Durham transition.

Bulls have four strongest pieces of evidence. First, debt and interest have genuinely fallen sharply after Chapter 11. This is an effective capital-structure change, not a story. Second, the large capex cycle is mostly over, the Durham 150mm device line has been closed, future capex will fall significantly, and operating leverage should theoretically begin to tilt toward shareholders. Third, AI data centers, industrial high voltage, grid, and defense applications open demand pools for SiC outside EVs. Fourth, the company does hold scarce 200mm vertically integrated assets. Bears have equally clear evidence. First, Q3 and Q4 remain in negative gross-margin territory, showing that the fixed-cost problem has not been solved. Second, customers prepared second sources during bankruptcy and some orders were pulled forward before the Durham line closure, which management itself wrote into the FY2026 Q2 outlook. Third, there are many post-restructuring dilution tools, and the June 2026 S-1 put 24.07 million resale shares on the table, with the company receiving none of those proceeds. Fourth, the AI narrative is hot, but absolute revenue scale remains unproven.

Valuation Analysis

For Wolfspeed, PE, Forward PE, and EV/EBITDA are secondary tools. In Q3 FY2026, the company reported a net loss of 119.9 million USD, Adjusted EBITDA of -62.0 million USD, and Q4 guidance still calling for negative gross margin. Discussing whether it is "expensive" through earnings multiples would be distorted. A better framework at this stage is to first look through cash flow, then consider EV/Sales, heavy-asset replacement cost, and per-share value under dilution constraints.

After cash-flow look-through, the conclusion is more conservative than the income statement. FY2026 Q2 and Q3 both had negative operating cash flow, and the company still needs to maintain some capital expenditure and minimum liquidity of 350 million USD. The boundary between maintenance capex and growth capex will become much clearer than in the past after the major expansion cycle, but at least in FY2026, "owner earnings" remain clearly negative. Therefore, headline PE is unusable, and FCF yield has no practical reference value. Current valuation is either built on the assumption that revenue and gross margin will materially recover over the next 2-3 years, or on the assumption that these manufacturing assets can be sold at a higher price in the future. It is not built on current-period earnings.

From a relative-valuation perspective, headline EV/Sales does not prove Wolfspeed is cheap. Using the latest available market cap of about 2.2 billion USD, post-restructuring debt of about 2.0 billion USD, and cash of about 1.2 billion USD, headline EV is roughly 3.0 billion USD. Against the current revenue run-rate of about 600 million USD, EV/Sales is about 5x. That number may not look absurd in isolation, because power semiconductor leaders such as onsemi, ST, and Infineon can also trade at several times sales when market conditions are strong and AI expectations provide support. But there are two limits. First, they are profitable. Wolfspeed is not. Second, Wolfspeed's share count is far from stable. Based on the potential conversion and incentive shares listed in the S-1, static potential share count could approach 123 million shares, meaning the per-share denominator risk is higher than headline EV suggests.

Valuation Scenarios

The table below is a research framework built from public information, not investment advice. I base it on two premises: first, FY2026 is still an early repair phase; second, valuation has the right to move from an "asset story" to an "operating story" only when revenue approaches or exceeds the company's own 800 million USD EBITDA break-even threshold.

On this basis, my three price signals are: bear range 18-24 USD, base range 28-38 USD, and bull range 48-60 USD. Considering the complexity of the share count and convertible notes, and the new June 2026 registration of 24.07 million resale shares, I leave a wider "clearly overvalued line" buffer in the final conclusion. I do not define the area around 48 USD directly as a bubble. Instead, I set 58-68 USD as the clearer overvaluation band. This is conservative, but it better fits the uncertainty of a newly restructured company.

Expectation Gap and Margin of Safety Review

What expectations does the market currently imply? I see at least three: first, EV demand will recover moderately over the next few quarters; second, AI data centers will contribute more revenue in 2026-2027 than the current "modest" level; third, the capital structure has been repaired enough to wait for the first two items to arrive. The most likely sources of expectation gaps are three hard metrics, not any single product launch: whether quarterly revenue can return above 180 million USD, whether GAAP gross margin can keep moving toward 0%, and whether operating cash flow can narrow to within -50 million USD per quarter. If any one of these remains stuck, the market will reprice the "turnaround stock" back into a "high-risk event-driven stock."

Under my bear scenario above, the current price offers no discount to conservative value. It carries a clear premium, so the margin of safety is zero. The most fragile assumption across the three scenarios is a basic one, not a requirement for AI to explode: Mohawk Valley and the materials side must push revenue across the company's own break-even threshold as soon as possible. If this assumption is haircut by 30%, meaning revenue recovery is materially slower than expected, the base range would shift toward the bear range, and reasonable per-share value would be closer to 22-30 USD rather than 28-38 USD. By this standard, the current price looks like a prepayment for "smooth repair." My conclusion on margin-of-safety sufficiency is: none.

Risk Analysis

I see five risks most likely to cause permanent capital loss.

The first is demand repair failing to keep pace with fixed-cost pressure. Probability is medium-high, and impact is high. Wolfspeed's most dangerous point is carrying the depreciation and labor of a 200mm fab when revenue is weak. Q2 management explicitly mentioned that some customers pulled orders forward before the Durham line closure, others prepared second sources during the company's bankruptcy, and EV demand was weak. Q3 and Q4 guidance then showed this gap had not been filled. The observable indicators are whether quarterly revenue returns above 180 million USD and whether gross margin keeps improving. If not, low utilization will keep pressing profits and cash flow down.

The second is dilution risk exceeding what headline market cap suggests. Probability is high, and impact is high. The June 2026 S-1 already registered 24.07 million shares for potential sale, and the company receives none of the proceeds from those sales. The same filing also listed many convertible notes, warrants, and incentive reserves. As long as the share price stays near the current level, multiple instruments are economically "in the money." For shareholders, this means that even if enterprise value improves, per-share value may keep transferring away. The observable indicators are new S-1/S-3 filings, conversion activity, Renesas warrants, and reduction arrangements by other major holders.

The third is AI data centers taking longer than imagined to move from narrative to revenue. Probability is medium, and impact is high. Wolfspeed is now talking about 10kV SiC, TOLT packaging, a dedicated team, and high-voltage infrastructure, but onsemi, Infineon, and Navitas are telling similar stories, and some are already closer to customer design centers. Wolfspeed also has not disclosed absolute AI data center revenue, implying scale is still early. The observable indicator is whether the company begins disclosing more specific customers, platforms, design wins, or absolute revenue mix. If the data remains at the level of "sequential growth and huge potential," the market will gradually discount this narrative.

The fourth is liquidity constraints becoming the core conflict again before operating repair arrives. Probability is medium, and impact is high. The company sounds well funded with 1.2 billion USD of cash, but first-lien notes require minimum liquidity of 350 million USD, operating cash flow is still negative, and Q4 gross margin is still guided negative. If revenue does not improve after several quarters and new external pressure appears, the market will again worry about financing costs and further share issuance. The observable indicators are cash balance, operating cash flow, and any new high-cost debt.

The fifth is China's SiC capacity expansion and price competition lowering the long-term gross-margin ceiling. Probability is medium, and impact is medium-high. This risk may be absent from Wolfspeed's own announcements; it will gradually appear through materials ASP, customer multi-sourcing, and competitors' delivery cadence. Wolfspeed's vertical integration can protect part of its supply security, with no guarantee that it protects all of its gross margin. If the industry becomes "materials-side price war plus device-side platform competition," first-mover advantage will also weaken. This risk is hard to quantify in the short term, but it directly affects the valuation center.

Catalysts and Tracking Indicators

The most important positive catalyst is simultaneous improvement in three hard metrics: revenue returning above 180 million USD, gross margin continuing to approach zero, and operating cash flow narrowing clearly. Only after that come hard evidence in new applications, such as disclosure of absolute AI data center revenue, major customer platform wins, continuous volume growth in industrial high-voltage devices, and further replacement of current high-interest debt with lower-cost capital. Negative catalysts are equally clear: another guidance cut, stalled gross-margin improvement, another deterioration in operating cash flow, concentrated selling driven by S-1/S-3 filings, or a continued delay in EV recovery.

Tracking Dashboard

These indicators matter because they correspond to four main lines: operating repair, cost absorption, cash runway, and per-share dilution. Wolfspeed is not a company where tracking revenue is enough. For it, gross margin and share-count changes often matter more for shareholder returns than revenue itself. Tracking sources should mainly be 10-Q, 8-K, quarterly press releases, and registration statements, supplemented where necessary by peer earnings as evidence of the demand environment.

Cross-Sectional and Longitudinal Synthesis

Longitudinally, Wolfspeed has genuinely proven two capabilities. First, it can industrialize a very difficult materials technology. Second, it is willing to place the heaviest capacity bets before an industry inflection. The first point shows this company is more than an empty story. The second shows it has strong industrial judgment and execution impulse. But the past few years have also proven another point: a correct technology roadmap alone cannot offset a wrong capital structure. When Wolfspeed succeeded, industry tailwinds, technology leadership, and capital markets' willingness to pay for scarcity stood together. When it failed, an EV cycle downturn, exposed fixed costs, and a colder financing environment also stood together. The company's path to this point was not luck; the mismatch between macro timing and industry cadence did magnify its wounds.

Cross-sectionally, Wolfspeed's real advantages versus peers are purity, materials-side accumulation, and 200mm vertically integrated assets. Its real weaknesses are the absence of diversified-business buffers, the absence of a mature cash cow, and the absence of large, already delivered AI data center revenue. This weakness is a structural result of the business model and cannot be fixed by short-term public relations. When the industry is strong, this structure gives it the most leverage. When the industry is weak, the same structure makes it the easiest to chase with cash-flow and share-count problems. Today's valuation is paying in advance for the possibility that the 200mm cost curve eventually works and that AI/industrial high-voltage demand fills the EV gap. It is not rewarding success that has already happened. The issue is that this possibility has not yet been verified simultaneously by revenue and gross margin.

I think the market is most likely to misjudge two things. The first is treating "debt reduction" as "automatic shareholder value recovery." In reality, restructuring improved the allocation of enterprise value, but per-share value still has to pass through dilution and conversion tools. The second is treating "AI data center product launches" as "a rewritten revenue curve." Navitas has already provided a useful reference: a company can capture the NVIDIA narrative first and then wait slowly for revenue to arrive. Wolfspeed is also walking this path today, but its assets are heavier, cash burn is higher, and room for error is lower.

The most important variable over the next 1 year is whether gross margin can continue repairing from deeply negative levels toward break-even. The most important variable over the next 3 years is whether EV demand recovery or new AI/industrial high-voltage applications feed Mohawk Valley first. The most important variable over the next 5 years is whether 200mm first-mover status becomes a real cost advantage or is diluted by broader industry expansion and price competition. Under what conditions would Wolfspeed become a better investment? The answer is plain: when the share price returns to a level that provides a margin of safety, and the company has proven over 2-3 quarters that the direction of gross margin and operating cash flow is no longer choppy. Under what conditions must the original view be overturned? If revenue falls below 140 million USD for two consecutive quarters, gross margin remains below -20%, operating cash flow worsens again, and the company starts another round of high-cost financing or share issuance, then this is a longer capital-consumption battle rather than a delayed turnaround.

Bull and Bear Cases

There are four bull points. First, debt and interest expense have fallen meaningfully after Chapter 11, and enterprise-level survivability has clearly improved. Second, the heaviest expansion cycle is basically over, and capex pressure over the next few years will be materially lower than in the past. Third, 200mm vertically integrated manufacturing assets have real scarcity. Fourth, AI data centers, industrial high voltage, and grid modernization provide new demand pools for SiC beyond EVs.

There are also four bear points. First, Q3 gross margin was still -27%, Q4 still guides for negative gross margin, and an operating turnaround has not happened. Second, customer second sourcing and order pull-forward already appeared in management's FY2026 Q2 commentary, showing that the order base is not firm. Third, post-restructuring convertible notes, warrants, prepaid warrants, and incentive reserves create high share-dilution risk. Fourth, the AI data center narrative is not an empty lane, with onsemi, Infineon, and Navitas all fighting for position on the same line.

Pre-mortem

If this investment loses 50% three years from now, I think the two most realistic scripts are as follows.

The first script is that from the second half of 2026 into 2027, EV demand continues to recover later than expected, Chinese supply-chain pricing moves lower and keeps pressuring SiC ASP, and Mohawk Valley utilization never rises enough. Company revenue stays stuck between 600 million and 800 million USD annualized for a long time, gross margin improves from negative only to the low single digits, and operating cash flow keeps consuming cash. Because of minimum liquidity constraints and repeated capital-market windows, the company is forced into another dilutive financing. A fall from the current 40-plus USD share price back toward 20 USD would not require a catastrophic assumption. The damage in this scenario comes from "revenue not high enough, repair not fast enough, and share count getting larger," not technology failure.

The second script is that from 2026 to 2028, AI data center 800V HVDC architecture does progress, but the players that win scaled orders are those with deeper system binding and more complete product stacks. onsemi keeps advancing through the NVIDIA ecosystem, Infineon accelerates through its existing power-supply share, and Navitas captures higher-density local segments. Wolfspeed wins technical qualifications and some high-voltage projects, but does not unlock enough absolute revenue. The market then shifts from "it can also get AI exposure" to "it can merely participate in AI." When event-driven high expectations compress and fundamentals have not delivered strong verification, a 50% share price decline is not excessive.

Final Research Conclusion

The most attractive thing about Wolfspeed today is that what it owns is genuinely scarce: SiC materials, 200mm device manufacturing, and finally a somewhat lighter debt burden after restructuring. The most dangerous thing also comes from these scarce assets: they are too heavy, too cash-burning, and too dependent on utilization to tolerate continued delays in delivery. I do not doubt that the company still has enterprise value. For the current share price, I do not see enough margin of safety. It can certainly go up. But if one treats it as an investment rather than a trade, it needs a better price and more operating proof.

My biggest concern is that after Chapter 11 ended, the market may again start ignoring that "per-share value is still being rewritten by the capital structure," rather than focusing on the fact that Chapter 11 already happened. The spring 2026 refinancing, the June S-1, and the large pool of potential conversions all remind investors that Wolfspeed still looks more like an event-driven asset than a standardized fundamental stock. If revenue, gross margin, and cash flow all verify improvement over the next few quarters, I would be willing to raise my view. Until then, waiting matters more than chasing price.

【Company Profile Score】

• Fundamental quality: Medium

• Growth: Medium

• Moat: Medium

• Financial stability: Medium

• Management credibility: Medium

• Valuation appeal: Low

• Risk level: High

• Suitable investor type: Not suitable for ordinary investors

【Investment Rating】

• Rating: Watch

• One-line investment thesis: The assets are scarce, but margins and per-share value have not truly climbed out of the hole.

• Three price signals: 【Ideal/Fair Buy Price】18-24 USD

• Rationale: This is close to the bear scenario and only begins to provide enough event and dilution safety buffer relative to the current price.

• Acceptable holding price: 28-38 USD

• Clearly overvalued price: 58-68 USD

• Current price classification: Outside the three ranges

• Is it worth waiting for a better price: Yes. A buying trigger would require a price below 24 USD plus continued gross-margin improvement or narrowing operating cash flow, with at least one of the two operating conditions met. Otherwise, cheap without improvement could still be a value trap.

• Target holding period: 1-3 years

• Expected annualized return: bear about -23% to -16%; base about -15% to -6%; bull about +2% to +11%

• Maximum loss risk: More than 50%. The trigger would be further delay in EV recovery, weaker-than-expected AI delivery, continued deterioration in operating cash flow, and resulting further dilution.

• Signals that would trigger reassessment: Revenue returning above 180 million USD for two consecutive quarters

• GAAP gross margin above -10% for two consecutive quarters

• Quarterly operating cash flow narrowing to within -50 million USD

• No new high-cost debt or large-scale primary equity

• The company begins disclosing more specific AI data center revenue or design-win scale

【Valuation Range】

• current: 45.54 (latest available quote after the 2026-06-11 close)

• bear (conservative · ideal buy zone): [18, 24]

• base (reasonable · acceptable holding zone): [28, 38]

• bull (optimistic · above the clear overvaluation line): [58, 68]

Key Data Tables

Post-Restructuring Operating and Cash Flow Snapshot

Data note: What improved in Q2 and Q3 was the loss magnitude, not the profitability state. Revenue and gross margin are still in a "low-level repair" phase rather than a "normalization" phase. This is why the current share price is better understood through scenarios than through simple multiples.

Post-Restructuring Share Count and Dilution Pressure

This table is a reminder, not a claim that all dilution will definitely occur in the near term: Wolfspeed's per-share value analysis must keep the potential denominator in mind. For a stock that has just completed bankruptcy restructuring and whose share price is already above multiple conversion/exercise prices, this issue is one of the main questions, not a footnote.

Research Uncertainties

This report still has four blind spots that require continued updating.

• The company has not disclosed the absolute revenue scale of the AI data center business. External research can confirm direction, but it is hard to confirm size.

• There are many post-restructuring convertible notes, warrants, and prepaid warrants. Although registration filings disclose potential shares, the actual settlement method, timing, and holder behavior remain uncertain.

• Public, timely, and comparable data are limited for SiC industry price competition, especially the pressure from Chinese supply expansion on materials-side ASP.

• After direct CHIPS grants did not finally land, there is currently no equally definitive first-hand new framework for long-term policy alternatives. What can be confirmed at this stage is that 48D tax refunds have arrived in large scale, while direct grants have not.

Reference Sources

This report is mainly based on the following public information:

• Wolfspeed's latest 10-Q, 8-K, S-1, and quarterly press releases.

• Wolfspeed's official announcements on Chapter 11 restructuring and changes to the board and management.

• Official information from the U.S. Department of Commerce / NIST on Wolfspeed's CHIPS PMT, and company announcements on 48D tax refunds.

• Reuters reporting on the 2025 bankruptcy, share price plunge, and exit from Chapter 11.

• Official annual reports, investor materials, and press releases from major competitors, including onsemi, Infineon, and Navitas.

Other Securities Mentioned in This Report

• ON.US - A direct comparison in SiC and automotive/industrial power semiconductors, and already positioned in the AI data center power chain

• STM.US - An integrated power semiconductor leader whose full SiC supply-chain capability serves as a horizontal reference for Wolfspeed

• IFX.DE - A system-level power semiconductor leader advancing both AI power and SiC, serving as a benchmark for valuation and execution

• 6963.TSE - A strong Japanese power semiconductor player, reflecting the regional competitive landscape of the SiC industry in the automotive chain

• NVTS.US - A high-beta comparison for the AI data center 800V HVDC narrative, illustrating that "cooperation" does not equal scaled revenue

• MTSI.US - The buyer of Wolfspeed's divested RF business, also useful for observing the company's boundaries after focusing on SiC

• 6723.TSE - Renesas Electronics, both a historical major customer and a key creditor/equity-related party in the restructuring

This report is based on public information and does not constitute investment advice. Markets carry risk; invest with caution.