• The silicon wafer market is a highly concentrated oligopoly dominated by companies with decades of accumulated process know-how. Five major players — Shin-Etsu Chemical, SUMCO, GlobalWafers, Siltronic, and SK Siltron — account for more than 90% of the market. Moreover, the two largest suppliers, Shin-Etsu Chemical and SUMCO, together control more than 50%.

• In 2020-2022, amid a surge in end-market demand and supply-chain disruptions, semiconductor manufacturers rushed to secure wafer volumes years in advance. In response, wafer suppliers, backed by long-term agreements, shifted from incremental capacity upgrades to large-scale greenfield expansion. This dynamic ultimately led to an unprecedented inventory build-up that the industry has been working through until recently.

• The rapid capacity expansion across the semiconductor industry is expected to drive substantial growth in silicon wafer demand. DRAM alone, including HBM, could generate incremental wafer demand of approximately 560,000 wafers per month by the end of 2027.

• However, wafer manufacturers are reluctant to pursue further capacity expansion because the economics do not work at current ASPs. We expect the wafer market to enter a major pricing cycle. The previous 2017-2018 supercycle provides a useful reference point, when prices for 300 mm wafers increased by approximately 50% over two years.

Subscribe now

The Product Behind Every Chip

Every semiconductor chip begins its life not in an Nvidia R&D lab or at a TSMC fab, but at the bottom of a quarry — as a piece of quartzite. Cargo ships carry this quartzite to the Norwegian town of Holla, where it is smelted at 2,000°C to produce metallurgical-grade silicon. It is then transported to Wacker’s facility in Burghausen, where the silicon is converted into gas, passed through distillation columns, and deposited back into solid form, producing polysilicon with a purity of 99.9999999999%, or 11-12N. This corresponds to just one impurity atom for every trillion silicon atoms.

Wacker crushes the polysilicon rods under specialized cleanroom conditions into larger chunks (~30-100 mm) and smaller chips (~5-30 mm). The resulting dark-gray fragments, with their characteristic metallic sheen, are then packaged and shipped to customers. Ironically, one of the world’s most technologically sophisticated products begins with an unremarkable gray rock.

Semiconductor-grade polysilicon costs around $50-60 per kg ($23-27/lb), compared with approximately $5-7 per kg for solar-grade material. The price of 12N-purity polysilicon can be even higher, reaching $120-150 per kg [20].

From this ultra-pure polysilicon, companies such as SUMCO, Shin-Etsu, and GlobalWafers produce silicon wafers. The process begins by loading 300-400 kg of polysilicon chunks and chips into a quartz crucible, where they are melted at approximately 1,420°C using a surrounding graphite heater heated by an electric current. A tiny seed crystal — a pencil-sized piece of silicon with the desired crystal structure — is then slowly lowered into the molten silicon. It is subsequently pulled upward just as slowly while being rotated.

As the seed is withdrawn, atoms from the molten silicon attach to its underside and solidify, replicating its crystal structure. Layer by layer, over the course of 24-48 hours, this process produces a cylindrical single-crystal silicon ingot approximately 100-300 mm in diameter and weighing around 200-300 kg. This process is known as the Czochralski method.

The ingot is then ground to the required diameter (200 or 300 mm) and sliced over the course of 8-12 hours using a diamond wire saw into hundreds of wafers approximately 775 μm thick — slightly less than one millimeter. The wafers are then ground, chemically etched to remove the damaged surface layer, and subjected to chemical-mechanical polishing (CMP), which makes the surface almost perfectly smooth. The result is a mirror-like finish with an average surface roughness of less than 0.1 nanometers. For comparison, if a wafer were scaled up to the size of a football field, the largest bump on its surface would be less than 0.5 mm high. Such demanding standards are necessary because transistors only a few nanometers in size will ultimately be fabricated on the wafer.

Each finished wafer is inspected across more than 50 parameters — including geometry, flatness, crystal lattice defects, and surface cleanliness — before being sealed in an airtight container under cleanroom conditions and shipped to customers such as TSMC, Samsung, or Intel.

The full manufacturing process, from crystal growth to the packaging of finished wafers, is well illustrated in this video:

Silicon wafers are classified according to several key characteristics, including diameter, crystal structure, surface treatment, and specialized applications for specific technologies, such as AI or power electronics. The primary classification is based on wafer size and processing technology:

• Size (primarily 200 mm and 300 mm diameter)

• Processing type (polished wafers vs epitaxial wafers)

Newer, more advanced 300 mm wafers have 2.25x the surface area of a 200 mm wafer but cost only 1.3-1.5x as much to process. This translates into a 30-40% reduction in cost per die when migrating from 200 mm to 300 mm. The 200 mm format served as the industry workhorse throughout the 1990s and 2000s and today primarily supports older-generation manufacturing, including automotive MCUs, industrial sensors, power management ICs, and RF components.

Fabs built in the 1990s and 2000s continue to use 200 mm wafers because their equipment was designed around this wafer size, while the cost of retooling an entire fab would exceed the potential savings per die. However, as older capacity is gradually retired, the transition toward 300 mm wafers will continue.

The transition has been rapid. Major semiconductor manufacturers such as Texas Instruments aim to increase the share of their internal production running on 300 mm wafers to 80% by 2030, inevitably reducing their demand for 200 mm substrates [5]. Today, on a surface-area basis, 200 mm wafers account for approximately 20% of the total market.

After the crystal is sliced and polished, there are two main product categories: polished wafers (PW) and epitaxial wafers (epi). A polished wafer is essentially the finished slice of a silicon crystal, polished to extreme smoothness through the CMP process described above. It is a more commoditized product and is therefore more exposed to industry cycles. The memory market (DRAM and NAND) is the largest consumer of PW wafers, while a significant share is also used to manufacture simpler controllers and chips on legacy nodes, as well as for monitoring and test wafers used to calibrate semiconductor equipment. Approximately 75% of all 300 mm wafers are polished wafers.

Because a large share of wafer shipments is made under long-term contracts with terms of three to five years, available pricing data varies considerably. For example, a former ON Semiconductor manager estimated the price of a 300 mm wafer at $125-130 in 2023 [20]. However, based on our calculations using SUMCO data, the ASP for these wafers appears to be closer to $90-100.

The product mix also matters significantly. Wafers with the lowest defect levels are sold as high-grade products to customers such as Intel, TSMC, and Samsung for 5 nm processes, commanding the highest prices. Wafers that do not meet this threshold can instead be used for 14 nm or 28 nm nodes, with pricing differentiated by grade.

An epi wafer starts with the same polished wafer, which serves as the substrate (or handle wafer). A thin layer of single-crystal silicon — the epitaxial layer — is then grown on the PW surface in a specialized reactor. This additional layer has a more perfect crystal structure than the underlying wafer, resulting in fewer defects, better electrical properties, and higher manufacturing yields for the chipmaker. The epi layer enables the wafer to meet the extremely demanding specifications required for leading-edge logic at 3 nm and 2 nm.

Epitaxial wafers are typically manufactured to customer-specific specifications. Each fab operates a unique manufacturing process with its own equipment parameters, lithography requirements, and thermal conditions, meaning that the wafer must be tailored to that specific process. In other words, a supplier does not simply sell a “300 mm epi wafer” — it sells a “300 mm epi wafer built to TSMC N3 specifications,” which differs from one designed to “Samsung 3GAE specifications.” This customization allows wafer manufacturers to achieve higher margins and more stable pricing compared with polished wafers.

“That said, as I have been saying at each of our recent meetings, 300-millimeter epitaxial wafers are used in made-to-order products while 300-millimeter polished wafers are used for memory, which is typically produced in anticipation of demand and, therefore, has a tendency to overshoot. So it is necessary to take into consideration not only wafer inventory levels but end product inventory levels as well, which appear to be high. As such, a recovery in polished wafers is likely to take more time than epi in my view.” — Mayuki Hashimoto, CEO & Chairman at Sumco [7].

As with polished wafers, there is no publicly available data on the average pricing of epitaxial wafers. Based on our estimates, their ASP is approximately $130-140, depending on the complexity of the product and contract terms. For the most advanced solutions, prices can reach $240-250 per wafer, for example, for wafers used in CMOS image sensors.

Thus, epitaxial wafers are significantly more expensive than polished wafers and require more sophisticated manufacturing equipment and deeper process know-how.

These extreme technical requirements help explain why the industry is a highly concentrated oligopoly dominated by companies with decades of accumulated process expertise. There are five major players in the market: Shin-Etsu Chemical, SUMCO, GlobalWafers, Siltronic, and SK Siltron. Together, they account for more than 90% of the market. Moreover, the two largest players — Shin-Etsu Chemical and SUMCO — alone control more than 50%.

Decades of process refinement and collaborative development with TSMC, Samsung, and SK Hynix create barriers that capital alone cannot replicate. You can build a wafer fab in three years, but you cannot replicate decades of accumulated process know-how in three years.

When a memory maker qualifies a new wafer supplier for HBM production, it must run thousands of test wafers through the full process flow, measure yields, correlate defect patterns, optimize process parameters, and validate reliability. This process takes 12-24 months and consumes engineering resources that could otherwise be devoted to production optimization. Even if the new supplier’s wafers perform well in testing, the customer faces switching costs in the form of lost engineering productivity and the risk that those yields may not translate to production scale. The rational decision is almost always to stay with the incumbent.

The economics further reinforce the moat. A silicon wafer accounts for roughly 3-5% of the total cost of the finished semiconductor device. But a defective wafer that causes yield losses can destroy value worth 10-50x the wafer’s price. This asymmetry means that, for advanced applications, customers are willing to pay a significant premium for wafers with proven, consistent quality rather than risk significant yield losses to save a few dollars per wafer. The value-to-cost ratio is so skewed that price sensitivity effectively disappears at the leading edge.

The table below outlines the approximate revenue mix for each of the “Big Five” in the wafer segment. These figures are derived from official company disclosures, earnings presentations, and Tegus expert interviews.

Shin-Etsu and SUMCO have the highest concentration in advanced 300 mm epitaxial wafers, precisely where AI demand is strongest, and margins are highest. GlobalWafers’ 200 mm exposure (35-43% of revenue) is the largest among the majors, limiting its AI upside but providing optionality if the 200 mm market stabilizes. SK Siltron data is estimated from limited public disclosure; the company is privately held within the SK Group and does not report a granular product mix.

Anatomy of a Wafer Cycle

The stability of these market positions is reinforced by how the industry structures its commercial relationships and capacity investments. The silicon wafer industry has a structural feature that creates a real investment opportunity: massive time lags between capex decisions and capacity coming online. Building a new wafer fab takes approximately three years from groundbreaking to production. These long lead times apply equally on the demand side: Micron’s new Idaho chip fab, for example, is on a three-year construction schedule, with production expected by 2027. Each new chip fab of this scale represents incremental wafer demand that is locked in years before the first lot runs. This matters for two reasons: when demand surges, supply cannot respond quickly; and when investment decisions made during booms bring capacity online just as demand softens, the result is an extended period of oversupply.

Customer relationships in the industry are governed by long-term agreements (LTAs). The process begins with a foundry receiving an 18-36-month forecast of end demand from its customers. The foundry then passes this demand signal upstream to wafer suppliers, providing roughly six months of additional forward visibility — a total horizon of 24-42 months — beyond the duration of its own customer contracts. The parties then enter into LTAs, typically with terms of three to five years.

During periods of strong demand and heavy investment in new production capacity, companies may enter into agreements lasting five to six years. In such cases, capacity expansion projects are often tied to customer commitments to purchase output for up to five years after the new capacity comes online.

These contracts typically include commitments to purchase specified volumes and tightly define product specifications, including substrate type, diameter, crystal quality, flatness, and defect density. Most contracts are based on fixed pricing, although some incorporate step-up mechanisms under which prices increase annually, for example, by 2-3%. Customers generally cannot cancel these contracts.

However, this mechanism — designed to protect wafer manufacturers from the industry’s inherent cyclicality — ultimately became the source of the unprecedented headwinds they have faced in recent years.

During the pandemic in 2020-2022, a sharp surge in end-market demand and widespread supply-chain disruptions pushed semiconductor manufacturers and their customers to secure access to critical components at almost any cost. As a result, major foundries and IDMs aggressively entered into LTAs with silicon wafer suppliers, locking in volumes several years in advance. Against this backdrop, wafer manufacturers shifted from incremental upgrades to existing production lines toward large-scale construction of new fabs. Importantly, much of this planned capacity had already been committed to customers under LTAs before construction even began.

The following excerpt from GlobalWafers’ Q4 2021 earnings call is particularly illustrative:

“At end of — as of end of 2021, the amount of our prepayment balance reached TWD 28.6 billion or around USD 1 billion. And this is TWD 6.2 billion higher than our prepayment balance as of the end of Q3. So in Q4 only, this one quarter, we received another new TWD 6.2 billion new prepayment — TWD 6.2 billion prepayment or USD 0.2 billion. And this prepayment amount keeps increasing in 2022 as well.

Our capacity expansion as we announced to the market that we have our brownfield and greenfield big capacity expansion. Overall, global digital transformation accelerated by COVID-19 has propelled the demand for cloud services -- server and also high-performance computing. 5G is rolling out an unprecedented pace and its uptake is far faster than 4G, so embodying people’s need for connectivity. Yes, connectivity, this is the key word for the growth. And those technological advancements and increasing silicon content could not be achieved under current level of worldwide semiconductor production capacity. Therefore, we are seeing very aggressive expansion program announced by almost all the key 1 — the key Tier 1 players, all key players in all sectors of semiconductor supply chain” — Hisu-Lan Hsu, Chairperson & CEO

In September 2021, SUMCO decided to invest a total of ¥228.7 billion ($2.08 billion) to expand its 300 mm epitaxial wafer production capacity. The company’s capital expenditures increased from ¥69.5 billion ($0.63 billion) in FY2021 to ¥130.8 billion ($1.05 billion) in FY2022.

Market leader Shin-Etsu pursued a more gradual expansion strategy, focusing on meeting demand through brownfield projects. Capital expenditures in its Electronic Materials segment totaled ¥106.5 billion ($0.97 billion) in FY2021, while the company guided to ¥300 billion ($2.40 billion) in FY2022 capex for the group as a whole.

Taiwan-based GlobalWafers redirected funds originally earmarked for the acquisition of Siltronic toward aggressive organic capacity expansion. In early 2022, the company announced a three-year capital expenditure plan (2022-2024) totaling TWD 100 billion ($3.57 billion). Approximately 55% of this amount was allocated to greenfield projects, including a new fab in Texas, while the remaining 45% was earmarked for expansions at existing facilities across six countries.

Germany-based Siltronic embarked on the largest investment project in its history. The company’s capital expenditures more than doubled, from €425.6 million ($0.46 billion) in 2021 to €1,073.6 million ($1.16 billion) in 2022.

The financial footprint was unprecedented. Sector capex roughly tripled — from a steady-state run rate of $1.5-1.8 billion per year in 2018–2021 to $3.8 billion in 2022 — before peaking at approximately $5.5 billion in 2023, followed by another $4.8 billion in 2024. Cumulative spending over 2022-2024 reached roughly $14 billion.

According to SUMCO estimates, global 300 mm wafer production capacity increased from approximately 6.0 million wafers per month in 2020 to more than 8.0 million wafers per month in 2022.

However, headwinds began to emerge as early as Q4 2022: shipments started to decline, while customer inventories rose rapidly, particularly in logic. Nevertheless, the prevailing industry view remained that LTAs and prepayments would cushion the cycle, as they had in 2019. The first warning sign came from Siltronic. In early February 2023, the company released preliminary results for 2022, providing a cautious but broadly neutral-to-optimistic outlook. Just one month later, Siltronic published its final results, highlighting a sharp deterioration in market conditions and order push-outs from customers.

The second, even stronger signal came from GlobalWafers. The company disclosed that, under the terms of its contracts, it had returned a portion of customer prepayments because customers had failed to take the volumes they had originally committed to purchase.

The third and clearest signal came from SUMCO. In February 2023, the company projected 2024 demand at 102% of supply — a tight, undersupplied market. By November 2023, that figure had been cut to 76%. The industry went from “structurally short” to “structurally long” in a matter of months.

On the one hand, LTAs supported revenues and pricing, as customers largely continued to honor their contractual commitments. On the other hand, those same LTAs forced customers to continue taking wafers they no longer needed, contributing to a further build-up of inventories.

Wafer manufacturers did not respond with panic. Even amid weak market conditions, SUMCO noted that customers continued to honor LTA pricing, while the company was still able to implement previously agreed price increases. In other words, the adjustment did not come through price capitulation, but rather through active management of shipment volumes and delivery schedules.

Manufacturers began accommodating push-outs, delivery adjustments, and delays. By early 2024, SUMCO management explicitly stated that, given extremely high customer inventories, the company had “no choice but to cooperate with customer requests to delay deliveries.”

At the same time, manufacturers began cutting production and slowing the ramp-up of new capacity. SUMCO, for example, continued with its domestic expansion projects but slowed the capacity ramp at its Taiwan/FST operations, which had greater exposure to polished wafers, where the correction was more severe.

Around the same time, a new structural challenge for the industry began to emerge: the accelerating entry of Chinese producers into the market. Chinese wafer manufacturers have made impressive progress, with combined capacity estimated at 1-2 million wafers per month across 200 mm and 300 mm wafers. According to J.P. Morgan, the market share of Chinese companies increased from 2.7% in 2021 to 19.4% in 2024 and is on track to exceed 30% [19]. However, it is important to note that Chinese producers are particularly strong in the 200 mm market, which primarily serves mature nodes.

By 2024, the industry recovery had become increasingly selective. The 300 mm segment, particularly applications related to AI and leading-edge technologies, was recovering more strongly, while 200 mm and smaller-diameter wafers remained weak. Shin-Etsu highlighted this divergence in demand particularly clearly. During an earnings call, management emphasized that, in the current market environment, only AI-related semiconductor devices were performing well, while other end markets remained stagnant.

Siltronic took the most decisive action: in March 2024, the company announced that it would exit the small-diameter wafer business by 2025, explaining that the segment had become too small and was facing intensifying competitive pressure from Chinese suppliers.

By May 2024, SUMCO’s CEO confirmed that 300 mm wafers had passed the bottom of the cycle, while 200 mm wafers continued to “crawl along the bottom.” Nevertheless, even in the 300 mm segment, a meaningful recovery in wafer consumption at semiconductor fabs did not translate into higher wafer purchases, as customers prioritized drawing down their existing inventories.

Overall, 2024 and 2025 remained years of sharp polarization, with explosive demand for AI-related products contrasting with prolonged stagnation across consumer and industrial end markets. At the same time, the industry faced an unprecedented increase in depreciation expenses as major investment programs reached completion, putting significant pressure on profitability in 2025 and contributing to a cautious outlook for 2026.

It was not until November 2025 that the shipment trend finally returned to 2022 levels, the industry’s previous peak. For the full year, global silicon wafer revenue reached $11.4 billion, with total shipments of 12.97 billion square inches (MSI). Shipments of 300 mm wafers averaged approximately 7.7 million wafers per month against installed capacity of roughly 8.8 million wafers per month, implying approximately 13% excess capacity in the 300 mm segment — or roughly 87% utilization. This supply-demand imbalance lies at the heart of the current cycle.

From Oversupply to Shortage

The inflection point for the industry came in late 2025 and early 2026, when the rapid increase in prices for memory and other AI infrastructure components triggered a sharp upward revision in industry capex expectations. In early March, GlobalWafers Chairwoman and CEO Doris Hsu told institutional investors that prices for both 200 mm and 300 mm wafers had bottomed in Q4 2025 — Q1 2026, rush orders were beginning to emerge, and advanced 300 mm capacity was fully utilized. By August, following the company’s latest reporting period, management noted that capacity was fully utilized across all segments and that supply of certain advanced 300 mm wafers had already become noticeably tight [19].

Executives at other companies also became significantly more optimistic about supply-demand dynamics. In May, SUMCO management noted that the company was still in recovery mode: customer inventories were declining, while some equipment remained underutilized. By August, however, President Ryuta highlighted that capacity utilization was approaching 100%, customers were increasingly concerned about securing sufficient wafer volumes for 2027-2028, and some were already considering rebuilding inventories. According to SUMCO, the 300 mm wafer market grew by approximately 15% year over year in Q1 2026 [19].

While expectations remained cautious at the beginning of the year, concerns about a potential shortage of leading-edge wafers have since intensified, prompting major customers to secure future supply. Micron Technology, for example, announced a strategic partnership with GlobalWafers involving a $500 million investment to expand U.S. silicon wafer production capacity, alongside a 10-year supply agreement that guarantees Micron access to significant volumes of 300 mm wafers. SK Group also abandoned its earlier plans to sell wafer manufacturer SK Siltron to Doosan [19].

Several factors are driving wafer demand and contributing to the expected supply shortage, including the memory supercycle, growth in leading-edge logic, and a recovery in mature nodes.

If you are even remotely familiar with the stock market, it is no secret that the memory industry is experiencing one of the most impressive upcycles in its history. While the prevailing view last year was that price increases would normalize within a reasonable timeframe, the consensus today is increasingly that the supply-demand imbalance is here to stay. Against the backdrop of LTAs signed with their customers — an interesting parallel with the wafer market — leading memory manufacturers such as Samsung, SK Hynix, and Micron have begun radically revising their investment and capacity expansion plans.

Samsung Electronics plans to increase its DRAM capacity from 640,000 wafers per month in Q1 2026 to 855,000 wafers per month by Q3 2027. SK Hynix is expected to expand its DRAM capacity from 565,000 wafers per month in Q1 2026 to 695,000 wafers per month by Q3 2027. According to TrendForce estimates, global 300 mm DRAM wafer capacity is expected to increase from 1.64 million wafers per month in 2024 to more than 2.5 million by the end of 2027. Over the longer term, total DRAM capacity is expected to increase by 2-3x compared with 2024 levels. Samsung Electronics’ capacity, in particular, could reach 1.35 million wafers per month, while SK Hynix could reach 1.0 million wafers per month by 2030 [19].

Equally important, HBM is accounting for a growing share of the production mix. According to Deutsche Bank estimates, HBM’s share of total DRAM bit production is expected to increase from approximately 5% in 2025 to around 10% by 2030. HBM production requires significantly more silicon wafers than conventional DRAM for the same amount of memory capacity. Specifically, HBM requires roughly three times more silicon wafer area per bit than conventional DRAM, while for HBM4/HBM4E, the ratio could reach 4:1 [20].

As a result, DRAM alone, including HBM, could generate incremental silicon wafer demand of approximately 560,000 wafers per month by the end of 2027. If the capacity expansion plans outlined by Samsung Electronics and SK Hynix are fully implemented, incremental wafer demand from these two manufacturers alone could exceed 1.1 million wafers per month by 2030.

Similar trends are emerging in leading-edge logic. To address the capacity shortage for AI chips, TSMC raised its 2026 capex guidance to $60-64 billion. The company expects investment over the next three years to be significantly higher than over the previous three. Based on our estimates, TSMC alone could generate incremental wafer demand of approximately 280,000-320,000 wafers per month by 2028.

A second tier of capacity additions could contribute another 100,000-120,000 wafers per month over the same horizon. Intel has reached approximately 30,000 wafers per month on its 18A process after addressing yield issues, with Fabs 52 and 62 designed for combined capacity of up to 40,000 wafers per month [14]. Samsung Foundry is targeting 50,000 wafers per month of SF2/SF3P capacity in Taylor, with a mature ramp realistically achievable by 2028 [15]. Rapidus, meanwhile, is expected to launch a pilot line with capacity of approximately 10,000 wafers per month from 2027 [16]. Combined, these additions imply incremental capacity of approximately 380,000–440,000 wafers per month.

SEMI expects advanced-node capacity (≤7 nm) to increase from 850,000 wafers per month in 2024 to approximately 1.4 million wafers per month by 2028, representing a CAGR of roughly 14% and implying approximately 550,000 wafers per month of incremental demand from leading-edge logic alone [15].

Thus, the current fab investment cycle is set to generate substantial incremental demand for silicon wafers. According to Barclays’ estimates, combined demand for leading-edge logic and DRAM wafers used in servers, smartphones, and PCs is expected to grow at a CAGR of approximately 17% from 2025 through 2030, adding roughly 1.0 million wafers per month of demand between 2025 and 2028, followed by another approximately 0.5 million wafers per month between 2028 and 2030 [16].

According to CTEE, growing demand from AI servers and advanced chip manufacturing could push the global silicon wafer supply shortfall to 8% by the end of this year, 15% by the end of next year, and 24% by the end of 2028 [18].

Leading wafer manufacturers such as Shin-Etsu, SUMCO, and GlobalWafers have noted that their 300 mm wafer production lines are already operating at or near full capacity, while supply is becoming increasingly constrained.

“With regard to production for 300-millimeter, we are very close to full capacity utilization... We will truly be at full capacity utilization in Q4.” — SUMCO Q2 FY2026 conference call

Based on the experience of previous cycles, one might expect the next step to be another investment supercycle in the silicon wafer industry. However, comments from wafer manufacturers point to a different scenario: rather than aggressively expanding capacity, the market could move into a pronounced supply shortage, creating a seller’s market and setting the stage for substantial price increases.

A particularly noteworthy exchange took place between a Goldman Sachs analyst and SUMCO’s President during the company’s latest earnings call. The analyst noted that, based on their calculations, wafer prices would need to be roughly twice their current levels for capacity expansion to generate an adequate return. Management confirmed that a 5-10% increase in ASPs would clearly not be sufficient:

“GS’ Analyst:

I am Ikeda of Goldman Sachs. In looking at the silicon wafer projections for servers you have outlined for 2027 and 2028 on page 15, the numbers appear very conservative if you take into account the expected increase in chips per end product or CBA. Have you baked in some supply risk into your forecast?

For 300-millimeter, I expect 2027 to be very tight and expect to see shortages in 2028. Despite this backdrop, I understand that you are hesitant to commit to greenfield investments, as you alluded to earlier. If you use a hurdle rate that is close to your weighted average cost of capital, you would need to see a doubling of wafer prices to enable new investments, or the economics would not work at all. It means that prices need to be significantly different from current levels.

With the profitability improving significantly at your customers and structural demand growing, what sort of negotiations are you currently conducting? Are there measures you can put into place, such as requiring upfront payments or extending the tenure of the LTAs, to ensure that you can sustain profit growth over the longer term? Please comment to the extent you can, including lesson learned in hindsight.

SUMCO’s Jiro Ryuta:

There are many things that I regret in hindsight in many different fields, which makes it difficult to talk about in short. It is true that the forecast for server-use silicon wafer demand takes into account the fact that the supply side is not in a position to increase supply at this time. Actual demand levels are higher, but at this stage, it is not possible to grow supply. Demand is likely to exceed supply for more than three years, in my view, because of such supply constraints.

In terms of the price level that we would need to increase supply capacity and invest to expand supply, we have made no specific statement at this stage. But what I can say is my personal sense is that prices would need to be dramatically higher. To your point, I think an increase of 5% or 10% would not have a meaningful impact. That’s my impression.”

Siltronic management also confirmed that, while spot prices have begun to increase selectively, they have yet to reach levels that would justify large-scale investment in further capacity expansion. As a result, the company is taking a selective approach, prioritizing contracts with more attractive terms rather than simply pursuing volume growth [17].

As the market leader, Shin-Etsu has noted that it is actively negotiating price adjustments even under existing contracts. The company is discussing LTA price revisions with customers in response to sharp increases in electricity, raw material, and logistics costs. In the memory segment, the company is receiving requests for LTAs extending through 2028 [18].

According to SunSirs Commodity Data Group, starting in May, leading manufacturers issued successive price increase notifications, raising prices for standard 12-inch wafers by 3-8% and high-end AI-specific wafers by nearly 20%, resulting in a cumulative annual increase of more than 15% [19].

Thus, the silicon wafer industry is set to benefit from substantial incremental demand, while manufacturers, having learned from the previous cycle, are publicly resisting aggressive capacity expansion because the economics simply do not work at current ASPs. The key factor is the asymmetry in pricing: the substrate accounts for only 3-5% of the cost of a processed wafer at the fab, meaning that even a doubling of wafer prices would translate into only a 3-5% increase in chip manufacturing costs — an amount that pales in comparison with the cost of leaving leading-edge capacity idle. The logical outcome is one of the most significant pricing cycles in the industry’s history, with the previous 2017-2018 supercycle providing a useful benchmark, when 300 mm wafer prices increased by approximately 50% over two years.

Sources

[1] https://www.wacker.com/cms/en-us/press-and-media/press/press-releases/2024/detail-228484.html

[2] https://en.wikipedia.org/wiki/Czochralski_method

[3] https://www.sumcosi.com/english/products/process/step_01.html

[4] YouTube

[5] https://investor.ti.com/static-files/fc9d9346-cf77-40db-902a-e9961e9c5736

[6]https://www.siltronic.com/fileadmin/investorrelations/2025/Q4/20260312_Siltronic_InvestorPresentation.pdf

[7] Sumco Corporation, 2023 Earnings Call, Feb 14, 2024

[8] https://www.sumcosi.com/english/products/process/step_03.html

[9] https://reports.valuates.com/market-reports/QYRE-Auto-21H9107/global-single-crystal-silicon-wafers-300mm

[10] GlobalWafers Co., Ltd., 2021 Earnings Call, Mar 15, 2022

[11] https://ssl4.eir-parts.net/doc/3436/ir_material_for_fiscal_ym18/130842/00.pdf

[12] https://ssl4.eir-parts.net/doc/3436/ir_material_for_fiscal_ym18/172131/00.pdf

[13] https://www.kad8.com/ai/intel-reports-18a-yield-breakthrough-30000-wafers-per-month-capacity-and-14a-roadmap/

[14] https://www.tomshardware.com/tech-industry/semiconductors/samsungs-taylor-texas-fab-could-herald-a-breakthrough-for-the-chipmaker-company-plans-2026-risk-production-new-production-flows-pellicles-for-euv-patterning-as-site-targets-50-000-wspm

[15] https://www.semi.org/en/semi-press-release/semi-forecasts-69-percent-growth-in-advanced-chipmaking-capacity-through-2028-due-to-ai

[16] Barclays, European Technology Hardware: From GW to WFE II

[17] https://www.siltronic.com/en/investors.html

[18] https://www.taiwannews.com.tw/news/6419954

[19] Bloomberg

[20] AlphaSense

Subscribe now

Disclaimer / Disclosures

The information contained in this report is provided solely for informational and educational purposes and does not constitute investment advice, investment research under applicable securities regulations, an offer to sell, or a solicitation to buy any security or financial instrument.

The opinions expressed in this report reflect the author’s views as of the date of publication and are subject to change without notice. This report has been prepared without regard to the specific investment objectives, financial situation, or particular needs of any individual investor. Readers should conduct their own independent research and seek professional financial, legal, tax, or other advice before making any investment decision.

The analysis presented in this report is based on information believed to be reliable, including publicly available sources, company disclosures, industry publications, and discussions with market participants. While reasonable efforts have been made to ensure the accuracy of the information, no representation or warranty, express or implied, is made as to its accuracy, completeness, or reliability. Any estimates, assumptions, or forecasts represent the author’s judgment and are inherently subject to uncertainty.

This report contains forward-looking statements, including expectations regarding future financial performance, industry trends, market developments, and company prospects. Such statements involve known and unknown risks, uncertainties, and assumptions that may cause actual results to differ materially from those expressed or implied. No assurance can be given that any projections or forecasts will be realized.

The author may hold, establish, increase, reduce, or dispose of positions in any securities discussed in this report at any time without prior notice. Such positions may create potential conflicts of interest. Nothing in this report should be interpreted as a recommendation to buy, sell, or hold any security.

To the fullest extent permitted by applicable law, the author disclaims any liability for any direct or indirect loss arising from the use of, or reliance upon, the information contained in this report.

When we created Grana Research, our goal was to show how investment decisions are made in practice within a small buy-side firm — from the initial screening of ideas and the construction of a watchlist to initiating a position and managing it over time. Every investment passes through a kind of funnel: dozens of companies may appear interesting at the top, but only a small number ultimately make it into the portfolio.

Ideas Hub represents the top of our investment funnel. Here, we highlight companies that have recently caught our attention and merit further research. Some ideas emerge from sell-side research, quarterly fund letters, or Substack publications; others arise from our own observations and analysis.

While most Ideas Hub posts will feature short collections of investment ideas, from time to time we will also publish standalone investment theses. The first of these is dedicated to Unity Software .

The Company

Unity is one of the leading platforms for creating, launching, and operating games and interactive content. The company offers a suite of tools spanning all major devices and environments, including mobile platforms, personal computers, consoles, and extended reality. In particular, Unity develops one of the world’s two leading game engines — software platforms used by developers to build 3D games and interactive content.

Unity’s platform consists of two complementary segments:

• Create Solutions. A suite of tools that developers use to create, launch, and maintain real-time 2D and 3D content. At the core of this segment is the Unity game engine. A game engine is the technical foundation and “construction kit” used to build a game or interactive 3D application: it determines how characters move, how lighting and graphics are rendered, and how sound, physics, interfaces, controls, networking, and player interactions function.
  Beyond the engine itself, Unity’s ecosystem includes an integrated development environment, graphics and animation tools, and solutions for audio, user interfaces, networking, navigation, and multiplayer projects.
  More than 70% of the world’s top 1,000 mobile games were built using Unity technologies, while the engine is used in 82% of the top 100 mobile games — not only during development, but also for subsequent updates and ongoing support.
  Unity also maintains a significant presence in the PC market: more than 50% of games on Steam are built with Unity.

• Grow Solutions. Solutions for growing and monetizing games, primarily related to Unity’s advertising business. They help developers acquire users, retain audiences, and monetize content, regardless of whether it was built with the Unity engine. This segment includes Unity’s advertising network, a platform for managing ad placements, tools for in-app advertising offers, and Supersonic, a publishing solution that helps smaller studios and independent developers launch mobile games and improve their commercial performance.
  Another product, Aura, helps developers promote applications directly on users’ devices, while enabling telecom operators to better monetize their audiences throughout the device lifecycle.

Taken together, Unity’s solutions support developers throughout the entire product lifecycle — from launching a game to acquiring users and monetizing it. This makes Unity one of the gaming industry’s core infrastructure platforms.

Importantly, Unity’s customer base has historically remained remarkably resilient despite years of operational missteps within the company (discussed below). Once a developer chooses Unity to build a game, they typically remain on the engine throughout the project’s entire lifecycle — and often use it again for future titles. Migrating to another platform requires rebuilding much of the development workflow: adopting new tools, transferring libraries of art and game assets, rewriting portions of the codebase, and re-establishing the production pipeline.

Unity’s large developer community also remains one of its key competitive advantages. As more studios build products using the Unity engine, third-party platforms have stronger incentives to integrate with Unity and streamline content deployment on their own ecosystems. This reinforces a powerful network effect: developers can more easily distribute games and interactive applications across multiple devices and platforms, while Unity retains its position as one of the industry’s leading development platforms for real-time 3D content.

Management Crisis

Unity is a strong business that has been held back by years of management missteps.

The company’s $4.4 billion merger with ironSource — an Israeli platform specializing in mobile advertising, app monetization, and user acquisition — proved unsuccessful and became one of the most controversial decisions in Unity’s history. ironSource went public at the height of the SPAC boom with an $11.1 billion valuation, making it the largest IPO ever by an Israeli company. By the time the merger with Unity was announced in July 2022, however, its market capitalization had fallen to roughly $2.3 billion. As a result, Unity’s offer represented a substantial premium of approximately 75% to the prevailing market value.

The rationale behind the transaction was straightforward: combine two complementary parts of the gaming ecosystem. Unity owned the development side through its game engine, while ironSource specialized in monetization, advertising, and user acquisition. In practice, however, Unity chose not to build upon ironSource’s advertising network. Instead, management shifted its focus to Unity’s own advertising infrastructure and the Vector platform. This strategic pivot was accompanied by the departure of several former ironSource executives, including founder and former CEO Tomer Bar-Zeev.

As a result, a significant portion of the acquired assets has since been either shut down or put up for sale. The closure of the ironSource advertising network and the planned divestiture of Supersonic effectively complete the dismantling of ironSource’s business within Unity. Supersonic is the mobile game publishing business behind titles such as Bridge Race, Going Balls, and Build A Queen.

Following Apple’s iOS 14 privacy changes, Unity also struggled with the accuracy of its advertising products and the consistency of advertisers’ return on ad spend (ROAS). Advertisers were no longer able to optimize campaigns effectively: even if they adjusted budgets to compensate for the algorithm’s previous week’s errors, the deviation could change the following week again. Combined with the post-pandemic slowdown in digital advertising spending, these shortcomings became increasingly apparent relative to competitors.

As a result, Unity’s advertising business rapidly lost market share to AppLovin .

This is how a former Senior Director at Unity described the situation:

“The problem that Unity had in addition to that is, because they were charging per impression and the accuracy wasn’t there, it was also not consistent. What would happen is, let’s say we had a goal, we were off by 30%. They adjust their budget, the goal, down by 30% so that they can actually hit the real goal, or up by 30%, other way. The problem is, the next week when it came back, we’d be off by more than 30% or less than 30%. They couldn’t play optimization game. Effectively, I think that’s the big thing.

The big problem is, the performance wasn’t there in ROAS and the consistency wasn’t there in ROAS product for that month period after the iOS 14 thing. When the world looked up again and people started playing less games, then the budget shrank a little bit, and everybody was deciding where they wanted to spend. It led to a larger [direct] client Unity than our competitors. That’s what happened in that time, from my perspective. Two things happened in concert. We had that weakness on our demand, which impacted our overall revenue, which impacted our share price in the short-term view for investing. At the same time, we were building and launching and we were signing contracts for Unity mediation. AppLovin and ironSource, one of those other people thought that as a threat.”

Notably, in August 2022 — after the ironSource acquisition had already been announced — AppLovin made a non-binding proposal to merge with Unity. The company valued Unity at approximately $20 billion, or about $58.85 per share, representing a 48% premium to Unity’s closing price on July 12. The transaction was structured as an all-stock deal. However, Unity’s board of directors unanimously rejected the offer.

In 2023, Unity made another major mistake by introducing the Runtime Fee, a new charge based on game installations. Until then, Unity’s pricing model had been relatively predictable: smaller studios could use the engine for free, while larger developers paid for licenses and subscriptions to development tools. Under the new policy, however, developers would be charged for every game installation once a title exceeded both $200,000 in revenue over the previous twelve months and 200,000 lifetime installs. The fee could reach as much as $0.20 per installation.

For developers, this amounted to changing the rules in the middle of the game. Many studios had spent years building projects on Unity, and the new pricing model could apply even to games that were already in development — or had already been released.

The backlash was immediate. Developers feared unpredictable and potentially uncontrollable costs. The controversy quickly spread beyond social media: studios signed open letters, threatened to move projects to competing game engines, and some mobile developers even disabled monetization through Unity Ads and ironSource in their games. The concern was straightforward: a user could download a game, spend little or nothing, uninstall it, and then reinstall it, while the developer could still be charged by Unity. Similar concerns applied to free giveaways, demo versions, and games distributed through subscription services. As a result, developers viewed the Runtime Fee as a threat to the economics of their businesses. Unity was ultimately forced to reverse course. In 2024, the company abandoned the Runtime Fee entirely and returned to a traditional per-seat subscription model.

The controversy came at a high cost. It severely damaged developers’ trust in the company and accelerated a leadership transition, with John Riccitiello stepping down as CEO in October 2023.

“When you unilaterally try to change the contract on everybody, game devs everywhere went, That’s terrible. Even though the CEO who did that is gone, they walked back those plans. They’ve tarnished the brand reputation pretty heavily. I’m not saying that nobody uses Unity anymore. Unity is objectively a more tool-rich platform than a place like Godot. Godot, as well, the other way people pronounce that, which is open source and free. There are a lot of indie developers who have switched over to an open-source engine where there is no cost.

There’s no risk of getting screwed by a company that has decided it needs to make more revenue off of you. That’s a real thing that has happened. As somebody on the game’s publishing side, 98% of everything that came through was in Unreal or Unity. Now it’s probably down to 90% is Unreal Unity and Godot. Godot used to be half of the 2%, and now it’s 15%, 20% of the total that comes through. It’s a noticeable and measurable shift in small-game developer behavior.” – Current Product Director at Epic Games

In May 2024, Unity appointed Matt Bromberg as CEO, and his background is well aligned with what the company needs today. Before joining Unity, Bromberg served as Chief Operating Officer at Zynga, the mobile and social gaming company behind franchises such as FarmVille and Words With Friends.

Over the following five years, Bromberg reshaped Zynga’s monetization strategy, data infrastructure, and live operations — including post-launch game support, regular content updates, in-game events, and player retention. By the time he departed at the end of 2021, Zynga had become one of the more profitable operators in the mobile gaming industry. Just a few months later, the company was acquired by Take-Two for approximately $9.86 per share — nearly four times the share price at which Zynga was trading when Bromberg joined.

A Next-Generation Advertising Platform

Following the ironSource acquisition, Unity’s advertising business became overly complex and fragmented. The company was simultaneously operating Unity Ads, ironSource’s advertising infrastructure, and related assets such as Supersonic. Today, Unity is simplifying the business by consolidating it around two core assets: the Unity Ad Network and Vector. In March 2026, Unity announced that it would shut down the ironSource Ads Network and pursue the sale of Supersonic, aiming to eliminate non-core advertising assets and accelerate the growth of its core advertising business.

The centerpiece of Unity’s strategy is now Vector — a next-generation advertising platform that replaces the company’s legacy systems with a single unified machine learning model for ad targeting.

Following Apple’s 2021 privacy changes, mobile advertising became significantly more challenging. Advertising platforms lost broad access to IDFA (Identifier for Advertisers), Apple’s device identifier that had previously enabled user tracking across applications. As a result, the industry was forced to rebuild targeting capabilities using device fingerprinting and probabilistic attribution models.

For Unity, however, these changes had the potential to become a competitive advantage. Unlike traditional advertising platforms, Unity is not only an ad network but also the game engine behind a large share of the world’s mobile games. This gives the company access to proprietary signals that do not rely on IDFA. These signals are generated within the application itself, on the user’s device, and with the developer’s consent.

Unity can observe not only ad clicks and app installs, but also how users behave inside the game: session length, in-game purchases, churn points, level progression, and many other engagement metrics. By comparison, AppLovin primarily sees users through the lens of advertising — bids, clicks, installs, and the post-install events that developers choose to share with the platform. While these data are highly valuable, they do not provide the same depth of insight into in-game behavior.

Previous management failed to capitalize on Unity’s unique gameplay data as effectively as competitors leveraged their advertising data. That is precisely where the new management team sees the opportunity. Vector is designed to unify Unity’s proprietary signals into a single machine learning model for ad targeting and user quality prediction.

According to the Head of Programmatic & Video at a leading media agency, Vector represents a significant step forward for Unity:

Interviewer

Have you tested Unity’s new Vector optimization platform? How does it compare to Unity’s prior optimization tools?

Head of Programmatic & Video at Media Agency

Yes. Vector’s new optimization platform compared to its legacy optimization tools, for example, with ironSource’s LevelPlaybasic optimizers, certainly represents a fundamental shift to more deterministic neural modeling.

Vector is not just really anupgrade but is a complete rebuild designed to achieve parity with AppLovin’s Axon 2.0.

Ultimately, the prior tools for Unity reallyrelied on heavily historical averages, broad cohort-based bidding, whereas now, Vector’s certainly moving towards more real-timeper impression-based decisioning.

That is certainly speeding up the learning phase now, cutting it down to several days in terms of the comparison legacy tools andusing more predictive lifetime value optimization goals as a lever, not just ROAS and cost per install.

Certainly, the Vector system isbuilding momentum, but warrants further testing to understand over time how does it continue to maintain heavy optimizationlayers.

The significant difference is also its integration with the Unity six engine, really understanding the ability for Vector to accessruntime data as this officially rolls out more at scale in new year, in 2026, and knowing what is happening with the player itself ingame context.

That really allows the algorithm to bid more aggressively for certain impressions that legacy tools would haveultimately ignored.

That should lead towards more lift and action-based input, action-based results specifically, as that is continuallyto be evaluated.

Interviewer

Is Vector improving scale, CPI efficiency, or ROAS any early performance differences across verticals?

Head of Programmatic & Video at Media Agency

Vector is improving scale and overall ROI efficiency due to those reasons outlined around better accessibility into game contextand consumers, how they’re interacting with the content itself, as well as understanding what are the most high value moments toshow up in. Beyond that, its integration layers with its overall ecosystem and Unity six engine, creating less load times, creating more playable, rewardable opportunities in the right moments are all contributing to better outcomes and abetter consumer experience.

Interviewer

Is Vector directly driving incremental spending increases on Unity?

Head of Programmatic & Video at Media Agency

Yes. Vector is primarily a driver for any incremental investment or spend to drive more investment on the platform as it relates todriving better outcomes with a stronger bidding algorithm.

Early adopters have already reported meaningful improvements. One advertising buyer, a director at a global digital marketing agency, noted that after testing Vector in the fourth quarter of 2025, they achieved exceptionally strong results, including roughly a 20% improvement in both user engagement and ROAS:

Interviewer

Have you tested Unity’s new Vector optimization platform?How does it compare to Unity’s prior optimization tools

Director at Global Digital Marketing Agency

Actually, we tested it in Q4 2025. Performance was very strong. This is one of those things I mentioned before about these twoplatforms specializing and boosting their technology stack, especially for in games and in apps. It’s one of the things that they’vejust pushed themselves in that space. This is one example here where this actually saw performance increase compared toprevious campaigns.

We actually compared it to Q2 2025 where we had a similar off ering, but without that technology update for optimization. In Q4, wesaw a boost, it was almost 20% engagement rate increase. We attributed that mainly to their optimization improvement because we didn’t change much else. That being the only variable in the equation was likely the reason for that.

Interviewer

Is Vector improving scale, CPI effi ciency or ROAS? Any early performance diff erences across verticals?

Director at Global Digital Marketing Agency

We’re using Unity mainly for B2C focused retail. Both of our clients that use it in that same space, but it improved overall. The thing is about conversions, our conversion isn’t a real conversion. It’s like a mid-funnel one where we want them to go do a certain thing. It has improved that, especially in terms of ROAS. It’s become better quarter over quarter for the two quarters we measured it.

We only have two quarters with the data here we have one without it and one with it. Q4 was with it and it was a better ROAS. I don’t have the exact percent, but it was somewhere in the notable range, I would say 20-ish percent that it caused a question from the client where they wanted to know what caused that. Again, I mentioned before, the only variable in the equation was in addition to the new optimization tactic.

A User Acquisition Manager at Voodoo, the French mobile games and apps developer, also confirmed that Vector’s algorithm works and has proven itself in practice. Before adopting Vector, the company had been unable to achieve consistently strong results on Unity Ads, leaving the platform in continuous testing for more than a year. According to the manager, Vector ultimately enabled Voodoo to achieve a fivefold outperformance relative to its ROAS target:

Interviewer

I got to switch over to Unity. The big debate around Unity is how good is Vector. I got to run through a bunch of questions quickly because we only have 20 minutes left, but what are your early impressions of Vector? Do you believe in it? Do you think it’s something that can catch up with Axon in terms of the efficiency provided? Just your early takes on Vector and how it compares to Axon.

User Acquisition Manager at Voodoo

It’s definitely working. This, I can tell you with clarity. We’ve been using it since late October, early November, probably maybe even earlier than that, 2024. We can see the impact that it had in our operations. I’m saying this with so much confidence because until that point, we never really found any kind of success with Unity but as part of our job, we have to keep testing and we have to keep experimenting and trying to make these channels work.

Unity, being one of the main channels, we never really stopped advertising on Unity. Because we couldn’t really make anything work, we were almost continuously in a testing phase for more than a year. By the way, very quick clarification that some people view this as a technology that advertisers choose to use, it’s not the case. Actually, it’s just the underlying algorithm and we really have no visibility exactly like Axon two. It’s really working. If you like, I can give you some more color on why it’s working and how it’s working, or we can move on to the next one.

The way Vector is helping Unity help us is it’s not like we, all of a sudden, started seeing 5X our ROAS targets that we set in the campaigns. Instead, of course it did bring in improved ROAS. That is absolutely there. The biggest contributing factor for us to keep spending more and more on Unity is we can keep increasing our daily budgets.

This is to a certain point, until we hit a ceiling, we can keep increasing our daily budget and the return on ad spend that we’re getting is not damaged by a meaningfully negative margin. We can go spending $500, $600, $1,000 a day to $10,000, $12,000, $15,000, $18,000, $20,000 and even above while still keeping to hit the targets. That’s the scalability is the real answer.

The expert also noted that the ability to integrate player behavior data directly from the Unity Engine into Vector’s machine learning model represents a “scarily powerful” long-term tailwind. Moreover, the expert suggested that there is a non-trivial possibility that this integration could eventually allow Unity to capture as much as 80% wallet share in the market, drawing a direct parallel to AppLovin’s position today.

Interviewer

Initially a lift to ROAS, then followed by ROAS declining as more bidders come in. In Q3, ROAS improving again would be how to summarize it. Do you agree with their thinking that there’s a lot of information within Unity runtime that they can incorporate into Vector in order to drive better results over time? If, theoretically, know more than anyone about player behavior within games, do you get the sense that that’s a long-term driver for them, a long-term potential tailwind?

User Acquisition Manager at Voodoo

Absolutely. That is a scary thought in the long run. Why I said this, I’ll try to explain it quickly. There’s indeed a lot of valuable data in what Unity can collect with the engine. By the way, it is currently already active with Unity engine version 6.2. From what I gather, nobody’s really opposed to opting in from the studios, from the publishers. Why this is a scary notion is because if they manage to execute this successfully, there are very small number of things that can disrupt that advantage.

As opposed to having mediation as your moat, there could be some other mediation partner innovating in the industry that can steal your place. What Unity is doing right now is another alternative. There’s a “independent mediation” that is coming up recently. I can’t remember the name of them, but they were quite vocal about their launch.

Anyway, this is why it’s scary because we were not happy with AppLovin’s dominance in the market because some games, from what I’ve heard, they would have 70%, 80% dependence or share of wallet to AppLovin and they couldn’t make anything work. That’s a scary thought.

If we live in a world where that is true for Unity in a couple of years, then that’s not very fun. Why I’m saying it’s a long term, not-so-great scenario for the industry as a whole is this feature is tied to the Unity version that the games are using. Once a game is launched, fully global launch, then they start marketing it, etc., there’s very low likelihood that they change the Unity version that game.

It’s not zero, but it needs to be a really big problem with the build or a really big opportunity to get in order to do that because it’s technically very, very complex and risky. We can only start seeing the impact of this, I think not earlier than at least two years.

It is no exaggeration to say that Vector has become the company’s primary growth engine. Launched in early 2025, the product generated $223 million of revenue by the first quarter of 2026, accounting for approximately 44% of Unity’s total revenue. Vector has delivered exceptional momentum, posting sequential revenue growth of roughly 15% for four consecutive quarters. In addition, management confirmed that the business operates at a gross margin of over 80%.

By 2025, Vector had already generated approximately $619 million in revenue, and Unity’s management expects the product to surpass $1 billion in revenue in 2026, implying roughly $400 million of incremental annual revenue. By comparison, the current sell-side consensus appears relatively conservative, forecasting annual revenue growth for the Grow Solutions segment of only $240-320 million through 2030.

As is often the case with companies of this kind, the majority of operating expenses are attributable to stock-based compensation. At Grana Research, we generally treat management’s adjusted financial metrics with a high degree of skepticism. Even when RSUs or PSUs have been granted but not yet vested, they still represent potential shareholder dilution and should therefore be taken into account.

Historically, Unity’s equity compensation plans included an automatic annual increase in the share reserve equal to 5% of shares outstanding. In December 2025, however, the board of directors approved a reduction in the annual share reserve increase to 2.5%. As a result, the company’s operating leverage is likely to improve meaningfully over time.

Even so, Unity’s current valuation does not appear particularly compelling relative to most companies in the advertising technology sector.

However, it is important to consider the economics of the advertising business. As advertising algorithms improve and return on ad spend (ROAS) increases, advertisers naturally allocate larger budgets to the platform until the marginal return on advertising converges with their target return. This flywheel was the primary driver behind AppLovin’s exponential growth following the launch of AXON 2.0 in the second quarter of 2023.

If Unity’s advertising performance can truly become competitive with AppLovin’s, it could represent an inflection point for both the company’s financial performance and its stock.

Disclaimer / Disclosures

The information contained in this report is provided solely for informational and educational purposes and does not constitute investment advice, investment research under applicable securities regulations, an offer to sell, or a solicitation to buy any security or financial instrument.

The opinions expressed in this report reflect the author’s views as of the date of publication and are subject to change without notice. This report has been prepared without regard to the specific investment objectives, financial situation, or particular needs of any individual investor. Readers should conduct their own independent research and seek professional financial, legal, tax, or other advice before making any investment decision.

The analysis presented in this report is based on information believed to be reliable, including publicly available sources, company disclosures, industry publications, and discussions with market participants. While reasonable efforts have been made to ensure the accuracy of the information, no representation or warranty, express or implied, is made as to its accuracy, completeness, or reliability. Any estimates, assumptions, or forecasts represent the author’s judgment and are inherently subject to uncertainty.

This report contains forward-looking statements, including expectations regarding future financial performance, industry trends, market developments, and company prospects. Such statements involve known and unknown risks, uncertainties, and assumptions that may cause actual results to differ materially from those expressed or implied. No assurance can be given that any projections or forecasts will be realized.

The author may hold, establish, increase, reduce, or dispose of positions in any securities discussed in this report at any time without prior notice. Such positions may create potential conflicts of interest. Nothing in this report should be interpreted as a recommendation to buy, sell, or hold any security.

To the fullest extent permitted by applicable law, the author disclaims any liability for any direct or indirect loss arising from the use of, or reliance upon, the information contained in this report.

• RS Technologies (3445 JP) evolved from a distressed asset — the 300mm wafer reclaim line that Rasa Industries sold for a fraction of its value following the Lehman shock — into one of the leaders in the small but growing silicon wafer reclaim market. CEO, founder, and controlling shareholder Ho Nagayoshi owns 43.87% of the company, while insiders and affiliated parties collectively control 50.59%.

• The wafer reclaim market is a concentrated oligopoly, with the four largest players controlling roughly 70% of global capacity. RS Technologies is one of the industry’s leaders. The business model is predominantly service-based: ownership of the wafer remains with the customer, while RST generates revenue by providing reclaim services. Test and monitor (”dummy”) wafers remain an essential consumable for maintaining manufacturing yield regardless of end-market semiconductor demand.

• The wafer reclaim business is the company’s core segment, generating approximately 35% of revenue, 64% of operating profit (or roughly 75% excluding subsidies in the prime wafer segment), with operating margins in the high-thirties and ROIC above 20%. Both of the company’s key facilities (Japan and Taiwan) are currently operating at 100% utilization, running 24/7 across three shifts amid the AI-driven semiconductor boom. We expect segment revenue growth to accelerate from the current mid-teens to the high-twenties over the next several years, while operating profit could increase by approximately 77% between 2025 and 2028.

• Demand is being supported by two structural tailwinds. The first is the construction of new fabs: more than 60 semiconductor fabs are currently under construction worldwide, including 11 in Japan and 9 in Taiwan, while TSMC has raised its 2026 CapEx guidance to $60-64 billion. The second is the migration toward leading-edge process technologies. Reclaimed wafer consumption intensity increases by roughly 3-4x for the same fab capacity (according to industry experts, from approximately 40k reclaim cycles per 50k wafer starts at 28nm to roughly 115k at 3nm and 135k at 2nm). To meet this demand, RST is investing ¥52 billion (~$324.6 million) to expand total reclaim capacity to 1.19 million wafers per month by 2028 (+72%), with Taiwan representing the primary expansion focus.

• The prime wafer business (approximately 26% of revenue and 26% of operating profit) is entirely China-based and is consolidated through publicly listed GRITEK (688432 CH), in which RST holds an effective 39.1% economic interest. This business is capital-intensive, generates ROIC below 5%, and is only beginning to emerge from a prolonged industry downturn. Its long-term outlook remains mixed due to the structural transition from 200mm to 300mm wafers and aggressive pricing by Chinese competitors. However, GRITEK’s public listing makes the segment a key source of hidden value within the group.

• Based on our model, we expect group revenue to grow at a 15.3% CAGR through 2028, operating profit at 19.5% CAGR, and net income at 19.8% CAGR, driven primarily by the increasing contribution of the higher-margin reclaim business. We forecast net income to exceed management guidance by 5%, 19%, and 23% in 2026, 2027, and 2028, respectively (management’s previous medium-term forecasts also proved conservative). While we expect free cash flow to remain deeply negative during the investment cycle, net debt should remain negative even if the entire expansion program is financed with debt.

• We see two key risks. First, the company may gradually lose share at TSMC in the fastest-growing leading-edge nodes to Taiwanese competitor Phoenix Silicon International. Second, capital allocation remains a concern. While the core reclaim business is capacity-constrained and generates ROIC above 20%, management continues searching for a “second cash cow” by deploying capital into businesses earning only 3-5% returns, diluting the group’s overall return on invested capital.

• Even after accounting for these risks, the valuation appears exceptionally attractive — both on conventional relative valuation metrics and on common sense. RST trades at a substantial discount to direct peers such as Phoenix Silicon (8028 TT), Kinik (1560 TT), and Scientech (3583 TT) despite offering comparable business quality and growth prospects. Moreover, RST’s 39.1% stake in publicly listed GRITEK alone was worth ~$2.9 billion as of July 29, while the entire market capitalization of RS Technologies stood at only ~$926.7 million. In other words, the market is effectively assigning a negative value to the company’s highly profitable and rapidly growing wafer reclaim business.

Subscribe now

The Company

Semiconductor demand seemed inexhaustible. A technology that had once been confined to engineering departments and research laboratories suddenly became indispensable to virtually every product, every industry, and every economy aspiring to compete in the future. Capital poured into the sector. Companies raced to build new capacity, struggling to keep pace with incoming orders. An entire nation came to believe it held the key to the next technological era. Analysts spoke of a paradigm shift, newspapers wrote about the race for global leadership, and investors watched fortunes being created in real time.

This is not the story of today’s artificial intelligence boom. It is the story of Japan in the 1980s. It was during this period that Japanese memory manufacturers rose to dominate the global DRAM market, while demand for silicon — the industry's fundamental raw material — grew exponentially.

Against this backdrop, Japan's traditional industrial conglomerates began searching for entry points into this young and promising market. For Rasa Industries, historically a chemicals and materials company, the move into electronics was not a one-off bet but a long-term strategic transition. The company launched high-purity red phosphorus for electronics in 1972, entered the gallium market in 1981, and in January 1984 became the first company in Japan to enter the silicon wafer reclaim business, making it one of the industry's global pioneers [1].

The economics of wafer reclaim are straightforward. Manufacturing semiconductor chips requires hundreds of process steps, with production quality verified at each stage using test wafers. Because prime wafers — the wafers ultimately used to manufacture chips — are expensive, it is far more economical to use reclaimed wafers for process monitoring.

By the early 2000s, the semiconductor industry was entering another transition. Following the prolonged downturn of the 1990s, chipmakers began migrating from 8-inch to 12-inch wafers (from 200 mm to 300 mm). The rationale was simple: larger wafers produce more dies per wafer, lowering the manufacturing cost of each chip. Between 2007 and 2008, Rasa invested approximately ¥5 billion to build a 12-inch wafer reclaim line with a capacity of roughly 100,000 wafers per month [2].

As is often the case in cyclical industries, timing proved unfortunate. The global financial crisis following the collapse of Lehman Brothers severely impacted semiconductor demand, fab utilization, and reclaim service pricing. In interviews conducted years later, RS Technologies' management described the situation candidly: Rasa was unable to absorb both the downturn and the burden of its recent investment, and in 2010 decided to exit the business altogether. The assets put up for sale included the Sanbongi wafer reclaim facility, along with its relatively new 12-inch reclaim production line.

It is at this point that Ho Nagayoshi — future founder, CEO, and largest shareholder of RS Technologies — enters the story. Ho was born in 1970 in China’s Jilin Province. His grandmother came from Japan’s Fukui Prefecture, making him a third-generation Japanese descendant. He moved to Japan in August 1988 at the age of sixteen, immediately after graduating from high school. After completing graduate studies in Japan, he founded Eiki Shoji in 1998, a company specializing in recycling and international trading of used computers, office equipment, and industrial machinery.

Eiki Shoji established a relationship with Intel: Intel’s irreclaimable silicon wafers, which had become waste, had to be disposed of, and Eiki Shoji purchased them for recycling — this marked its first connection to the wafer industry. When Rasa announced its exit from the business in 2010, Eiki Shoji became one of the bidders for the assets [4]. Three overseas buyers also competed for the assets. According to participants in the negotiations, they were willing to pay considerably more for the technology. Nevertheless, Ho’s group ultimately prevailed after committing to continue operations at the site, lease the factory from Rasa, and preserve manufacturing in Sanbongi [5].

Public sources are remarkably consistent regarding the economics of the transaction. Assets into which Rasa had invested an estimated ¥10-20 billion (~$125-250 million at the average 2011 exchange rate) were acquired for roughly ¥400 million — only $4.8-5.0 million [3].

Later, RS director Satoru Endo would say:

“It was a transfer made to facilitate Rasa’s exit after the Lehman shock, and for us it was a fortunate coincidence—we were able to start the business without bearing the burden of enormous capital expenditures.”

Under the terms of the transaction, only the equipment was transferred. The buildings at Sanbongi were leased, and 55 former Rasa employees who had been laid off were rehired. By that point, Rasa’s entire customer base had already moved to competitors, and the company started with zero revenue. Thus, RS Technologies was established in December 2010.

RS Technologies commenced operations in January 2011. However, in March, the Tohoku earthquake — the strongest in the country’s history, known as the Great East Japan Earthquake — struck. The company was able to reach the break-even point only in the second half of 2012 [6].

The company’s next step was expansion into Taiwan. Ho personally went to TSMC to offer reclaimed wafers. However, TSMC’s position was firm: the product quality was good, but the delivery cycle was too long — three weeks; if it could be reduced to two weeks, they could begin discussing orders. In February 2014, RS established its Taiwanese subsidiary, and in December 2015 completed its factory in Tainan. In its 2015 presentations, the company stated that the new Taiwanese facility had an initial capacity of 100,000 12-inch wafers per month. The company also became publicly listed that same year.

In March 2015, RS went public; in June 2015, it completed the eighth building at its Sanbongi plant, and following the IPO, it began rapidly expanding capacity. Toyo Keizai writes that in 2015 alone, the company invested a total of ¥7.7 billion (~$63-64 million) in the reorganization and expansion of the Sanbongi facility and in the construction of its new factory in Taiwan [7].

The second strategic turning point was the company’s entry into the prime wafer market. In December 2017, RS signed a joint venture agreement with China’s GRINM and Fujian Cangyuan, allowing it to enter the prime wafer market for the first time. Shandong GRITEK manufactures and sells 5-, 6-, and 8-inch wafers (125, 150, and 200 mm), and since 2020, through the SGRS joint venture, has been conducting research and development aimed at the mass production of 12-inch (300 mm) wafers.

At the same time, RS began assembling a broader portfolio of assets, including a business manufacturing quartz glass components and parts, the distribution of power semiconductors, the production of electrolyte for vanadium redox flow batteries, and other businesses.

Today, RS Technologies is one of the leaders in the small, capital-intensive, yet critically important silicon wafer reclaim market, with a broad portfolio of additional assets. Ho Nagayoshi remains the company’s CEO and controlling shareholder, owning 43.87% of the shares (8.04% directly and 35.83% through RS TECH Hong Kong). In total, insiders and affiliated parties own 50.59% of the company.

Wafer Reclaim Business (~35% of Revenue and ~64% of Operating Profit)

Two types of silicon wafers are used in semiconductor manufacturing: prime wafers and test wafers. Prime wafers are the highest-quality silicon substrates, manufactured from single-crystal silicon and intended for the fabrication of integrated circuits, processors, and memory chips. Test wafers are consumable silicon substrates used at semiconductor fabs such as TSMC, GlobalFoundries, and Intel for equipment qualification, process monitoring, and experimentation.

After use, a test wafer is not discarded but reclaimed and reused. Following a process run, the wafer is covered with process films, contaminants, and microscopic defects — but if these are carefully removed, the wafer becomes suitable for reuse. According to RS Technologies, test wafers account for about 20% of all wafers used in semiconductor manufacturing, and roughly 80% of them are reclaimed wafers. Each wafer can typically go through the reclaim process up to approximately ten times before it becomes too thin and must be scrapped.

Fabs prefer reclaimed wafers because they are significantly less expensive than new ones. According to a Former Manager at GlobalFoundries, a reclaimed wafer costs $20-30 per wafer, compared with around $100 for a new one [20]. A Former Manager at Phoenix Silicon, RS Technologies’ main competitor, provides a similar estimate:

Interviewer

Do you know the pricing, how much it costs for one wafer reclaim cycle for each of these nanometers?”

Former Manager at Phoenix Silicon

The business department is definitely more in charge of the pricing. The exact numbers are company trade secrets. I can provide you some ranges. In general, the wafer reclaim price is 1/3 of the prime wafer price. It will change depending on some factors including the particle size requirement, the outer volume by that customer. Those things will change the actual price.

Interviewer

What would you categorize as the price of a prime wafer here?

Former Manager at Phoenix Silicon

In general, a prime wafer costs $100-$120, but if we’re looking at the really advanced nodes it will be quite a bit more expensive. Based on publicly available information, I believe those more advanced process wafers are two to three times as expensive as the general wafer price.

In some cases, reclaimers purchase wafers outright and resell them after cleaning. However, the majority of revenue is generated under a service model. Ownership of the wafer remains with the customer, while RS earns revenue by providing the reclaim service. The reclaim process is as follows:

1. Shipment. The fab collects used test wafers and ships them to the reclaim supplier in batches.

2. Incoming inspection and sorting. The wafers are classified by diameter, film type, and degree of wear, as some are already unsuitable for reclaim because they have become too thin. At this stage, 10-20% of the wafers are rejected.

3. Film removal and wet cleaning. Deposited films and contaminants are removed through chemical cleaning.

4. Chemical Mechanical Planarization (CMP). The wafer surface is restored to a condition comparable to that of a new wafer while maintaining geometry and thickness within specification.

5. Final inspection and return to the customer. For high-grade wafers, the reclaim yield is typically 80-90%. For medium-grade wafers, it is 70-85%, while for mechanical-grade wafers, it is 60-75%.

When the fab and the reclaim supplier are located nearby, the entire cycle — from sorting to return — typically takes 1–2 weeks. With regional transportation (for example, Taiwan to Japan), the cycle extends to 3-4 weeks, with logistics costs of ~$2-4 per wafer. Intercontinental shipments (for example, the U.S. to Asia) take 6-8 weeks and add roughly $8-10 per wafer in logistics costs.

Lead time and logistics are important considerations for customers. While wafers are in transit and undergoing the reclaim process, they remain tied up in the supply chain, requiring customers to maintain buffer inventory throughout the entire cycle. This represents additional working capital tied up in inventory. This is likely what explains TSMC’s firm stance on delivery cycle time. As one Consultant in Semiconductor Wafer explains:

“When you do the wafer reclaim... there’s a lot of wafer accumulated in the whole supply chain. In RS Technologies from door-to-door means day one leave the TSMC door until the day that arrive TSMC door again, it takes around like three to four weeks... It means on average sent to RS Technologies, you accumulate four weeks wafer volume in the supply chain. It means money. However, if you use [a local supplier], it takes around just one week. Four weeks versus one week, the money accumulated right there is significant, especially when TSMC is big enough.”

Wafer Reclaim is the core business of RS Technologies, accounting for ~35% of revenue and ~64% of operating profit (or ~75% of operating profit excluding subsidies received by the Prime Wafers segment). The company’s production capacity is distributed as follows:

Sanbongi Plant (Japan)

• 8-inch: 150,000 wafers/month;

• 12-inch: 340,000 wafers/month;

• Planned to increase capacity to 520,000 wafers per month (+53%) by 2028.

Tainan Plant (Taiwan) – operated through the subsidiary RSTEC Semiconductor Taiwan

• 12-inch: 300,000 wafers per month;

• Planned to increase capacity to 570,000 wafers per month (+90%) by 2028.

China Plant

• 12-inch: 50,000 wafers per month;

• Planned to increase capacity to 100,000 wafers per month (+100%) by 2028.

According to management, both the Japan and Taiwan plants are operating at 100% utilization due to the AI boom. Moreover, the company has stated that there are periods when the facilities have no available capacity at all, operating 24 hours a day in three shifts. As a result, RS Technologies is forced to turn away customers — indicating not only full utilization, but also unmet demand.

RS Technologies plans to invest ¥52 billion (~$324.6 million) in these facilities to increase their combined monthly capacity to 1.19 million wafers by 2028. Notably, the expansion in Taiwan will be more aggressive, which is particularly important in the context of supply chain localization for TSMC.

The segment has two key customers: Sony (~20% of the group’s total revenue) and TSMC (~18%). Together, they account for the vast majority of the company’s wafer reclaim revenue.

The wafer reclaim market is a small oligopolistic market, with the four largest players controlling about 70% of global capacity. RS Technologies is one of these market leaders, with management estimating its market share at approximately 31%. However, based on our analysis, the company’s actual market share may be lower, as discussed in the Risk Factors section.

It is also worth noting that several industry experts describe the wafer reclaim business as countercyclical, since monitor wafers and dummy wafers are essential consumables required to maintain manufacturing yield, regardless of end-market chip demand.

Within its Wafer Reclaim business, the company has two key growth drivers: first, the construction of new semiconductor fabs around the world; and second, the transition to more advanced process nodes.

New Semiconductor Fabs Worldwide

Every new semiconductor fab represents a new, structural source of demand for both new and reclaimed wafers. According to RS Technologies, more than 60 new fabs are currently under construction worldwide, including 11 in Japan and 9 in Taiwan.

Against the backdrop of the AI boom, capital expenditures across the industry continue to accelerate. In January 2026, TSMC raised its full-year capital expenditure guidance to $52-56 billion, up from $40.9 billion the previous year, and following its most recent earnings report, increased the guidance further to $60-64 billion. According to Goldman Sachs, TSMC’s manufacturing capacity is expected to expand by approximately 25% by the end of 2028 [21].

Memory manufacturers are also expanding capacity aggressively in response to the severe supply shortage. According to Barclays, total DRAM wafer supply is expected to increase by ~70% by 2030, implying a CAGR of ~11%. These projections are indirectly supported by Sumco, one of the world’s largest manufacturers of prime wafers, which forecasts demand for 12-inch wafers to reach 10 million wafers per month by 2028. For comparison, demand stood at 7.8 million wafers per month in Q1 2026 [9].

On May 8, 2026, Sony and TSMC signed a memorandum of understanding to establish a strategic partnership for the development and manufacturing of next-generation image sensors. Most of the new production capacity will be deployed at the new fab currently under construction in Koshi, Kumamoto Prefecture, Japan. Sony will hold a majority and controlling stake in the joint venture.

The JASM Phase 1 and Phase 2 (Japan Advanced Semiconductor Manufacturing) facilities are expected to support process nodes ranging from 40nm to the leading-edge 3nm node. Combined production capacity will reach approximately 100,000 wafers per month. Given the proximity of RS Technologies’ operations, the company is likely to capture the largest share of wallet from this project.

Transition to More Advanced Process Nodes

For those familiar with TSMC, it is no secret that with each new generation of chips, the company seeks to increase the number of transistors that can be integrated onto a single wafer. This is known as process node scaling in accordance with Moore’s Law, while the progression from one generation to the next is referred to as the transition to more advanced process nodes (advanced nodes).

Historically, this progress was measured in nanometers, which once reflected the actual physical dimensions of a transistor’s gate. Today, these figures have largely become commercial generation labels, but the underlying engineering challenges remain fundamentally the same.

Today, the most advanced process nodes (7nm, 5nm, and 3nm) account for 39% of TSMC’s manufacturing capacity and 74% of its revenue. The share of advanced nodes is expected to exceed 60% of total capacity by the end of 2028.

The transition to more advanced process nodes is a key driver of demand for reclaimed wafers, because as transistor density increases, the number of lithography, etching, and deposition steps also rises, along with the number of test, monitor, and dummy wafers required for each production wafer. In other words, demand for wafer reclaim increases even without the construction of new fabs, simply as a result of migration to more advanced process nodes.

The scale of this effect is illustrated by a Former Manager at Phoenix Silicon. According to the manager, the intensity of reclaimed wafer usage increases significantly as fabs transition to more advanced process nodes, even when fab capacity remains unchanged.

Interviewer

Let’s see. In terms of the overall demand for wafer reclaim, let’s say you have a fab doing 50,000 wafer starts per month. How much wafer reclaim demand are you going to have? I think this was one of the questions but breaking it out by nanometer as well would be helpful.

Former Manager at Phoenix Silicon

I’m not that involved in capacity planning. I can just give you some rough numbers. For 28 nm, if the wafer start number is 50,000 per month, then we will need 40,000 reclaim wafer. However, that doesn’t mean the absolute number of reclaim wafers we will be using because each wafer can be reclaimed five to 10 times.

If we divide that 40,000 by 10, the actual reclaim wafer that will be used in that month will be 4,000 pieces.

However, they need to perform wafer reclaim for 40,000 times to be able to meet the requirement. For three nm, for the same capacity, we will use the wafer reclaim process for 115,000 times. For the two nm, there is a rough calculation. They suspect in the future, we will use 135,000 times of the wafer reclaim.

According to a Consultant in Semiconductor Wafer, the transition to more advanced process nodes can increase demand for reclaimed wafers by 3-4x, even with the fab’s manufacturing capacity remaining unchanged:

Interviewer

Just on the advanced nodes for TSMC, can you talk me through the shift to advanced node and what the impact is on the reclaim business? Obviously, the understanding is because its more advanced node the dummy ratio or the test wafer ratio is going to be higher, there’s more wastage.

Consultant in Semiconductor Wafer

Yes, I have a number.

Interviewer

Yes, this is obviously positive for the volume of reclaim. I just wanted to understand, is it more the benefit for reclaim? Is it more that it’s a higher dummy ratio or more that the percent of dummy ratio that is reclaimed is higher? Is it both?

Consultant in Semiconductor Wafer

Let me share the number with you. All the data are for 12-inch. Every 1,000 product wafers, they need 600 dummy wafers and monitor wafers. If we combine everything into like every 1,000 require 600 for the reclaimed wafer. However, when we move to the 28 nm, every 1,000 requires around 1,500-2,000 reclaimed wafers. The ratio from one to 1.6, 1.5 to 2.0.

When you move to the like five nm, it can move to more than 1,000-2,500, which means every 1,000 product wafer you require like 2,500 recurring wafers, which means even TSMC do not establish the new fab but their technology should not keep migrating from 65 nm to 28 nm to seven nm.

Even the same fact of 50,000 wafer per month capacity, the recurring way the recurring wafer demand increase three to four times easy. This is the power of the migration. Meanwhile TSMC keep building the new fab as you know, so two factors multiplied together.

The phrase “two factors multiplied together” neatly captures the essence of the investment thesis: migration to advanced process nodes, which require 3-4x more reclaimed wafers per fab, is compounded by the growing number of semiconductor fabs being built worldwide. This shift in the industry mix toward advanced nodes is precisely the growth driver behind RS Technologies’ plan to expand its reclaim capacity by 72% by 2028. According to management, the company’s capacity expansion investments have been coordinated with its customers’ production plans.

Prime Wafer Business (~26% of Revenue and ~26% of Operating Profit)

If reclaimed wafers are the semiconductor fab’s consumable, then a prime wafer is the monocrystalline silicon substrate on which integrated circuits are actually built. Its journey begins with the growth of a single-crystal silicon ingot using the Czochralski process. The ingot is then sliced into wafers, ground, mirror-polished, and, where required, an epitaxial layer is deposited. The entire manufacturing process — from crystal growth to the packaging of finished wafers — is well illustrated in the following video.

The 12-inch (300 mm) wafer segment accounts for the majority of the market — approximately 75% by value. These wafers are used primarily in the production of logic and memory chips. 8-inch and smaller wafers (5-, 6-, and 8-inch, or 125 mm, 150 mm, and 200 mm) are primarily used for power semiconductors, analog integrated circuits, discrete devices, and sensors.

RS Technologies entered the prime wafer market in 2018 through a joint venture with the Chinese state-owned General Research Institute for Nonferrous Metals (GRINM). A special-purpose vehicle, Beijing GRINM RS Semiconductor Technology Co., Ltd. (BGRS), was established. BGRS owns a 30.8% stake in the publicly listed GRITEK (Ticker: 688432 CH), which in turn owns 85% of the operating company Shandong GRITEK.

RS Technologies owns 45% of BGRS and a 25.2% stake in publicly listed GRITEK directly. As a result, RS Technologies’ effective ownership interest in GRITEK is 39.1% (25.2% direct ownership plus 13.9% indirect ownership through BGRS).

Although RS Technologies’ total economic interest is 39.1%, the company controls GRITEK and therefore consolidates its financial results. RS Technologies paid ~$47 million for its 45% stake in BGRS. The ownership structure also includes Fujian Kuramoto, a company affiliated with Ho Nagayoshi.

This complex ownership structure serves a single purpose: to ensure that the Chinese business is treated as a domestic enterprise and therefore qualifies for Chinese government subsidies. It is worth noting that these subsidies account for approximately 13% of GRITEK’s ordinary income [11].

A separate joint venture, Shandong GRINM RS Semiconductor Materials Co., Ltd. (SGRS), was established for the 12-inch (300 mm) wafer business, in which the publicly listed GRITEK holds a 28% stake. As a result, RS Technologies’ effective economic interest in the 12-inch prime wafer business is 11.2%.

Within the segment, 56% of revenue is generated by sales of 5-, 6-, and 8-inch prime wafers. The production capacity is distributed as follows:

• 5-inch: 50,000 wafers/month;

• 6-inch: 200,000 wafers/month;

• 8-inch: 250,000 wafers/month.

Approximately 32% of segment revenue comes from silicon components, including consumable parts used in wafer etching equipment — such as shower plates, polycrystalline silicon rings, focus rings, and other silicon components used to distribute process gases, control plasma, and ensure etch uniformity. These components are gradually eroded during operation and must be replaced regularly, creating a recurring revenue stream tied to semiconductor fabs’ capacity utilization.

A further 12% of revenue comes from sales of complete silicon ingots used to manufacture wafers, including ingots produced using the float-zone process.

In our view, the outlook for this business is mixed. On the one hand, China continues to aggressively promote the localization of its semiconductor supply chain. For example, the government supports the adoption of domestically produced wafers even when their yield is lower than that of international competitors, fully compensating chip manufacturers for the resulting losses.

On the other hand, the silicon wafer industry has experienced a deep cyclical downturn globally in recent years. While demand for 12-inch wafers has begun to recover amid the current boom in advanced process nodes, the recovery in the mature-node segment (consumer and industrial electronics) remains weak, with wafer prices continuing to decline due to the persistent imbalance between supply and demand. Customers are also continuing to normalize 8-inch wafer inventory levels.

The situation is further complicated by the ongoing migration of many semiconductor manufacturers from 8-inch to 12-inch wafers. A 12-inch wafer has 2.25x the surface area of an 8-inch wafer, while processing costs are only 1.3-1.5x higher. As a result, the cost per die declines by ~30-40% following the transition.

Fabs built during the 1990s and 2000s continue to use 8-inch wafers because their manufacturing equipment was designed for that wafer size, while the cost of converting an entire fab exceeds the potential savings achieved on a per-die basis. However, as older manufacturing capacity is retired, the transition to 12-inch wafers is expected to continue.

In the West, this transition began many years ago, with demand for 12-inch wafers surpassing demand for 8-inch wafers as early as 2007 [12]. Companies such as Texas Instruments are targeting 80% of production on 12-inch wafers by 2030 [13]. Today, measured by total wafer surface area, 8-inch wafers account for only around 20% of the global market.

Similar trends are evident in China. For example, at SMIC, the country’s largest foundry, more than 80% of revenue is generated from 12-inch wafers. Nexchip, UNT, Silan, and CR Micro are also actively expanding 12-inch capacity for the production of analog and power semiconductors [20].

Against this backdrop, the following comments from SUMCO’s management during the Q2 2025 earnings call are particularly noteworthy:

“The trend reflects the combination of this China impact as well as the fading of demand for 200-millimeter wafers. The fading of demand is the result of weaker economic conditions as well as the transitioning of some applications to 300-millimeter wafers. The fall in demand as a result of weaker macro conditions should recover when economic conditions improve. But as the loss of demand on the back of applications transitioning to 300-millimeter represents a structural change, demand here will not come back.”

Also noteworthy are the comments from the Chief Commercial Officer of NSIG, China’s largest silicon wafer manufacturer, which accounts for ~6% of the global market:

Local competition from China for the power application part also for the logic part, I think other parties, they’re transition into 300 mm, especially in China. There’s been quicker transition to 300 mm than it was expected in this power and legacy market area. Even, let’s say, 2021, it was still expected that it’s much lower, the speed of the transition. Even some of the processes that it’s not expected to be transferred to 300 mm, but it has happened quicker especially there in China, also in Europe, and with some of the U.S. players too.

In Taiwan, for example, TSMC’s even talking about reducing the 200 mm, totally cutting down closing the 150-mm line but reducing 200 mm manufacturing space and this way, also volume. That’s affecting the transition for the volumes and the China local suppliers. I think these are two very important elements there. There have been rumors about that some of these big players are even cutting shutting down some of their 200 mm manufacturing capabilities factories.

I have not heard anything specific related to SUMCO on this field. If you look at the public numbers for the 200-mm manufacturing, it is lower than it used to be. In most of this data, China players are still missing. Missing the data from the China players, but the trend is not upwards. Also, I think that’s what SEMI forecasted, organization for semiconductors. The 200 mm will be in the lower growth percentages than the 300 mm.

RST/GRITEK’s management also recognizes the structural importance of 12-inch wafers. The company forecasts that demand will increase from 2.5-2.6 million wafers per month today to 4 million per month by 2030, with RST targeting a 7% market share. In 2026, the company plans to invest ¥3.80 billion ($23.95 million) to expand 8-inch wafer production capacity to 300,000 wafers per month, as well as ¥8.0 billion ($50.4 million) to increase 12-inch wafer capacity to 150,000 wafers per month. In 2027, it plans to further expand 12-inch capacity to 250,000-300,000 wafers per month, investing a substantial ¥17 billion ($107.2 million).

Investments in China are primarily made through the establishment and recapitalization of joint ventures in which RST holds ownership stakes.

RS Technologies’ prime wafer segment is a capital-intensive business whose core production assets are subject to ongoing technological obsolescence. As substrate requirements for cleanliness, flatness, and defect density become increasingly stringent with each successive process node, maintaining competitiveness alone requires continuous reinvestment. At the same time, the structural growth in demand within China is unevenly distributed. It is concentrated in the 12-inch segment serving advanced logic and memory, whereas GRITEK’s prime wafer business is focused on the 8-inch segment, where unit prices remain under structural pressure. To capture the faster-growing portion of demand, RST is forced to invest in the 12-inch business through SGRS, which has yet to enter mass production and is currently loss-making. Meanwhile, Chinese wafer manufacturers are aggressively competing on price while simultaneously shifting their own production from 8-inch to 12-inch wafers.

Assessing the potential and quality of this business is ultimately a matter of debate. Nevertheless, the prime wafer business represents an important part of the RS Technologies investment thesis. As noted earlier, GRITEK is a publicly listed company trading under ticker 688432 on the STAR Market of the Shanghai Stock Exchange. As of July 29, 2026, GRITEK’s market capitalization stood at RMB 49.86 billion (~$7.36 billion). Accordingly, RS Technologies’ 39.1% stake is worth ~$2.9 billion, while RST itself is valued at only $926.7 million.

It is worth noting that following GRITEK’s IPO on the STAR Market, RST became subject to a 36-month lock-up period. In Q1 2026, RS Technologies reduced its ownership stake by 1%, selling shares in the open market. We spoke with the company’s Investor Relations team. According to management, they recognize the company’s undervaluation and understand the levers available to unlock shareholder value.

Semiconductor-related Equipment & Materials Business (~39% of Revenue and ~10% of Operating Profit)

This is the company’s most heterogeneous segment, assembled through a series of M&A transactions. While it is formally the largest segment by revenue, it contributes only around 10% of operating profit.

The largest subsegment by revenue is RS Precision Device Huizhou (RSPDH), the Chinese business manufacturing optical pickup modules and in-vehicle camera modules, acquired from Sony in the fall of 2024. In 2025, this business accounted for 50% of segment revenue.

The second-largest business is DG Technologies, a manufacturer of consumable components for wafer etching equipment, acquired in 2019. The third is Union Electronics, a distribution company supplying Minebea power semiconductors and Renesas microcontrollers, acquired in 2018. Together, these two businesses account for approximately 10-15% of segment revenue.

A brief history of the acquisitions follows:

• 2018 — Union Electronics Solutions. A trading company supplying Minebea power semiconductors and Renesas microcontrollers. At the time of the acquisition, the company had ¥27 million (~$244.5 thousand) in equity.

• 2019 — DG Technologies. RS acquired 100% of DG Technologies, a manufacturer of consumable components for semiconductor equipment and processing systems, for ¥906.5 million (~$8.32 million at the time of the transaction). At the time of the acquisition, DGT generated ¥1.27 billion (~$11.65 million) in revenue and ¥82 million (~$0.75 million) in operating profit.

• October 2023 — LE System. The company is Japan’s only manufacturer of electrolytes for vanadium redox flow batteries (VRFBs). The global VRFB market is relatively small, with an estimated size of $580 million. Nevertheless, there is a possibility that VRFBs could capture a meaningful share of the backup power market for data centers [14].

• 2024 — Sony assets. In December, RST acquired 100% of Sony Precision Devices (Huizhou), Sony’s Chinese manufacturer of optical pickup modules and automotive camera modules. Optical pickup modules are used to read data from CDs and DVDs — a structurally declining business — whereas automotive camera modules used in ADAS represent a growing market.

Expanding through the acquisition of unrelated businesses is generally not an optimal capital allocation strategy. Companies pursuing this approach typically trade at a meaningful discount to peers focused on organic growth. Likely, RS Technologies’ active M&A strategy has indeed contributed to a valuation discount.

It is worth noting that management has not overpaid for the assets it has acquired. For example, the Sony assets appear to have been purchased at a discount to book value (the acquired company’s equity was ~$81.4 million), allowing RS to recognize approximately ¥1.5 billion in other income, increasing net profit by 22.6%. The group’s balance sheet contains only ¥669 million ($4.2 million) of intangible assets against ¥205.2 billion ($1.29 billion) of total assets, indicating that the company carries virtually no goodwill.

Despite operating several sizeable subsegments, management’s primary focus remains on the dry etching consumables business, which, according to the company, is currently operating at 100% utilization:

“Utilization rate is 100%. I mean, their demand is very, very strong… they are running at full capacity. And the top line is growing every month… Our major customers are quite in a good position, both TSMC and Tokyo Electron. It is growing.”

Dry etching consumables represent a relatively small but growing market, estimated at ¥150 billion (~$922.5 million). RST’s current market share is only a few percent, while management is targeting 10% in the medium term and 30% over the long term, implying multiple-fold growth of the business.

The market is highly competitive. In the segment for silicon electrodes and etch rings, RST competes with manufacturers such as Techno Quartz, ThinkonSemi / Fujian Dynafine, Chongqing Genori, Ruijiexinsheng, KC Parts Tech, SICREAT, and Coma Technology. These are complemented by larger specialized suppliers of quartz and silicon components, including Shin-Etsu Quartz, Tosoh Quartz, Mitsubishi Materials, Hana Materials, SK Enpulse, Ferrotec, and CoorsTek, among others.

Nevertheless, RS Technologies has a modest competitive advantage through cross-selling. The company already serves a global customer base through its wafer reclaim business, allowing it to offer dry etching consumables to the same semiconductor fabs. As noted in the quotation above, the largest customers for this business are TSMC and Tokyo Electron.

It should be noted, however, that shareholders’ economic interest in DG Technologies has been substantially diluted. In January 2026, RS Technologies transferred a 70% stake in DGT to its own consolidated Chinese subsidiary. As a result, GRITEK now owns 70% of DG Technologies, while RST retains the remaining 30%. DG Technologies remains a consolidated subsidiary of RST, and its operations continue to be reported within the Semiconductor Equipment and Materials segment. However, the share of net income attributable to the parent company’s shareholders and included in the consolidated financial statements has declined from 100% to 57%, consisting of RST’s direct ownership interest together with its indirect ownership interest held through GRITEK.

Financials and Valuation

Due to the limited level of disclosure typically provided by Japanese companies, our ability to analyze RS Technologies’ unit economics is constrained. Nevertheless, the wafer reclaim business is sufficiently straightforward to allow for reasonable assumptions regarding its medium-term trajectory, particularly given the exceptionally high capacity utilization currently observed across the industry.

As discussed earlier, wafer reclaim is RS Technologies’ core business. Over the past five years, segment revenue has grown at a 19.2% compound annual growth rate, while operating margins have consistently remained in the high-thirties.

Segment revenue is driven by three key variables: a) available capacity, b) capacity utilization, and c) the average reclaim price per wafer. Ultimately, all three factors are determined by underlying demand, which we expect to accelerate over the coming years.

Despite serving a relatively limited number of customers, RS Technologies has increased revenue per unit of capacity by an average of 7.9% annually over the past five years. We expect this metric to grow at an average annual rate of 4.9% between 2026 and 2028, with the anticipated moderation reflecting the record pace of capacity additions.

We also expect the company’s capacity utilization to remain exceptionally high, supported by continued expansion of semiconductor manufacturing capacity and robust demand for its services.

As a result, we expect wafer reclaim revenue growth to accelerate from the current mid-teens to the high-twenties over the next several years. Segment operating margins are likely to come under pressure due to a significant increase in D&A expense. Nevertheless, we forecast operating profit to increase by approximately 77% between 2025 and 2028.

Until recently, the prime wafer business had been going through a deep downcycle. In 2024, despite a 36% increase in shipment volume, revenue grew by only 3.8%, as average selling prices declined by ~25%. In 2025, net sales totaled ¥20.8 billion, up 2.2% year over year, driven by a more than 20% year-over-year increase in 8-inch wafer shipment volume. On the other hand, the average unit price declined by ~10% year over year.

Notably, 2025 already showed the first signs of market stabilization. Average selling prices for power semiconductor wafers — the company’s core product — declined only modestly. The largest price declines were seen in general-purpose products, such as IGBT and memory wafers, which account for 20-30% of the company’s product mix.

In Q1 2026, prime wafer segment revenue increased by 32% year over year, while operating profit grew by 22% year over year, supported by ASP stabilization and higher shipment volumes. We believe RST is at the beginning of an upcycle in this segment. ASP for 8-inch wafers is expected to increase by 10-15% by the end of 2026. This is driven by extremely high capacity utilization at Chinese fabs (above 110% at industry leader Hua Hong) and rising prices for power semiconductors, such as IGBTs.

In addition, the share of production capacity dedicated to the structurally declining 5-inch and 6-inch wafer segments is expected to decrease from 41% to 32.5% by year-end, as RST plans to expand 8-inch capacity by 20% and 12-inch capacity by 50% year over year. Even assuming only a 10% increase in ASP by year-end, segment revenue growth could exceed 35%. It is worth noting that the sell-side consensus currently forecasts 26% year-over-year revenue growth for 2026.

The Semiconductor Equipment and Materials business is likely to lag. Last year, approximately half of segment revenue came from RSPDH, the business acquired from Sony, generating ¥15 billion in revenue versus management’s original guidance of ¥10 billion. The outperformance was driven by earlier-than-expected deliveries of optical pickup modules under customer orders inherited as part of the acquisition. However, those accelerated shipments in 2025 effectively pulled forward revenue from future periods. As a result, we expect RSPDH revenue to decline in 2026. RST management is guiding for RSPDH revenue of ¥10 billion for the full year [15].

Weakness at RSPDH is likely to be partially offset by growth at DG Technologies and Union Electronics, as both businesses appear to be at the beginning of an upcycle. Nevertheless, we model a 13.7% year-over-year decline in segment revenue, assuming ¥4.5 billion of revenue at DGT and 3% growth across the remaining subsegments. We also forecast an operating margin of 5.3%, in line with the segment’s six-year historical average.

According to our model, we expect the company’s total revenue to grow at a compound annual growth rate of 15.3% between 2025 and 2028, while gross profit is projected to grow at a CAGR of 14.8%. Despite a significant increase in depreciation expense, we expect gross margin to remain broadly stable throughout the forecast period, driven by the increasing contribution of the higher-margin wafer reclaim business, whose share of total company revenue is expected to increase from 35.9% to 46.5%.

Benefiting from operating leverage, we forecast operating profit to grow at a CAGR of 19.5%, while net income is expected to increase at a CAGR of 19.8%. Our model does not assume any additional acquisitions, despite the company’s relatively aggressive approach to M&A.

RS Technologies has presented its financial targets through 2028. According to our estimates, the company’s net income could exceed management’s guidance by ~5%, ~19%, and ~23% in 2026, 2027, and 2028, respectively. It is worth noting that management’s previous medium-term guidance, covering the periods through 2023 and 2026, also proved to be conservative [16] [17].

Despite continued growth in operating cash flow, we expect the company’s free cash flow to remain negative over the next several years. Nevertheless, RST’s net debt is likely to remain negative even if the planned investments are financed entirely through debt.

Valuation is the most important part of the investment case for the company. As discussed earlier, RS Technologies owns a 39.1% stake in GRITEK (688432 CH), a publicly listed Chinese manufacturer of silicon wafers. As of July 27, GRITEK’s market capitalization stood at RMB 49.86 billion (~$7.36 billion). Accordingly, RS Technologies’ 39.1% stake is worth ~$2.9 billion, while RST itself has a market capitalization of only $926.7 million. In other words, the market is assigning a negative value to the company’s profitable and rapidly growing wafer reclaim business.

The gap is equally pronounced from the perspective of a traditional comparable valuation. The wafer reclaim market is highly concentrated, with RS Technologies’ largest competitors being the Taiwanese companies Phoenix Silicon International (8028 TT), Kinik (1560 TT), and Scientech (3583 TT). The figure below compares RST’s forward valuation multiples (based on our estimates) with those of its peers (based on sell-side consensus). RS Technologies is trading at a substantial discount to its peers across valuation metrics.

It is worth noting that management is well aware of the company’s significant undervaluation — and says so explicitly. Investor Relations acknowledges that RST’s valuation multiples are materially below those of comparable companies and that senior management fully recognizes this. Moreover, the company’s CEO closely monitors the share price on a daily basis and, as a major shareholder, is personally dissatisfied with the company’s current valuation.

There are also multiple avenues to unlock shareholder value. In our view, the most straightforward approach would be to deconsolidate the prime wafer business by selling shares in GRITEK, followed by share repurchases or the payment of special dividends. Another effective option would be a spin-off of the wafer reclaim business. As a standalone company, this business could potentially be worth several times more than the current market value of RS Technologies as a whole.

Key Risks / Bear Case

Loss of Share at TSMC

RS Technologies is the global leader in the wafer reclaim market, with an estimated market share of roughly one-third. The issue is that this represents the market of yesterday. In the fastest-growing part of the industry — advanced process nodes at the world’s largest customer — TSMC — the company has been losing ground for several years to Taiwan’s Phoenix Silicon International (PSI).

RS Technologies denies that it has lost market share. However, the publicly available data suggests otherwise. Although PSI has historically been smaller than RST, it plans to increase its 12-inch wafer reclaim capacity from 630,000 wafers per month in 2024 to 950,000 wafers per month by 2026.

After conducting a series of expert interviews, we concluded that TSMC has increasingly favored the local Taiwanese supplier, particularly at advanced process nodes, for three reasons. First, quality control. Shipping wafers back to Taiwan by air after processing in Japan introduces edge defects and surface particle contamination, which is unacceptable for leading-edge manufacturing processes.

Second, logistics costs. Unlike prime wafers, which are worth approximately $80-100 per wafer and for which transportation costs are relatively insignificant, wafer reclaim is a much lower-value service, making logistics a meaningful component of the total cost.

Third — and most importantly — working capital. RS Technologies’ end-to-end door-to-door turnaround time is approximately three to four weeks, whereas PSI, located just three minutes from TSMC’s fabs, completes the cycle in about one week. Given TSMC’s scale, the additional three weeks of wafers tied up in the supply chain represent a significant amount of working capital.

One industry expert described the situation as follows:

Interviewer

When you talk about TSMC, obviously PSI are known to be doing the wafer reclaim business with TSMC particularly focused on their more advanced nodes, for example two nm. TSMC by all accounts used to use more a Japanese competitor and Phoenix has taken market share. My understanding is that they took market share because their Japanese competitor essentially just didn’t have enough capacity. What do you know about the situation with regard to TSMC, the reason for Phoenix’s market share gain and how they compare versus in the industry to the Japanese competitor?

Consultant in Semiconductor Wafer

Yes, RS Technologies. The couple reasons, number one is in the earlier stage like 15 years ago, RS Technologies are doing much better than PSI because PSI, if you have been inside their facility, they have a lot of problems. For example, many steps are manual and not as automatic as the RS Technologies. Also, their quality control and the EAR are not as good as RS Technologies as well.

However, there are a couple points. Number one is, recycle recurring wafer. TSMC always prefer the local manufacturing or local supply due to couple reasons. Number one is the quality control because when the ship to Japan and ship-back which means the quality control sometimes the shift can happen especially when logistics like an aircraft.

You take the passenger aircraft. Your life is sometimes got thrown in the crack. Similar situation happens during the wafer reclaimed specially ship-back. You ship to Japan, that’s fine. When you finish everything and ship-back, when the transportation is now well taken care then the forearm particle or the wafer edge chipping or crack can happen which means the quality control is always a problem.

This is why TSMC won the local supply and as close as possible. That’s the reason if you know that PSI headquarter to TSMC it’s just around by scooter maybe three minutes, two to three minutes, this is at most. In Taichung and in Thailand, PSI have the facility as well. As you can imagine in the past, TSMC can ship all the wafer from Taiwan to Japan.

How come they cannot just have one facility in Taiwan and ship all the wafer around Taiwan to that point? This is the proof that TSMC want PSI attached to their facility and do the local supply and shorten the supply chain because the 1) quality and 1) is the price. Since we can cut the transportation price, then it can reflect to the cost of the reclaim wafer.

The transportation you can almost ignore. However, for the reclaim wafer, it’s not high price. Unfortunately you have to try to set the cost as much as possible. The transportation cost is another thing as well. 3) is maybe the most important because when you do the wafer reclaim, which means the supply chain, there’s a lot of wafer accumulated in the whole supply chain.

The TSMC should wafer until the wafer go back to TSMC. In RS Technologies from door-to-door means day one leave the TSMC door until the day that arrive TSMC door again, it takes around like three to four weeks. I think you get a similar number as well. It means on average sent to RS Technologies, you accumulate four weeks wafer volume in the supply chain. It means money. However, if you use PSI, it takes around just one week.

Four weeks versus one week, the money accumulated right there is a significant, especially when TSMC is big enough. That is also the reason why the Micron Technology in Taichung, they use Kinik, the competitor of PSI as they work on Korean partner as well just because of the same reason. The money accumulated in the supply chain sent to RS Technologies four weeks. You can imagine the amount TSMC uses every single day. Four weeks versus one week is a significant difference. Quality and the shipping cost and all the total amount of money accumulated right there.

We have also learned that when TSMC decided to build its fabrication facilities in Arizona, it asked RS Technologies to establish a wafer reclaim facility there. The company declined. Management’s rationale was that labor and material costs in the United States are significantly higher, making it impossible to maintain the current level of profitability at a U.S. facility.

At first glance, this appears to be a disciplined capital allocation decision. From the perspective of customer retention, however, it meant declining an invitation to participate in the fastest-growing part of the supply chain of the company’s largest customer. Phoenix Silicon International (PSI) accepted that invitation.

As early as January 2025, PSI, while explaining its capacity expansion plans, explicitly cited the overseas expansion of one of its key customers — rapidly increasing production of leading-edge chips in the United States — as one of the main growth drivers. The company did not identify the customer by name, but at the time there was only one manufacturer producing 4nm leading-edge chips in Arizona.

We would also highlight the company’s response. We were told that PSI’s expansion plans were difficult for them to understand, that the competitor was not yet fully utilizing its existing capacity, and that TSMC would not, in any case, rely on a single supplier.

In our view, it is unlikely that one of the largest players in this highly concentrated market — one that has consistently gained market share in recent years — would undertake a major capacity expansion without coordination with TSMC, particularly given that RST itself invests only after securing customer commitments. The argument that TSMC will not rely on a single supplier is, strictly speaking, correct. However, it does not suggest that RST is in a position to increase its market share.

Capital Allocation

Capital allocation — one of the most important considerations for any disciplined investor — is one of the weaker aspects of the RS Technologies investment case.

The wafer reclaim business accounts for only one-sixth of the group’s assets while generating approximately two-thirds of operating profit. By contrast, the prime wafer business in China absorbs roughly two-thirds of the group’s assets yet contributes only about one-quarter of operating profit. Moreover, that profit includes Chinese government subsidies. Excluding these subsidies, the core wafer reclaim business would account for approximately 71-75% of the group’s operating profit.

During our discussion with the company, management explicitly acknowledged the difference in profitability between the two businesses. Return on invested capital (ROIC) exceeds 20% in the wafer reclaim business but is below 5% in the prime wafer business. At the same time, demand for reclaimed wafers is so strong that RS Technologies is forced to turn away some customers due to insufficient capacity. Nevertheless, management continues to allocate a significant portion of capital to substantially lower-return businesses.

A notable example is the energy project, to which the company committed ~¥8 billion (~$49 million) earlier this year. Commercial operation is not expected to begin until October 2029. Formally, the investment is justified by its vertical linkage to subsidiary LE System, which manufactures electrolytes for vanadium redox flow batteries. In practice, however, RS Technologies is entering a capital-intensive infrastructure business where the group currently lacks any obvious operational expertise, competitive advantage, or pricing power [18].

This capital allocation policy has resulted in a sustained decline in the group’s return on invested capital. While the trend may begin to reverse over the coming years as the wafer reclaim business accounts for a larger share of earnings, the underlying issue remains unchanged: capital allocation discipline continues to be one of RS Technologies’ key weaknesses. In our view, if management concentrated its resources on the company’s core business, the market would assign a materially higher valuation to RS Technologies.

The explanation we received, in our view, accurately reflects management’s underlying rationale. Because the wafer reclaim market is relatively small, the company is constantly searching for a “second cash cow.” This explains its acquisitions, expansion into prime wafer manufacturing in China, and investments in the energy sector. We believe this approach is misguided for three reasons.

First, it is difficult to argue that market size represents a genuine constraint when the company already controls roughly one-third of the market and is still unable to satisfy existing demand. The real constraint is production capacity, which can be addressed through additional investment in a business generating returns on invested capital above 20%.

Second, the wafer reclaim market is growing faster than one might expect based solely on semiconductor production volumes. The transition to more advanced process nodes dramatically increases the number of test wafers required for each finished wafer. As a result, this is not a mature market with limited growth potential, but rather a niche benefiting from a structural acceleration in demand.

Third — and most importantly — the objective of a public company is not to maximize its own size, but to maximize shareholder value. If high-return investment opportunities within the core business are genuinely exhausted, excess capital should be returned to shareholders, who can allocate it more efficiently themselves. Diversifying into businesses generating 3-5% returns does not solve the problem of a “small market” — it merely reduces the group’s overall return on invested capital.

Conclusion

RS Technologies has evolved from a distressed asset into one of the leaders in a small, capital-intensive, yet structurally growing oligopolistic market. The core of the investment case — the wafer reclaim segment — is a high-quality business operating at 100% capacity utilization, with long-term demand supported by two structural drivers: the global expansion of semiconductor fabrication capacity and the migration to advanced process nodes, which increases reclaimed wafer consumption by 3-4x for the same level of fab capacity.

The investment case is not without flaws, and we do not seek to downplay them. The most significant is the gradual loss of share at TSMC in the fastest-growing leading-edge process nodes, where Phoenix Silicon has been gaining ground. RST’s decision not to build a wafer reclaim facility in Arizona appears disciplined from a margin perspective, but in practice it meant declining the opportunity to participate in the fastest-growing part of the supply chain of its largest customer — an opportunity that Phoenix Silicon accepted. The second weakness is capital allocation. Despite operating a core business constrained only by capacity and generating ROIC above 20%, management continues, in its search for a “second cash cow,” to allocate capital to businesses earning 3-5% returns — including the Chinese prime wafer business and an infrastructure project related to vanadium redox flow batteries (VRFBs) — thereby diluting the group’s overall return on capital.

Even after fully accounting for these negative factors, however, RS Technologies’ valuation appears unjustifiably low — both from the perspective of comparable company valuation and from the standpoint of simple economic logic.

On a relative valuation basis, the company trades at a substantial discount to its closest peers despite a core business of comparable or superior quality and growth. From a fundamental perspective, the disconnect is even more striking: RST’s 39.1% stake in publicly listed GRITEK alone is worth approximately $2.9 billion, while RS Technologies’ entire market capitalization is only around $926.7 million. In other words, the market is assigning a negative value to a profitable, rapidly growing wafer reclaim business generating ROIC above 20% and operating at full capacity utilization.

Such a valuation dislocation is unlikely to persist indefinitely. Management — and particularly CEO and controlling shareholder Ho Nagayoshi — recognizes the company’s undervaluation and has several obvious levers to unlock shareholder value. Logical steps could include deconsolidating the prime wafer business through the sale of GRITEK shares, followed by share repurchases or special dividends, or alternatively a spin-off of the wafer reclaim business, which, as a standalone public company, could plausibly be worth several times the current market capitalization of RS Technologies. Even under a conservative scenario in which market share losses at TSMC continue, and capital allocation remains suboptimal, the valuation discount is simply too large to justify. At today’s share price, investors are effectively acquiring the company’s high-growth core business for free, while any future value-unlocking initiatives represent embedded upside optionality.

Sources

[1] https://www.rasa.co.jp/company/history/history.html

[2] https://yunogami.net/ej/_src/1020/EJ11年3月号.pdf

[3] https://www.rs-tec.jp/en/press/press_images/20180410.pdf

[4] https://shachomeikan.jp/presidents/detail/10062694/5195

[5] https://bridge-group.co.jp/ipo_media/6978/

[6] https://toyokeizai.net/articles/-/935426?page=3

[7] https://www.rs-tec.jp/corporate/history/

[8] https://contents.xj-storage.jp/xcontents/AS02916/68cc9041/1a69/4951/8014/b845895bed3a/140120260514534869.pdf

[9] https://www.sumcosi.com/english/ir/library/presentations/

[10] YouTube

[11] https://contents.xj-storage.jp/xcontents/AS02916/4dfb2d10/d8e7/4345/b803/f1e0c7b10577/140120260410502015.pdf

[12]https://www.siltronic.com/fileadmin/investorrelations/2025/Q4/20260312_Siltronic_InvestorPresentation.pdf

[13] https://investor.ti.com/static-files/fc9d9346-cf77-40db-902a-e9961e9c5736

[14] https://www.listcorp.com/asx/cmg/critical-minerals-group-limited/news/investor-presentation-3364878.html

[15] https://contents.xj-storage.jp/xcontents/AS02916/ca064325/4738/4e9c/9256/af5d7cf9ca3d/140120260514534875.pdf

[16] https://contents.xj-storage.jp/xcontents/AS02916/30bd651d/ce27/42a5/a881/1d796c683425/20230214172151528s.pdf

[17] https://contents.xj-storage.jp/xcontents/AS02916/f053a8d2/a8d9/4b7b/92e3/bee3912a8d8b/140120240216538951.pdf

[18] https://prtimes.jp/main/html/rd/p/000000007.000110789.html?utm_source=chatgpt.com

[19] https://www.psi.com.tw/en/share_news2.asp

[20] AlphaSense

[21] Bloomberg

Subscribe now

Disclaimer / Disclosures

The information contained in this report is provided solely for informational and educational purposes and does not constitute investment advice, investment research under applicable securities regulations, an offer to sell, or a solicitation to buy any security or financial instrument.

The opinions expressed in this report reflect the author’s views as of the date of publication and are subject to change without notice. This report has been prepared without regard to the specific investment objectives, financial situation, or particular needs of any individual investor. Readers should conduct their own independent research and seek professional financial, legal, tax, or other advice before making any investment decision.

The analysis presented in this report is based on information believed to be reliable, including publicly available sources, company disclosures, industry publications, and discussions with market participants. While reasonable efforts have been made to ensure the accuracy of the information, no representation or warranty, express or implied, is made as to its accuracy, completeness, or reliability. Any estimates, assumptions, or forecasts represent the author’s judgment and are inherently subject to uncertainty.

This report contains forward-looking statements, including expectations regarding future financial performance, industry trends, market developments, and company prospects. Such statements involve known and unknown risks, uncertainties, and assumptions that may cause actual results to differ materially from those expressed or implied. No assurance can be given that any projections or forecasts will be realized.

The author may hold, establish, increase, reduce, or dispose of positions in any securities discussed in this report at any time without prior notice. Such positions may create potential conflicts of interest. Nothing in this report should be interpreted as a recommendation to buy, sell, or hold any security.

To the fullest extent permitted by applicable law, the author disclaims any liability for any direct or indirect loss arising from the use of, or reliance upon, the information contained in this report.

Against this backdrop, the performance of LVMH Moët Hennessy ($LVMH, $MC.FP) shares appears almost paradoxical: the largest company in the sector would seemingly be among the primary beneficiaries and grow at least in line with the market, yet over the past three years it has consistently lagged behind its peers, while also dragging down the entire thematic benchmark. One of the most common explanations, widely cited by the sell-side, is the company’s active M&A activity, which, among other factors, contributes to an increased “conglomerate discount.”

Analysts, including Morgan Stanley, UBS, and HSBC, described both the latest annual report and the previous one as “mixed but encouraging” in their post-earnings reviews. At JP Morgan, the negative impact of large M&A transactions is highlighted as one of the key risks that could lead to a downgrade of the “Neutral” rating and target price.

To be fair, deals on the scale of Tiffany, valued at $15.8 billion, are relatively rare. Over the past year, the company has focused less on large-scale portfolio expansion and more on strengthening the ecosystem around its core brands (Louis Vuitton, Dior, etc.), occasionally opening a window into media, gastronomy, and lifestyle. These efforts are aimed at increasing customer engagement. Accordingly, the downside risk discussed on Wall Street is not tied to a one-off financial impact from such transactions, but rather to a gradual compression of operating margins if the number of such non-core acquisitions increases — particularly in a scenario where the core business slows.

However, it appears that such deals will continue. During the Q3 2025 earnings call, management emphasized that these initiatives help drive traffic and create an emotional connection between the customer and the brand.

Frankly, it has long been difficult to classify LVMH Moët Hennessy unequivocally and without reservations as “hard luxury” — unlike Hermès ($RMS.FP), which relies on a narrow but highly loyal client base, tightly controls supply (going as far as legal disputes over the right not to sell its exclusive $15,000 handbags to insufficiently loyal customers), and maintains strong pricing power. While LVMH is trying to strengthen “brand connection” through ventures like bakeries, Hermès bags continue to fly off the shelves like hot cakes. According to Sotheby’s, their average resale price increased by around 35% year-on-year, while demand from collectors remains strong: sales of Birkin and Kelly bags rose by 44% in 2025 compared to 2024 and by 55% relative to 2023.

Therefore, beyond M&A activity, LVMH has already accumulated a whole set of challenges: limited room for further price increases, risks of brand dilution, frequent turnover of creative directors, and increasingly visible signs of succession tensions. We discuss these in more detail below.

Louis Vuitton Has Run Out of Price Hikes

Remember the “shut up and take my money” meme? Until recently, it fairly accurately described the behavior of luxury consumers. Now the tone is shifting: an uncomfortable question is increasingly being asked — what exactly am I paying for?

The personal luxury market is facing the consequences of years of price increases, which have led to the loss of “aspirational” demand and a weakening of interest from younger consumers, including Generation Z, in favor of so-called affordable luxury. The number of customers has declined from 400 million in 2022 to around 340 million in 2025, and could shrink by another 20–30 million. Notably, even the wealthiest clients — who account for roughly 46–47% of the market — have started to spend less. The contraction of the customer base and the cooling of key buyers are setting a more subdued trajectory for financial growth.

The Louis Vuitton brand has historically enjoyed strong pricing power: a 5% increase in prices for key products — primarily iconic handbags — would theoretically add around 2% to segment revenue. This implies that handbags account for roughly 40% of Fashion & Leather Goods (F&LG) revenue. And the company has actively leveraged this.

The brand has for many years raised prices on its key models annually — and sometimes more frequently — reinforcing Louis Vuitton’s positioning as an ultra-luxury brand and supporting high residual values in the secondary market. In 2020-2022, LV implemented several notable rounds of price increases, but the pace of indexation has since slowed. Moreover, in recent earnings calls, C-level executives have placed unusually strong emphasis on maintaining price discipline. In effect, this suggests that Louis Vuitton has temporarily (though potentially for an extended period) lost one of its key revenue growth levers.

Here are several comments from Tegus insiders and analysts from client-only reports regarding the current situation:

“For the current luxury, the narrative is still spreading that luxury is a scam or people feel they are more [rational]. The whole luxury disenchantment movement is still ongoing and more and more people are believing that luxury is not something appealing or it’s not something they aspire to. I think, on the demand side, there’s not only income issue but also this narrative spreading on social media.” - Luxury Critic, Senior Editor at LADYMAX

“If you look at LV probably last year, massive increase in price. Bring new creative director. Increase the prices because they know they're going to do less volume just by nature of the business and hence, trying to offset their sales reduction or traffic or volume by the increase in price. Didn't work out.” Former CFO Americas, Fendi at LVMH

In our view, the situation is complex and has two sides:

Negative: loss of part of the customer base due to resale and the risk of losing “desirability.” According to Bain & Company, spending on new luxury goods declined by 7% among Generation Z and by 2% among millennials in 2024. “Appetite for these brands and products remains high, but the willingness to pay current prices is low,” says Claudia D’Arpizio, Global Head of Fashion & Luxury at Bain & Company. Against this backdrop, some consumers are shifting toward the secondary market. Among the most in-demand brands on resale platforms are Louis Vuitton, Chanel, and Gucci.

The secondhand luxury market is estimated at $56 billion — nearly three times larger than a decade ago. Increasingly, consumers check resale prices before purchasing new items. For brands with strong residual value (such as Louis Vuitton or Bottega Veneta), this is a positive: their bags retain on average 89% of their original price upon resale.

Brands are now actively using reseller data to track demand. For example, the price of Chloé Paddington bags on The RealReal increased from $217 in 2024 to $724 in 2025, prompting the brand to quickly reissue the model. Louis Vuitton and Balenciaga have also revived archival designs. For some LVMH brands, this serves as an additional marketing resource, but in the event of declining “desirability,” the situation could change dramatically. This could lead to an accelerated migration of customers toward resale, other brands, and even counterfeits.

“Even for some well-educated people, they’re opening to buy takedown version of luxury or even counterfeit… People don’t have luxury shame. A lot of people, they are very close. They will disclose they wear counterfeit and the genuine products at the same time.” - Luxury Critic, Senior Editor at LADYMAX

Positive: the emergence of more affordable products and, as a result, a larger total addressable market (TAM). LV is actively lowering the entry price point into the brand, which may attract new customers:

“The entry ticket in order to have something originally from Vuitton was very expensive. Now there was a change in strategy in Vuitton. I would say most luxury brands, they started to develop wallets, shoes, tennis shoes. There is some people wearing Vuitton because they buy a $1,000 tennis shoes and it’s an original Vuitton tennis shoes. That’s in my perspective the way they are approaching younger generations or lower segment of income, that’s one way, by the average selling price of the ticket”. - CFO at LVMH

At present, Louis Vuitton is effectively serving as a platform for testing new retail formats and expanding its product matrix — which appears risky, given that the brand is also the primary engine of the entire conglomerate. The company is trying to sit on two chairs at once: to preserve a sense of exclusivity while simultaneously driving sales through more accessible products and a softer pricing policy. This likely implies more moderate annual price increases and the development of entry-level categories, particularly small leather goods.

However, this ambiguity dilutes the brand’s positioning and makes it harder to compete with brands where pricing logic and the perception of scarcity remain crystal clear to customers — such as Hermès and Chanel.

Creative Directors vs. Creative Direction

In fashion, the creative director is effectively the architect of product strategy, the cadence of novelty, and the codes a brand will commercialize. This season has been one of the richest in fashion history in terms of leadership reshuffles. Following the debut of Matthieu Blazy as Creative Director of Chanel at Paris Fashion Week, perhaps the second most significant event was the appointment of Jonathan Anderson at Dior (the second-largest brand within the LVMH Moët Hennessy portfolio by revenue after Louis Vuitton).

In a positive scenario, analysts compare the current situation to the late 1990s, when John Galliano, Alexander McQueen, Nicolas Ghesquière, and Martin Margiela fundamentally reshaped fashion and pop culture, successfully balancing creative innovation with commercial success. Chanel and Dior appear to be potential beneficiaries of this industry “reset,” while some players risk losing market share, according to Bloomberg.

As noted by The Business of Fashion, a strong start is crucial for brands today: early signs of success can help regain consumer attention and trust, but the effectiveness of these changes can only be assessed after several seasons. Dior’s transition under Anderson is a calculated risk. The scale of the brand is enormous, but his experience at Loewe increases the likelihood of a successful transformation, even if it takes time. Most of his products will only be available in stores toward the end of this year, and the full impact of his appointment will likely become clear only by mid-2026.

To better understand the situation, we decided to draw parallels between Dior’s two most recent creative directors:

Maria Grazia Chiuri (Womenswear): 2016-2024

Under Chiuri, Dior’s revenue increased more than fourfold. Her “perpetually safe” Dior was not considered particularly exciting by the niche fashion community and was often criticized for a lack of edge and creative boldness. However, it was crystal clear to the core base of affluent female clients — and therefore commercially highly successful.

So why was there a need to part ways with Chiuri? Prices for luxury goods continue to rise, which means that either product quality (where Dior remains strong as a largely craftsmanship-driven brand) or creative direction must evolve — otherwise consumers will simply turn away. Under her tenure, there was a growing sense of narrative fatigue: audiences were craving something new and exciting. As a result, conglomerates across the industry are increasingly embracing change.

“There has also been a significant amount of bad press surrounding the cost of labor and materials to create some coveted luxury goods vs. the retail markup… turning customers off”. - Former Manager, Media & Digital (Van Cleef & Arpels)

“There’s a shift towards commercial luxury items and concerns about lack of newness… I think it was because the artistic director was the same. It really started because of the product strategy. They just introduced the right product at the right time [before], but now that’s missing”. - Former Head of Sales - Europe & Middle East at LVMH

“Dior must reinvent the experience and find a way to surprise again and again their customers…” - Former Visual Merchandising Director — EMEA at Christian Dior Couture (29 Apr 2025)

Jonathan Anderson: 2025-Present

The first designer since Christian Dior himself to simultaneously oversee all divisions of the house — menswear, womenswear, couture, and accessories. At least, this is how Anderson is presented by the Arnault family; formally, a similar level of authority was once held by Marc Bohan. In any case, the responsibility is enormous. The pressure is immense — especially considering how much LVMH Moët Hennessy has monetized Dior’s womenswear division under his predecessor.

Anderson must attract a new audience without alienating the existing one:

“You can’t assume that even if you lose one existing client, you will simply gain two new ones thanks to unique creative solutions. The days when you could join a brand and start from a blank slate are over,” - Robert Williams, an editor at The Business of Fashion specializing in luxury brands

Anderson is a driven art enthusiast, long favored by a similarly niche or aspirational audience. He has successfully managed both his eponymous brand, JW Anderson, and Loewe. Under his leadership, Loewe’s revenue grew over 11 years from approximately €230 million to €1.5–2 billion. However, Jonathan is now compelled — and intends — to appeal to a much broader audience. Will it work? We will see. That said, sentiment among observers and insiders regarding Anderson’s era at Dior is generally bullish:

“I think that Dior will probably gain market shares. LOEWE has been, so far, a very desired brand… They were leading the luxury arena when JW Anderson was at the helm of the creative direction of LOEWE. I think this is going to be the case for Dior. The industry, and probably, all the clients of the luxury industry are waiting to see its first collections hitting the stores.” - Global Head of Sales Merchandising - Ready-to-Wear at Chloé

“He does have a great track record. I think what he did at Loewe was incredible. It was an unknown name to a very strong player in the industry… it’s really interesting and the right move that they have him overseeing both men’s and women’s, because I think from a brand direction perspective, it’s always been so disconnected… To bring it all under one designer again for the first time since Christian Dior himself, that is a big, bold move.” - Former Senior Director, West at Dior

“He has a very strong point of view on what the luxury market should be… He takes a point of view from a different generation… He has a very sharp point of view on what are the codes that a luxury maison should be using today to be relevant with that Gen Z cohort of people.” - Former Brand President, North America at Bulgari

There are already some early positive signals. Jonathan Anderson’s first womenswear collection at Dior became the second most viewed Spring/Summer 2026 collection on Vogue Runway, уступив лишь дебютной коллекции Matthieu Blazy at Chanel.

Anderson did exactly what the market was expecting: he engaged with the house’s core codes. While the Bar jacket was not radically reinterpreted, the key accessory — the Lady Dior — underwent a meaningful “reboot.” The bag remained rooted in Dior’s DNA but evolved in form: softer versions emerged, alongside new materials and unconventional proportions. This was critically important.

Across Tegus materials and public sources, there has been a consistent narrative around the lack of new strong product hits, particularly in leather goods. Insiders pointed to the exhaustion of flagship lines, a limited set of products capable of consistently driving turnover, and growing consumer fatigue with logomania:

“There has really not been any innovation in product in the last six or seven years… there is a finite amount of Lady Diors and Book Totes and Saddle Bags that are going to drive your business, especially as you get to this $1+ billion mark.” - Former Senior Director, West at Christian Dior Couture (30 Jul 2025)

At Louis Vuitton, creative leadership is split between two figures: Nicolas Ghesquière (womenswear) and Pharrell Williams (menswear). This has created a structural setup the brand has not encountered in this form before, and it carries a number of implications:

• A split in the brand’s visual and stylistic codes. The menswear line, following the arrival of Pharrell Williams, has taken on a bold, pop-cultural direction — collaborations, media visibility, streetwear influence, and strong resonance on social media. He has revitalized men’s Louis Vuitton, making it more visible and desirable. Meanwhile, the womenswear collections under Nicolas Ghesquière are increasingly perceived as somewhat dated and predictable. His approach leans toward a refined but expected aesthetic, which experts often describe as “redundant” and “lacking innovation.”

• A divergence in commercialization. The menswear segment — particularly ready-to-wear — benefits from the buzz surrounding Pharrell, whereas the womenswear division, which remains the key revenue driver, lacks a comparable creative “spark.” This creates an imbalance: the brand is perceived as highly “hot” in menswear from a media standpoint, yet financially it remains dependent on the more conservative womenswear segment.

• Complexity in product strategy and communication. Two creative teams with different approaches are producing collections, campaigns, and collaborations with distinct aesthetics. This makes it more difficult to build a cohesive narrative in global marketing campaigns — especially in Asia, where consumers are particularly sensitive to a unified “brand image.”

• Risk of audience fragmentation. The menswear line is more successful in attracting a younger and broader audience, while womenswear remains focused on traditional HNW clients. Without alignment in positioning, the brand risks being perceived as two parallel worlds rather than a single, unified Louis Vuitton house.

Here are several insider comments on the current situation:

“If I think about Vuitton internally, they have Pietro Beccari — he's fantastic. From a leadership perspective, Pietro has been able to move them forward and resolve internal challenges. Obviously, the appointment of Pharrell Williams was really interesting and made a lot of waves in the menswear space, which you have to do when you're a luxury player in that segment.” - Former Senior Director, West at Christian Dior Couture

“Even if we think of Louis Vuitton, they still have two different creative directors. They have Nicolas Ghesquière for the women's universe and Pharrell for men's, but they still rely on two separate creative directors, which is quite a challenge because sometimes we see a gap between the women's and men's categories. From my perspective, Pharrell has made the men’s universe very, very attractive. Nicolas Ghesquière has been at the helm of womenswear for such a long time — it feels very redundant and lacking innovation. This is probably what disappoints many of the brand’s clients. Again, in tough markets like Asia, a lack of innovation is a major challenge.” - Global Head of Sales Merchandising – Ready-to-Wear at Chloé

At present, there is no publicly available or insider information regarding potential changes to the creative director roles at Louis Vuitton.

Catching a “Wolf in Cashmere” by the Tail

The internal dynamics within Bernard Arnault's family are increasingly being compared to the plot of the TV series Succession: all five of his children are involved in managing LVMH Moët Hennessy and are competing for influence within the group. A de facto “Darwinian” model has formed within the company, with competition for positions and a gradual selection of a potential successor. While the market may tolerate uncertainty around succession as long as the business is growing rapidly, in a phase of “normalization” or stagnation, any managerial uncertainty becomes significantly more costly.

Children from the first marriage (with Anne Dewavrin):

• Delphine Arnault — the eldest daughter. Since February 1, 2023, Delphine Arnault has been Chairman and Chief Executive Officer of Christian Dior Couture. She is a member of the LVMH Board of Directors and the Executive Committee.

• Antoine Arnault — Director of Image & Environment, also a member of the Executive Committee; Chief Executive Officer and Vice-Chairman of the Board of Directors of Christian Dior SE.

Children from the second marriage (with Hélène Mercier):

• Alexandre Arnault — one of the younger sons, Deputy Chief Executive Officer of Moët Hennessy, and a member of the LVMH Board of Directors. Previously Executive Vice President at Tiffany & Co.

• Frédéric Arnault — Chief Executive Officer of Loro Piana; previously CEO of TAG Heuer and head of LVMH’s watchmaking division.

• Jean Arnault — Director of Marketing and Development for Louis Vuitton Watches, overseeing the brand’s watchmaking activities.

Here’s what matters:

• Shareholders approved a change to the company’s bylaws allowing Bernard Arnault to remain Chairman and CEO until the age of 85. The decision was supported by 99.18% of votes. The age limit had previously already been raised — from 75 to 80 in 2022. Arnault, now 77, continues to combine the roles of CEO, Chairman, and controlling shareholder of LVMH Moët Hennessy.

• The family has strengthened its control. In February 2026, regulatory disclosures showed that the Arnault holding structure crossed the 50% ownership threshold (50.01%). This is symbolically important: an absolute majority stake reinforces the family’s “defensive shield” against external pressure.

The extension of Bernard Arnault’s tenure to age 85 reduces near-term uncertainty. However, the succession question remains open and acute: this is more a postponement of the decision than a resolution. Shareholders of LVMH Moët Hennessy are increasing pressure on the company, demanding greater transparency on this issue. A concrete succession plan has still not been disclosed, and any changes at the senior management level continue to fuel speculation about a potential distribution of roles within the family.

The governance structure reinforces this dynamic. All heirs are regularly involved in strategic discussions — Bernard Arnault holds monthly meetings with them — and the market interprets key appointments as signals of the balance of power within the family. At the same time, control is evenly distributed: each of the children holds an equal stake in the holding structure that is expected to assume control after Arnault steps down.

The family’s key holding structures include Financière Agache and Agache. Following the 2022 reorganization, all five heirs — Delphine Arnault, Antoine Arnault, Alexandre Arnault, Frédéric Arnault, and Jean Arnault — each hold a 20% stake in Agache Commandite SAS, which is set to control the family holding in the future. Corporate governance experts warn that such a structure increases the risk of disagreements in decision-making and may complicate the selection of a single leader.

The configuration of “centers of power” within the family/group can be clearly inferred from official appointments.

At the same time, even large institutional shareholders acknowledge that they have no clear understanding of how a potential transition of power would be structured. Moreover, the issue is hardly discussed within the family itself. Hélène Mercier, the wife of Bernard Arnault, stated in an interview with Libération that the topic is “almost never” raised and that the decision rests entirely with Arnault. Potential frictions among the heirs and the absence of a clear succession mechanism increase the likelihood of long-term governance risks and may be reflected in the company’s valuation through a discount to multiples.

So where does this leave us? Bernard Arnault is already 77 and has no intention of stepping down: he has extended his ability to remain CEO until 85 and has explicitly stated that the succession issue can be revisited “in 10 years.” This sets the baseline framework — formally, there is ample time, but the market perceives the situation differently. Investors are already citing the lack of a clear plan as a source of risk and opacity, which is beginning to weigh on the company’s valuation.

At the same time, the governance structure is designed so that all five children remain actively involved. They hold operational roles in key group assets — Dior, Moët Hennessy, Tiffany, and Loro Piana — and are not merely nominal heirs but active managers. Simultaneously, they each own equal stakes (20%) in the family structure that will ultimately control the holding, implying a model of collective decision-making. Against this backdrop, Arnault has been steadily increasing their presence at the center of governance. In 2026, Antoine Arnault joined the executive committee — the inner circle where key decisions are made — becoming the second family representative in that body after Delphine Arnault.

In our view, he is the most likely successor to the “wolf in cashmere.”

• The remittance market consists of three main types of participants: banks, traditional money transfer operators (MTOs), and digital fintech platforms. Banks account for approximately 40% of the market, while traditional MTOs and fintech services each hold about 30%. Fintechs are rapidly gaining share due to their ability to offer lower transaction costs compared to banks and MTOs.

• Fintechs achieve greater cost efficiency through: (a) digital channels, which eliminate intermediaries at the customer interaction stage; and (b) liquidity management (local pools, netting), which minimizes the need for correspondent banks during the actual movement of funds.

• Wise (WISE.L; WPLCF) offers the lowest-cost solution in the industry: the average take rate for personal transfers is 0.57%, and for business transfers – 0.39%. The blended take rate stands at 0.52%, down from 0.64% a year earlier. For comparison, Western Union’s rate remains around ~3.5%, and Remitly’s at ~2.3%.

• The low take rate is a result of Wise’s advanced infrastructure, which took years to build. Replicating it is difficult, requiring dozens of integrations, licenses, and regulatory approvals across jurisdictions. This creates a strong competitive moat, which Wise continues to deepen by deliberately lowering its take rate as it scales.

• The low take rate drives higher transfer volumes and growing customer balances held in Wise’s multi-currency accounts – a key source of interest income. This creates a self-reinforcing cycle: lower fees drive volume growth, which in turn enables further price cuts, making Wise increasingly competitive.

• Wise has multiple growth levers. Its core business remains the B2C segment with the Wise Account, which drives growth in customer balances, card revenues, and interest income. The B2B segment (Wise Business) also holds significant potential, as the corporate remittance market is much larger and Wise’s current penetration remains low.

• Wise Platform enables fintechs and banks to embed transfers into their products. While it currently accounts for less than 5% of volume, it is already used by players like Morgan Stanley, Interactive Brokers, and Raiffeisen Bank, and could become a major volume growth driver. Management expects the Platform to reach 10% of total volumes in the medium term and over 50% in the long term.

• While Wise trades at a premium to most remittance peers, we believe current valuation levels are reasonable for initiating a position: the company is trading at a 2025 FCF yield (to EV) of ~4.61% and a 2026 yield of ~5.84%, with expected Free Cash Flow growth of 18.2% in 2026 and a projected FCF CAGR of ~17.2% through 2030.

• We include Wise in our cohort of high-conviction ideas with a target allocation of 10% of NAV. We are initiating one-third of the position at current levels.

• Our financial model for the company is available here.

Subscribe now

The Company

The story of Wise began back in 2007, when Kristo Käärmann, a Big Four consultant living in London, and Taavet Hinrikus, then Director of Strategy at Skype, realized they shared the same problem: high currency exchange costs when sending money to and from their native Estonia. Hinrikus received his salary in Estonian kroons and converted them into British pounds each month; Käärmann did the opposite, converting his GBP salary into kroons.

The future co-founders came up with a simple solution: each month, Käärmann would top up Hinrikus’s UK account, and Hinrikus would do the same for Käärmann’s Estonian account, setting the amounts based on the mid-market exchange rate.

The idea quickly spread among other Estonian expats and evolved into a Skype chat group. In 2011, the informal money exchange forum transformed into TransferWise [1]. That same year, the company raised a $1.3 million seed round with participation from IA Ventures, Index Ventures, Max Levchin, and Peter Thiel. Since then, Wise has completed seven more funding rounds, growing from a startup into a public company with a valuation of £8.75 billion at the time of its IPO in July 2021 [2].

Today, Wise is one of the largest players in the industry, with a market share of less than 5% in P2P remittances and under 1% in small business transfers. Over the past four years, the company’s active customer base has grown 2.8x, cross-border payment volume by 2.9x, and revenue by 3.1x [3].

Wise remains founder-led. Kristo Käärmann is the CEO and owns 18.23% of Class A shares and 46.6% of Class B shares, giving him control of approximately 49.3% of total voting rights. Taavet Hinrikus left the company in 2021 but retained 5.4% of Class A shares and 8.4% of Class B shares, representing about 12.2% of voting power [4]. Class B shares do not grant rights to the company’s earnings but carry nine votes per share [5].

The Money Transfer Industry

The global money transfer market consists of two main segments:

• Personal transfers (P2P) — money sent between individuals. This segment is commonly referred to as remittances. While it accounts for a relatively small share of total payment volume, it generates a disproportionately high share of industry revenue due to steep fees. According to McKinsey, low-value cross-border transactions — including P2P remittances, small P2B, B2C, and SME B2B payments — represent only about 10% of global payment volume, which is estimated at $162 trillion in 2024 [6]. Yet, these small transactions are thought to generate up to one-third of total market revenue, thanks to a high average take rate of around 6.6% [7]. Wise estimates the annual remittance volume at $2 trillion [3].

• Corporate transfers — cross-border payments between companies, financial institutions, trade settlements, and more. This segment dominates by volume, accounting for 90%+ of global flows by value. B2B payments by large corporations and banks alone amounted to tens of trillions of dollars in 2024. The total volume of all cross-border business payments is estimated at $146–150 trillion annually [6]. These payments are typically processed through banks and the SWIFT correspondent network, with lower fees thanks to large transaction sizes. However, even in the corporate space, a clear divide is emerging: while large transactions remain with traditional banks, small and medium-sized enterprises (SMEs) are increasingly turning to fintech providers for smaller international payments. In fact, 35% to 50% of SMEs in North America, Europe, and emerging Asia have used non-bank fintech services for international payments at least once a year [6]. Wise estimates that small businesses transfer around $12 trillion annually, with this segment growing at a 9% CAGR over the past two years.

The US dollar and the euro are the dominant currencies in the industry, facilitating the majority of global payment flows. According to SWIFT data, USD and EUR together accounted for over 70% of all international payments in 2024 [8].

The money transfer market is made up of three main types of players: banks, traditional money transfer operators (MTOs), and fintech platforms. MTOs and fintechs combined handle approximately 60% of global transfer volumes, while banks account for the remaining 40% [9].

Within the non-bank segment, fintech is rapidly gaining ground. In the 2010s, the majority of non-bank transfers were processed via offline networks like Western Union and other MTOs. By the mid-2020s, however, roughly 50% of all transfers were already being made through digital channels [10]. This trend is expected to continue, with the digital remittance market projected to grow at 16–17% annually through 2030, significantly outpacing the overall market growth rate of ~5–6% [11][12].

The disruptive nature of fintechs lies in their ability to offer significantly lower transaction costs compared to both banks and MTOs. For end users, the key parameters when sending money are the transfer fee and the exchange rate margin. According to the World Bank, the current average cost of sending money from G20 countries stands at 6.77% [11].

Below is a comparison of average fees for sending $200 based on provider type [7].

Banks are the most expensive option. Their infrastructure relies on the SWIFT interbank messaging system and a network of correspondent accounts. In a cross-border bank transfer, funds move through a chain of institutions: the sending bank sends a SWIFT message to its correspondent (or directly to the recipient bank if a direct relationship exists), and the money is routed through Nostro/Vostro accounts — accounts that banks hold with one another to settle international transactions — before finally being credited to the recipient’s account. This process is relatively complex and can take several days, especially when multiple intermediaries are involved.

It is precisely the fact that funds pass through a chain of intermediaries — each aiming to profit from the transaction — that makes the overall cost to the customer so high. Notably, since bank transfer fees are typically fixed, banks tend to earn more from the FX spread. This makes their pricing model less transparent.

The largest players in the bank-led segment of the money transfer market are primarily those with extensive international networks of subsidiaries. These subsidiaries allow banks to use their own internal FX rates and settlement systems, bypassing SWIFT and third-party correspondent banks. They also enable banks to act as correspondents for others.

The top banks by transfer volume include JPMorgan Chase, Bank of America, Wells Fargo, HSBC, and Citi. JPMorgan is the only institution with transfer volumes exceeding $25 billion. For comparison, Wise — the largest player in the non-bank segment — transferred $145.2 billion in 2025, up from $118.5 billion in 2024 [13].

Traditional Money Transfer Operators (MTOs). This category includes companies such as Western Union , MoneyGram, Ria (Euronet Worldwide, ), and numerous regional providers. Their infrastructure differs significantly from that of banks: MTOs have built their own global networks of agents and branches, enabling them to send and disburse cash nearly anywhere in the world within minutes.

At the core of an MTO system is a proprietary payment network. For example, Western Union operates over 500,000 service locations worldwide — ranging from urban branches to small agent outlets in rural areas. When a sender brings funds (in cash or by card) to a Western Union location, the agent enters the transaction details into the company’s centralized system. Within minutes, the information becomes available to the receiving agent, and the recipient can collect the funds almost immediately by presenting identification and the tracking number.

The actual movement of money occurs through a system of inter-agent settlements with the central company — typically using netting (offsetting incoming and outgoing flows) and periodic reconciliations between Western Union and its agents in each country. Other MTOs operate similarly.

Many traditional MTOs have expanded their infrastructure to include digital channels. Today, both Western Union and MoneyGram allow customers to send money via website or mobile app, with payment from a bank account or card, and receive funds not only in cash, but also via bank deposit, mobile wallet, or card payout. This brings their infrastructure and cost structure closer to those of fintech players.

However, the core of the MTO model remains a hybrid of a digital platform and an agent-based network. For instance, only 28% of Western Union’s revenue comes from digital channels, while 45% of its total expenses are still paid out as agent commissions [14].

Fintechs. The leading players in this category include Wise, Remitly , WorldRemit, as well as the fully digital segments of traditional MTOs. These providers were built from the ground up as digital and distributed platforms, with no need for physical branch networks. This gives them a structural cost advantage and allows them to offer lower fees to end users.

The core principle behind fintech infrastructure is managing local liquidity pools. For example, Wise maintains bank accounts or local partnerships in each country it sends money to or from. When a customer in the U.S. sends $100 to a recipient in India (in rupees) via Wise, the funds may not actually move cross-border. Instead, Wise credits $100 to its local U.S. account, and an equivalent amount in INR is paid out from Wise’s account in India to the recipient — effectively two local transfers.

The currency conversion is done at the mid-market rate, and Wise charges a fee on top. This creates a P2P-style liquidity pool, where funds flow through the platform without using SWIFT or correspondent banks — significantly reducing both cost and time. Other fintechs operate with similar models.

Here’s how a former Wise product manager explained the movement within liquidity pools in an interview on Tegus:

“You've got many teams in Wise who manage those liquidity pools, which means that they are forecasting and estimating how many yuan or Singaporean dollars or whatever will be needed for tomorrow. Actually, they're forecasting three, four days out. What they then do is go, "What's the likelihood of us getting pay-ins to have that amount? What is the gap? From which market should we transfer to top that up so that amount is correct for the payout estimate?" and then they make those transactions on the normal SWIFT network.

If you're looking at that, and let's just say they pay 4% for those conversions and they send maybe U.S. dollars to Singaporean dollars and they have to pay 4%, it's only about 10% is the gap. Effectively, they can turn that 4% into 40 bps because all they're doing is they're just topping up the difference. Even though they might be paying 4% to do that, because they're distributing it across the other 90% of liquidity that they had, the actual underlying cost drops down the 40 bps.”

In other words, effective liquidity pool management requires deep integration with banking infrastructure and local payment systems. Fintechs achieve greater cost efficiency in money transfers through two key mechanisms:

• Digital distribution, which eliminates intermediaries at the customer interaction stage — unlike traditional MTOs, where agents act as a layer between the company and the customer, adding cost.

• Liquidity management via local pools and netting, which minimizes the need for intermediaries (i.e., correspondent banks) in the actual movement of funds.

Their Margins is Our Opportunity

This quote is often attributed to Jeff Bezos, the founder of Amazon. And it perfectly captures the essence of the strategy that allowed the company to turn e-commerce — a business with relatively low barriers to entry — into a capital-intensive infrastructure “fortress.” For years, Amazon sacrificed margins in order to gain market share and strengthen its competitive moat by building a level of infrastructure that no one else could easily replicate.

In early April, during Wise’s Investor Day conference call, CEO Kristo Käärmann echoed that famous Bezos quote. Wise’s strategy in cross-border payments indeed mirrors Amazon’s approach — especially in its:
a) aggressive fee compression in pursuit of scale,
b) vertical integration, and
c) reinvestment into proprietary infrastructure.

Today, Wise has the lowest take rate in the industry (i.e., the fee it charges for international transfers). In FY2025, which ended on March 31, the average take rate for personal transfers was 0.63% (with average volume per active customer at £7.2k), while for business transfers it was 0.44% (VPC at £55.8k). The blended take rate came in at 0.58%, down from 0.67% a year earlier — a record low for the company, achieved without sacrificing operational efficiency [15].

On its website, Wise maintains a public comparison tool showing rates charged by different providers across currency corridors [16]. At first glance, the market appears crowded with competing offers at comparable price points. In the “USD/EUR” corridor, for instance, Wise trails behind Xoom (PayPal ), Instarem, MoneyGram, and Remitly.

However, competitors often subsidize their rates in the most competitive corridors by offering less favorable pricing in other markets. Wise deliberately avoids cross-subsidization — expensive or complex routes are not financed by the more profitable ones. Each corridor must be profitable on its own.

Management believes that price reductions should follow cost reductions, not precede them.

“Our pricing is a function of our cost to serve a corridor, plus a small margin. That is why we can keep reducing prices sustainably.”— Kristo Käärmann, CEO

Wise’s take rate is significantly lower than that of publicly traded competitors.
Western Union, the second-largest player in the money transfer market by volume, maintains a take rate of approximately 3.5% [16]. Remitly — Wise’s closest and effectively only major publicly listed fintech rival — operates with a take rate of around 2.3%. Given that private fintech competitors operate at a much smaller scale, their take rates are likely even higher.

Low fees are one of Wise’s key competitive advantages. Thanks to its pricing edge, the company has been steadily gaining share in the consumer remittance market.

The massive gap in take rates is primarily a result of infrastructure — infrastructure that took years to build. Wise holds over 70 Electronic Money Institution (EMI) and Payment Institution (PI) licenses, enabling it to provide money transfer services, hold customer funds, and issue cards across various jurisdictions. The company is integrated with more than 90 local banks and 8 domestic payment systems: FPS in the UK, SEPA in the Eurozone, GIRO/AFR in Hungary, Pix in Brazil, Zengin in Japan, FAST in Singapore, NPP in Australia, and InstaPay in the Philippines. Wise is currently working on a direct connection in the United States.

Direct integration with national payment systems removes unnecessary links in the transfer chain — whether correspondent banks or payment aggregators — allowing Wise to operate in these countries as a fully-fledged settlement participant. This dramatically lowers costs and speeds up transfers. For example, following its integration with the Philippines’ InstaPay system, up to 90% of Wise’s local transfers are now processed instantly, and the base cost per transaction dropped eightfold [17].

Why don’t other players replicate Wise’s infrastructure? Traditional banks avoid doing so for two main reasons:

1. Money transfers account for a negligible portion of their business — roughly 0.5% of total bank revenue and around 1–1.5% of consumer banking revenue [18];

2. Banks rely on a mutual “barter system” of services, where informal reciprocity replaces direct integration.

Here’s how a former Head of International Payments at Barclays explained it in an expert interview on Tegus:

“Sure. Because the wallet share of correspondent relationships and reciprocity is much bigger in value for a bank like Barclays than trying to bypass them and build something new. For example, we were making circa GBP 150 million of revenue from financial institution clients — easy money. If I used Deutsche as my correspondent to send payments into Europe, I paid a lifting fee, and in return Barclays got reciprocity — Deutsche might, for instance, clear payments in the UK for us. <...> The sheer size of this reciprocity business made it more attractive to preserve the status quo.

Now imagine I bypassed Deutsche and used local balances across markets — that would remove their business, and in response they'd cut off reciprocity to us. The value of that reciprocity was still higher. <...> With the rise of Swift GPI and more visibility, this model is under pressure, but behavior hasn’t changed dramatically. Even if Deutsche charges GBP five per transaction, Barclays still benefits from the incoming Deutsche flows into the UK, and those volumes remain significant.”

MTOs cannot replicate Wise’s infrastructure due to a complex mix of regulatory, technological, and economic barriers. First, a large portion of their operations still relies on cash disbursement via agent networks, which requires maintaining physical infrastructure — a costly endeavor that accounts for roughly half of total expenses at many MTOs.

Second, the core customer base for MTOs consists of migrant workers and blue-collar earners with low volume per customer (VPC) — less than $500 annually at Western Union and MoneyGram. These transaction sizes typically fall under simplified regulatory thresholds, allowing MTOs to avoid building out their own KYC/AML infrastructure. By contrast, direct integration with domestic payment systems requires a full audit of all KYC/AML procedures before licenses can be granted.

Third, MTOs often operate at the low end of the pricing spectrum in certain corridors (e.g., India, the Philippines), where they heavily subsidize transfer fees. For them, investing $20–30 million into direct connections with national payment systems is not economically viable when the average revenue per customer is just $1–2 per year.

Notably, the One Big Beautiful Bill Act proposes a 3.5% tax on money transfers that do not go through financial institutions and are not funded via card payments [19]. The bill is intended to reduce the share of "gray" transfers that fall under simplified KYC requirements. This should incentivize MTOs to increase the share of fully digital transfers and invest in internal compliance systems. However, building a KYC/AML stack on par with Wise could take years, even with a dedicated team and sufficient capital. KYC/AML is not just about technology — it involves reporting processes, independent audits, and coordination with Financial Intelligence Units (FIUs) and tax authorities.

Here’s how the company’s management explained the difficulty of replicating its infrastructure during the latest Investor Day:

“So the first part of the question about whether the progress that we make in these direct connections, is this making it easier for newcomers? The answer is yes and no. Yes, in the sense that all our efforts with policymakers, with the regulators at the G20 level as well is making progress towards removing this barrier entry to their payment systems that traditionally was reserved only to banks. <...> And of course, when the regulation changes, when the law changes, it changes for everyone. <...> And as I was explaining, it is not for everyone. It is not for every nonbank financially regulated institution. That is the first step, getting this license.

And once this license comes in, it goes -- we go through a process with a payment network that requires really high standards from a technology, from an operational perspective, requirements around the type of servers or cloud based around our resilience, our business continuity plans. <...> Also the difference is we already have a very strong head start. We have done this in 6 countries. We are about to go live into more. <...> I think it's a very good example that this year, we are delivering to one in Brazil and the other one all the way in Japan.”

Also notable are the comments from Wise’s former Head of Pricing and former Senior Product Manager on how difficult it is to replicate the company’s infrastructure and cost structure:

Tegus Client

We talked about why no one has been able to copy Wise's cost structure. Right now, no one has been able to copy Wise's cost structure. But if you look at three or five years, what is the likelihood that Wise still has such a unique advantage in its cost structure relative to peers?

Former Head of Pricing at Wise

I think if there would be a company that would go full speed on, and they would really focus only on the cross currency, not on all the other features that all the competitors are trying to launch at the same time, they could get to a place where they would start to get a little bit closer to the cost structure. But Wise has too big of advantage.

They're like different cost structures and different ones require very different skills and efforts to optimize. But even the most common example is the bank fees. So all of these challengers who would want to essentially process conversion, they need to have partnerships with banks. And having partnership with banks, the commercial model is much higher than the commercials you get if you directly integrate within a payment system.

To give you more like a tangible sense of what this means, it's more than 10x different, what you pay if you're directly integrated versus if you pay to a partner. So these direct integrations, they take a long time and licenses, and Wise has couple of them already. So just to level the playing field on this one cost item would already buy Wise many years.

Tegus Client

That's interesting. On the point about the central bank integrations, how difficult is that? A big part of Wise's infrastructure is using domestic payment systems, presumably that is something that competitors can still do, so just trying to understand.

Former Senior Product Manager at Wise plc/ADR

Every time Wise does a direct bank integration, they're opening the doors for Currencycloud to follow, because the Wise approach goes through regulatory rehashing of what the rules are. Why can a fintech like Wise be allowed in there, even though they're not a bank? How can they work within those systems? Wise goes through lots of lobbying, lots of conversation, takes years in most markets. The U.S. one is still ongoing and it's been ongoing for a long time. Those things take time.

Once it's done, they're opening the doors for any competitors to catch up, but no one's taking them up on that yet. I think once they realize, then they've got all these markets that was already existent and then they can try and maybe lead in some other market. If they do, then it also opens the doors for Wise. It's not cheap to do. It's not something that I've seen anyone adopt as a strategy apart from Wise.

Tegus Client

Again, supposing that they did try to replicate that, is it still a case that Wise just has the scale of volumes, that the price still wouldn't be competitive? I don't know where volumes would compare versus, say, Currencycloud?

Former Senior Product Manager at Wise plc/ADR

Let's go back a step. The biggest money maker is having £1,000 and €1,000 in the bank at the same time and being able to pay out whoever needs to pay in or pay out the exact amount without having to convert anything, because then you're just paying for the servicing cost, which is your AMLs, your RegTech stuff, your compliance things, your customer support, your chargeback costs, your infrastructure, your AWS bill, that's all you're paying for.

Wise is able to do that in many markets. They don't actually have to exchange. The moment they do have to exchange, that's when they start paying some fee to someone who's helping them with that, or if they've got those liquidity buckets, then they're paying themselves a reasonable fee.

For Currencycloud or anyone to get up to that point, they have to build those liquidity pools. They have to have those inflows and outflows balanced as close as possible. They have to have really robust hedging solutions for the forex fluctuations. If they do want to offer a rate lock like Wise does, if they don't, then they can save money there like Revolut does. There's lots of things where they can move. To really get down to that lowest cost, you need to be optimizing all those areas, the customer support, the transactions, the handling, the liquidity, all of those things.

The central bank integrations help with that. There's a big factor there, because you get to drop the banking partner. That means that you are predictably a bank in that region, which means you've got access to that cash and you don't have to pay any extra markup for those. That is a huge cost. That's half the cost in most other markets.

Most of Wise’s competitors still rely heavily on partner banks and third-party aggregators. Direct, full-stack access to payment systems is rare, and Wise is several years ahead in this regard. However, a number of fintech providers — including Airwallex, Nium, and Currencycloud (Visa, ) — are working to build similar infrastructure.

The closest comparable is Nium, as it operates in both the consumer and business segments, like Wise. However, Nium is significantly smaller: in its most recent funding round, the company was valued at $1.4 billion — 30% below its 2022 valuation — and processes around $25 billion in annual transactions, compared to $199.2 billion (£145.3 billion) for Wise.

Wise also competes with Airwallex in the business segment, where the latter processes about $130 billion, versus $53.4 billion (£38.9 billion) for Wise [20]. The companies have comparable take rates — 0.39% for Wise vs. 0.55% for Airwallex — though it’s worth noting that Airwallex’s volume includes acquiring, which may skew the figures. Also worth mentioning: only 22% of Wise’s revenue comes from the business segment.

An executive officer at a fintech company offered a concise comparison of the two firms. In his view, Wise and Airwallex target different client profiles. Wise is stronger in serving banks and fintechs, offering deeply integrated solutions and access to local payment rails, while Airwallex more often serves corporate clients in e-commerce, travel, and B2B services:

Tegus Client

Given that you mentioned Wise and Airwallex sometimes can be head-to-head competition, is there a rough win rate for you when you're encountered with Wise? What is the typical scenario that you will win over Wise or vice versa?

Director - Strategy & Operations | Executive Office at a Global FinTech Company

Yes. I don't have that. Maybe commercial team track that I don't. Wise is very stronger when it comes to offering their product on the infrastructure side to banks and other fintech’s. Even if you go on their website and click on the platform part, not the personal business but the platform part you will see their focus is that "Okay, the power banks, the power fintech’s and the power enterprises."

Versus Airwallex is stronger in powering the enterprises versus banks and fintech’s. I think it's just by the way of design that how they have built their infrastructure business. If you've seen recent press releases, I think now they're powering some currencies for Standard Chartered. I'm talking about Wise and now they're also powering payments for Morgan Stanley.

If you can see, with just these two press releases and also from the past, most of their big customers are in the banking and the fintech space, financial services space versus Airwallex's customers are very diverse but mostly from the e-commerce space, travel segment, wealth management, professional services.

They are powering. You can say they are more popular among the enterprises who want to become fintech’s by leveraging our infrastructure but don't have their own financial services business yet versus Wise is stronger on the other side. That's one difference.

The other thing is as I mentioned, Airwallex also offers acquiring. If an infrastructure wants card acquiring in their product mix, card acquiring or local payment methods acquiring, then they tend to work with Airwallex versus Wise.

That's one of the other advantages that they have. Wise is stronger when it's competitive fintech’s and banks and Wise may have I would say more direct connections for the banks for the local payment rails versus Wise and I think it's just because they started way before us. Yes, that's the focus. For major currencies now Airwallex also competes with Wise for that.

Currencycloud is a B2B platform for international payments, serving banks, fintechs, brokers, and enterprises. It operates via APIs and white-label solutions, enabling clients to quickly launch financial products while outsourcing much of the backend — including compliance and KYC — to Currencycloud. The company is particularly notable for being a Visa subsidiary and for powering the payment infrastructure behind European fintech giant Revolut [21]. Wise competes with Currencycloud in its platform offering (a subset of its business segment), which currently accounts for less than 5% of the company’s TPV.

While Currencycloud processes around $100 billion in transactions annually, a former senior product manager at Wise notes that the company still lags significantly behind Wise in terms of infrastructure, and that its take rate is roughly twice as high:

Tegus Client

Currencycloud, would you say they're the biggest competitor on the platform side before they were bought?

Former Senior Product Manager at Wise plc/ADR

Yes. I don't see anything good coming out of it since the acquisition. For me, it's a very stagnated product, and I don't understand why.

Tegus Client

Do you have any idea of the overall experience between the two platforms now? I don't know how the fees might look and I have no idea really how what the fee rates will be for Wise. Presumably, it will differ for each customer, but in general?

Former Senior Product Manager at Wise plc/ADR

I would say in general, double. Especially on the embedded routes where there is a central bank integration, Currencycloud is way over 1%. They can't even get below that because they're still reliant heavily on banking partners, but in those other areas where Wise had central bank integrations, they're really able to do 30 bps, 40 bps, just on the consumer level.

You've got additional discounts depending on the client that you get on board. If they take on the customer service and support role and they're doing that internally, then again, they're getting a much better rate because then Wise isn't offering that as a service. There's a lot of leverage there further. I would say, even on those routes, it could be 3X or 4X cheaper than Currencycloud can even offer without subsidizing.

Interestingly, according to him, Currencycloud was once considered the gold standard — and Wise lagged far behind. But today, the situation has changed dramatically:

Tegus Client

You paint a pretty bleak picture of Currencycloud then by comparison between two.

Former Senior Product Manager at Wise plc/ADR

It was the gold standard. The other thing that they did really well was they were able to plug things in quickly. Wise, it took years from an initial conversation to getting things on the rails. I know my team worked on a component of that which reduced things from nine months to a month, and it took us 18 months to build. I know how far behind Wise was.

When you now look at them achieving parity and then having these central bank integrations, plus the liquidity pools of having money coming in on those locations and real time in the amounts that they need them, through very targeted marketing, through very targeted customer relationship building, that is like a secret sauce that Currencycloud never even attempted to pick up or try and replicate.

They were just saying, "Look, here's what we can do. Here's the cheapest we can do it. Here's how you can integrate." It did the job, which everyone got excited and rightly so, because they went from something that cost 4% to something that cost 1.5%. That's revolutionary. They just lost their edge. I don't think it's completely bleak. I think they can still wake up and start playing the game. I think they just celebrated the acquisition and dropped the ball a bit.

Wise’s infrastructure is unique in its combination of depth and scale. As a result, the company has virtually no direct competitors in terms of transaction cost structure. Moreover, Wise continues to deepen its competitive moat by deliberately lowering its take rate as it scales. Management has publicly stated its long-term goal of making international transfers free. In other words, the ambition to be the cheapest and most transparent cross-border payment provider is not just a strategy — it’s part of the company’s founding culture.

And the strategy is working. Over the past five years alone, Wise’s active user base has grown 2.6x, reaching 15.6 million, while cross-border transfer volumes have increased 2.7x.

Around 67% of new customers join Wise through referrals and recommendation links shared by existing users. Amusingly, after Wise was discussed in one of our internal investment committee meetings, our trader turned out to be a customer — and promptly dropped his referral link into the team chat.

The company relies heavily on word-of-mouth growth rather than expensive marketing, which helps keep operating costs low. Despite increasing its marketing budget by 47% year-over-year in FY2025, total marketing spend amounted to just £53.8 million. Meanwhile, operating margin reached 47.8%, and the marketing-to-revenue ratio remains among the lowest in the publicly listed peer group.

As more users not only transfer money but also hold funds in their Wise accounts, interest income is becoming an increasingly important part of the company’s business model. In FY2025, interest income from customer balances totaled £594.3 million, compared to £840.4 million in cross-border revenue and an additional £371.5 million from card and other income. By year-end, total customer balances exceeded £17 billion, with £10.5 billion held by personal users and £6.6 billion by business clients.

This structural shift strengthens the resilience of Wise’s model: the more funds customers keep in their accounts, the higher the interest income — which the company can reinvest into further take rate reductions. But more on that in the next section.

To summarize, here are the key ingredients behind Wise’s low pricing:

• a well-developed proprietary infrastructure and direct integration with national payment systems;

• low customer acquisition costs driven by a high share of organic traffic;

• a growing volume of customer deposits.

Together, these factors form a resilient business model that enables the company to reduce its take rate without compromising profitability, while simultaneously widening its competitive moat. The low take rate, in turn, drives growth in transfer volumes and increases the balances held in Wise’s multi-currency accounts — a key source of interest income. This creates a self-reinforcing cycle: lower fees stimulate higher transfer volumes, which in turn allow Wise to lower prices even further — making its offering increasingly competitive over time.

Business Model and Growth Drivers

Wise operates across several interconnected segments, all built on a single platform but targeting different customer types and use cases. From a customer-type perspective, the company distinguishes the following segments:

• Personal (B2C). In this segment, Wise provides individuals with a platform for international money transfers, multi-currency storage, and usage — all with minimal fees and currency conversion at near-market rates. Customers can open accounts with local bank details in multiple currencies (Wise Account), receive cross-border transfers, convert currencies, and spend using a Wise debit card.

  
  The segment serves 14.9 million active customers. In FY2025, they transferred £106.4 billion in total, or about £7.2k per customer — accounting for approximately 73% of total volume. The aggregate balance held in Wise Accounts within this segment stood at £13.4 billion (averaging £899 per customer).

  
  Wise earns revenue through transfer fees, and card-related fees such as a one-time card issuance fee, interchange, and ATM withdrawal fees beyond a free limit.

  In Q1 FY2026, the cross-border take rate was 0.57% (vs. 0.63% in FY2025), and the card and other take rate was 0.26% (flat vs. FY2025), resulting in a total take rate of 0.83% (vs. 0.89%). In addition, Wise earns interest income by investing customer balances in liquid assets. The Personal segment accounts for ~78% of total revenue and underlying income (revenue + interest income on the first 1% of balances).
  

• Business (B2B). Wise’s B2B segment offers small and medium-sized businesses a multi-currency financial platform for cross-border payments, local account details in multiple currencies, payouts, international revenue management, and FX conversion at interbank rates with low fees.
  
  The segment serves 697,000 business customers, who transferred a total of £38.9 billion, or roughly £55.8k per customer. The total B2B balance reached £6.8 billion (~£9.76k per customer). The monetization model mirrors the B2C segment: Wise charges transfer fees, applies card fees, and earns interest income on deposited funds. In Q1 FY2026, the cross-border take rate was 0.39% (vs. 0.44% in FY2025), the card and other take rate was 0.23% (vs. 0.24%), and the total take rate was 0.62% (vs. 0.68%).
  

• Wise Platform (B2B2C, part of Business segment). This is Wise’s infrastructure-as-a-service product, allowing third-party companies to embed Wise’s cross-border payment capabilities via API-based white-label solutions. It is used by banks and fintechs that prefer to leverage Wise’s infrastructure and licenses instead of building their own global banking integrations. Currently, the platform generates a modest share of total volume — less than 5%.

From a revenue composition perspective, Wise categorizes its income as follows:

• Cross-border revenue: 46.5%

• Card and other revenue: 20.6%

• Underlying interest income (first 1% of balances): 8.3%

• Interest income above 1% threshold: 24.6%

The full revenue breakdown is shown below:

The B2C segment has always been the company’s primary growth driver.
Since 2019, the number of active customers in this segment has grown at a CAGR of ~28.8%, transfer volumes at ~29.4%, and revenue at ~37.2%. Total income from the segment — including both revenue and total interest income — has expanded at an impressive ~43.7% CAGR. Despite a declining take rate, the segment’s financial performance has significantly outpaced its operational metrics. A major contributor to this outperformance has been the Wise Account, a product launched in 2018 [22].

The Wise Account is the company’s flagship product and one of its core growth drivers. It’s a multi-currency account for personal users, allowing them to hold balances in over 50 currencies and convert between them at the real mid-market rate. Users can link a debit card for online payments and ATM withdrawals, send international transfers, and make purchases — all within the same account. Thanks to the Wise Account, card and other revenue grew from £7.5 million in 2019 to £279.7 million in 2025, increasing from 5% to 29.5% of B2C segment revenue.
According to our model, this share could reach approximately 48% of B2C revenue by 2030.

The Wise Account was designed to eliminate friction between a customer’s traditional bank account and transfers made through Wise. Here’s how a former financial controller at the company explains the product’s value proposition:

“The friction is getting from me in my pounds account, in HSBC, for example, into the Wise GBP account to get across the euro accounts... So you take that out and say, well, you keep a Wise account, you can have all these currencies in it. The proportion of them, you can have bank account details like your local in Australia or local in Canada or whatever. And therefore, you can get money in, and you can pay money out in all those accounts.

So that just reduces the friction of the whole flow and gives you that flexibility. And that was really the driver of it. And then when you're looking now it's billions on the balance sheet now in the Wise account because it's in the balance sheet. It's safeguarded money, it's on the balance sheet.”

The Wise Account is a key lever for lowering transfer fees and scaling the business further. The more funds customers hold in their Wise Accounts, the less frequent currency conversions become — and the cheaper transfers get. In other words, Wise uses customer funds held in Wise Accounts as a form of “localized liquidity.” These same funds also generate interest income. Total interest income from customer balances grew from £75.2 million in FY2023 to £360.1 million in FY2025.

The Wise Account has become the entry point into a broader product stack — including cards, auto-conversion, and B2B services. It helps Wise retain users, increase lifetime value, and lower customer acquisition costs:

“That was the explosive growth when Wise accounts came about... you can look back at the prospectus to see that.” — Former Group Financial Controller at Wise

Since Wise began investing in the Wise Account, product adoption has grown impressively. Roughly 53% of personal customers now use the account — up from just 24% adoption in 2022 [23]. According to Wise’s former Head of Pricing, adoption is likely to reach 70%. This would imply a near doubling of issued debit cards (as not all account users have a card), which could in turn double interchange fee revenue, currently accounting for about 10% of the company’s total revenue. These estimates are based solely on Wise’s existing customer base, without factoring in expected customer growth.

The Wise Account and debit card are also available to B2B clients, where adoption stands at ~60%, up from 49% in 2022.

Wise Business holds substantial potential as well: since 2019, active customer count in the segment has grown at a 34.2% CAGR, transfer volume at ~43.3%, and revenue at an impressive ~47.5% CAGR. As in the B2C segment, the faster growth in revenue relative to operational metrics is largely due to Wise Account adoption, which enabled card revenues to grow from 0% to 35% of segment revenue, and interest income in the segment to increase from £65 million in 2023 to £234.2 million in 2025.

Wise Business primarily serves small and medium-sized businesses (SMBs) — including freelancers, e-commerce merchants, software companies, and suppliers operating across borders. For this user cohort, international payments via traditional banks are often expensive, slow, and opaque. Wise Business addresses these pain points by offering real mid-market FX rates, transparent fees, and the ability to make instant payments. The platform allows users to create and send invoices, manage corporate expenses, and simplify payments across distributed teams — all within a single unified interface.

Since 2023–2024, Wise has been placing increased strategic focus on B2B, and the reason is clear: according to the company, SMBs transfer roughly 4.7x more money than individuals, while Wise’s market share in this segment remains below 1% [13].

Wise Platform also offers a vast runway for future growth. Through this product, Wise provides payment infrastructure that banks, fintechs, and other large enterprises can embed into their own services to enable fast and low-cost transfers. The partner retains full control over the customer experience: they display the FX rates (sourced from Wise), lock them in, initiate the transfer, and notify the end user. Wise then automatically processes the payment through its global network of local settlement accounts. In essence, Wise handles the technical delivery of funds along the optimal route, while the partner manages all customer-facing interactions.

Wise Platform clients include names such as Nubank, N26, Monzo, Google Pay, Bolt, Agoda (Booking), Brex, Deel, Spendesk, and Interactive Brokers. Notably, traditional banks are increasingly joining the platform, including EQ Bank, Itaú, Raiffeisen Bank, HSBC, IndusInd Bank, Standard Chartered, Morgan Stanley, and UniCredit. Notably, several of the most prominent banks joined the platform only recently: Standard Chartered in November 2024 [24], Morgan Stanley in December 2024 [25], Raiffeisen Bank in May 2025 [26], and UniCredit in July 2025 [32].

Today, Wise Platform accounts for approximately 4% of total payment volume (TPV).
The company’s management expects this figure to reach 10% in the medium term, and that the platform will eventually drive the majority of TPV over the long run. A former Senior Product Manager at Wise believes the Platform unlocks disproportionate opportunities for the company:

Tegus Client

Right. They talk about platforms still being very small today and they think it's going to become a big, a majority of volumes going forwards. Is that internal consistent with internal messaging in terms of really where they see the future opportunity? Should we really expect that the platform business is actually much more prosperous potentially than the majority of the consumer business that it is today?

Former Senior Product Manager at Wise plc/ADR

The focus has always been on the customers, because the customer is king. What that means is that everything is done for the customer first. Customers have funded Wise, they've invested in it, they've made it what it is today. The honoring of that is to try and give them the best possible rate in the world. The platform offering grows volume. The volume grows, that means that the consumers get a benefit of that because as the liquidity pools get bigger, as the volumes get bigger, all the underlying costs get driven down and contribute to a lower cost for them.

The platform offering has the biggest opportunity internally to grow volumes faster than anything else. That's why there's a big investment there, because signing up one big client can drive volumes in the same way as taking on a whole other country. That's where the disproportionate opportunities.

Beyond onboarding new fintechs and digital services, a clear growth driver is the potential migration of traditional banks to the platform. Given that Wise already partners with several large global banks, this scenario no longer seems far-fetched. Here’s what a Product Lead at N26 — the German neobank and a Wise Platform client — had to say on the matter:

Tegus Client

Okay. Could you imagine that the incumbents, really, the typical old banks, would move to Wise?

Product Lead at N26

To be honest, I think yes, for two things. First, on the international level, and I've been involved in a couple of discussion, there is a trial how can we move the corresponding bank into a more efficient way. There's a lot of progress in some of the corridors. In Europe, they are trying to do the SEPA instant scheme, one-leg transfer out that basically when you want to transfer funds externally, the payment doesn't go via SIWFT, it could be a SEPA instant until the corresponding bank to make it more efficient and cheap.

There's also similar things in the U.K. and U.S. We have these three big currencies all creating the infrastructure to leverage how the money can flow across continents. However, for the small currencies or the long tail, I think this infrastructure will not be in place for next 10 years which always going to grant why isn't social similar to them an edge compared to incumbent spending.

I believe that as the digital nature of this generation of millennials go to the premium of wealth into the business size more and more, they are aware of how these native digital solutions are built and how trusted they can be, increasing the overall volumes of Wise into the future. With this, I think once they see the drop or the volumes, incumbent will see, "Okay, what can we actually integrate to maintain a competitive offering and retain part of their business model revenue?" I think Wise going to be in the front line.

It’s worth noting that N26 initially partnered with Banking Circle, but the competitor’s infrastructure proved less mature: the promised FX functionality was not fully available, and implementation timelines kept slipping. As a result, N26 switched to Wise Platform in early 2024.

Thanks to its robust product ecosystem and unique infrastructure, Wise has multiple growth vectors. The B2C segment remains the company’s core business, with the Wise Account driving growth in customer balances, card revenues, and interest income. The B2B segment (Wise Business) also holds significant potential, given that the corporate money transfer market is much larger in size, while Wise’s current penetration remains low. Finally, the scaling of Wise Platform delivers a disproportionate contribution to TPV, lowers unit costs, and amplifies network effects. Together, these three segments reinforce one another, positioning Wise as a resilient and highly scalable business with a long runway ahead.

Financials and Valuation

The financial analysis and valuation section is typically the least exciting part of any investment thesis. However, it’s precisely the work with numbers that provides the clearest picture of how the business operates — and where its growth levers lie.

You can view and download our financial model via the link. The model includes seven key tabs:

• DCF – a full discounted cash flow valuation with sensitivity analysis;

• Multiples – historical and projected profitability metrics and valuation multiples;

• Growth – historical and forecasted trends in core operational and financial indicators;

• Inputs – detailed past and projected operational metrics by business segment;

• IS / BS / CFS – consolidated income statement, balance sheet, and cash flow statement for historical and forecast periods.

The core assumptions are outlined on the "Inputs" tab, as well as at the bottom of the Income Statement, Balance Sheet, and Cash Flow Statement pages. Assumptions are highlighted in yellow. Our projections are based on a combination of historical trends, management commentary, and our own estimates.

Following the latest reporting period, Wise provided guidance indicating that it expects underlying income (revenue + interest income on the first 1%) to grow at a CAGR of 15–20% over the medium term and for FY2026. Wise also expects to operate with an underlying profit before tax margin of 13–16%, and for FY2026, management expects to be near the top of that range. The company additionally reaffirmed its commitment to investing ~£2 billion into infrastructure, product, and marketing over the coming years [27].

Management guidance implies a 7.3% year-over-year decline in underlying profit before tax in the optimistic scenario (assuming 20% YoY growth in underlying income and a 16% underlying PBT margin), and a 29% YoY decline in the downside case (15% YoY income growth and a 13% margin). It’s worth noting that in Q1 FY2026, underlying income grew by 11.2% YoY. The slowdown was primarily driven by a 19% YoY drop in cross-border take rate, from 0.64% to 0.52%.

In building our model, we followed a conservative approach. We forecast underlying income growth of 14.9% YoY and an underlying PBT margin of 16.2% for FY2026 — 21 bps above the upper end of management guidance.

It’s worth noting that we model administrative expenses of £1.86 billion (excluding D&A) and CapEx of £54 million over the next two years, which is roughly in line with the company’s expected £2 billion in investments (Wise is not a capital-intensive business; most of the investment will flow through the P&L). Approximately 57% of administrative expenses are projected to come from employee benefits.

Projected trends in operating and financial performance are outlined below:

Based on our model, underlying PBT is expected to decline by 10% YoY in 2026, compared to 16.7% growth in 2025. The anticipated decline in bottom-line performance and margin compression may put pressure on the stock in the short term.

While Wise trades at a premium to most remittance peers, we believe the current valuation levels are reasonable for initiating a position: the company is trading at a 2025 FCF yield (to EV) of ~4.61% and a 2026 FCF yield of ~5.84%, with expected Free Cash Flow growth of 18.2% in 2026 and an estimated FCF CAGR of ~17.2% through 2030.

Valuation multiples are shown below:

According to our DCF valuation, the fair value of Wise shares is approximately £13.2, which is about 28% above the current price.

Sizing: we are initiating approximately one-third of the position at the current price, targeting a maximum allocation of ~10% of NAV.

About Crypto Risk

Crypto is arguably the most prominent risk factor that concerns investors in fintech. The market is constantly seeing the emergence of new crypto startups aiming to disrupt traditional MTOs and fiat-based fintechs — from the launch of Ripple in 2012 to Coinbase’s crypto remittance platform in 2022 [28], and Kraken’s launch of Krak in June 2025 [29].

And while crypto enthusiasts promise cheap, instant, and decentralized money transfers, crypto startups have failed to capture even 1% of any segment of the market. The reasons behind this failure go deeper than just regulatory barriers:

• The last-mile problem. The end recipient doesn’t need a token — they need money. Receiving USDC, XRP, or Bitcoin isn’t enough. The token must be exchanged, cashed out, or spent somewhere — all of which introduces additional transactional costs. A textbook example is the Ripple–MoneyGram partnership from 2019 to 2021 (Ripple even acquired a 10% stake in MoneyGram) [30]. The transfer process worked as follows:

Instead of directly converting fiat currencies, MoneyGram was buying and selling XRP — the cryptocurrency issued by Ripple. Due to the relatively low trading volume of XRP, its FX spread was significantly higher than that of fiat currencies — and the spread had to be paid twice.

Another example is the partnership between Coinbase and Remitly, launched in 2022. It aimed to solve the last-mile problem. While the exact FX margin under this partnership is unknown, Remitly’s CEO noted that the deal is highly profitable for the company:

“And I can't talk about Coinbase specifically, obviously. But what we like about that business is the fact that from a profitability perspective, it's very advantageous to us and their strategic benefits to more transactions flowing through our system because it gives us more leverage with our partners the ability to invest more in our network, which helps both our remittance consumer business as well as the overall RFD offering. So mix of fee and foreign exchange generally with a nice profitability profile on a unit economic basis for that business.”

Thus, the last-mile problem has so far always been solved the same way — by involving traditional MTOs or fintech companies. So the question is: what value do crypto providers actually add in such operations? Arguably, it’s mostly marketing.

• The question of economic feasibility. Decentralized systems rely on miners or validators to operate, and these participants must be compensated for processing transactions. Of course, fees on networks like Tron or Solana are negligible (less than $0.001). However, MTOs and fintechs also incur very low costs for moving money internationally. For example, Wise pays bank and partner fees amounting to just 0.2% of volume. Theoretically, even if these costs increased several times over, Wise could absorb them without significantly impacting its margins.

  
  The majority of expenses for MTOs and fintechs are related to payments for receiving and disbursing funds, customer acquisition, customer service, risk management, and KYC/AML compliance. Even if new operators emerged with blockchain-based infrastructure, they would still face the same cost structure. And given the last-mile problem and additional transactional costs, their take rate would likely be significantly higher than Wise’s.

Thus, as of today, crypto does not pose a meaningful risk to Wise. It offers a weak value proposition — both for end users and as an infrastructure solution. Any savings on network fees are marginal compared to the FX margin from double conversions. It’s worth noting that Wise’s management is regularly asked about the risks they see in crypto. Here’s what Kristo Käärmann said on the topic in November 2024:

Analyst

This past month, we saw Stripe do an acquisition in the Stablecoins space. Just curious how over the next, I don't know, 5 years, you think Stablecoins will play a role with Wise?

Kristo Käärmann

I can't -- every time there is a crypto or stable coin question, I think our opinion hasn't changed, that if we see something that enables the movement of funds faster, cheaper and more conveniently, we will totally consider that. So the stance hasn't changed. But we also have nothing to report that we would have seen anything interesting. So I believe their circumstances are maybe perhaps why they did that acquisition. Must be something different than ours.

Conclusion

Wise is a rare example of a company that doesn’t just compete in a vast and fragmented market — it is fundamentally reshaping it. Through years of infrastructure investment and consistent fee reduction, the company has built a resilient model with a deep competitive moat. Direct integration with local payment systems, organic customer acquisition, and a scalable operating model enable Wise to lower prices while remaining one of the most profitable players in the sector.

The core of Wise’s model is a self-reinforcing cycle: lower fees drive higher transfer volumes, which in turn enable further price cuts, making the offering increasingly competitive. This strategy cannot be easily replicated without comparable scale and years of infrastructure development.

At the same time, Wise remains founder-led, continues to reinvest in its product, is expanding in B2B and platform-based solutions, and maintains strong financial discipline. While most competitors either operate at smaller scale or with higher unit costs, Wise is strengthening its position in both the consumer and corporate segments.

Wise is an infrastructure asset in an active growth phase, with a vast runway ahead. We include Wise in our high-conviction ideas cohort with a position size of up to 10% of NAV. In our view, the current price presents a reasonable entry point.

Sources

[1] https://moneyweek.com/economy/people/603574/the-story-of-wise-a-multi-billion-fintech-started-by-accident?utm_source=chatgpt.com

[2]https://tracxn.com/d/companies/wise/__YAuKfycFmfMvs6MmH65xgzFBaakq0ec3oksdVFrUAWI/funding-and-investors?utm_source=chatgpt.com#funding-rounds

[3] https://wise.com/imaginary-v2/images/5b2bb84ff23eb690fda40fa9a92c27ed-HY_FY25_Sep-24_Presentation.pdf

[4] https://www.londonstockexchange.com/news-article/WISE/holding-s-in-company/16609867

[5] https://wise.com/imaginary-v2/images/b4e39297919e9749b16f13f504929e75-Wise_2024_Annual_Report_and_Acounts.pdf

[6] https://www.mckinsey.com/industries/financial-services/our-insights/banking-matters/how-banks-can-win-back-lower-value-cross-border-payments-business

[7] https://www.worldbank.org/en/news/press-release/2024/06/26/remittances-slowed-in-2023-expected-to-grow-faster-in-2024#:~:text=Sending%20remittances%20remains%20too%20costly,technological%20advancements%20in%20reducing%20the

[8] https://www.fxstreet.com/analysis/the-market-moving-data-this-week-comes-from-everywhere-other-than-the-us-202405201254#:~:text=%E2%80%9CThe%20euro%20and%20U,%E2%80%9D

[9] https://wise.com/imaginary-v2/images/915eb3b394df2a0159500f72b8034ff2-2023%20Remittances%20Report%20-%20FINAL.pdf#:~:text=networks%20who%20can%20operate%20with,Furthermore%2C%20compliance%20with

[10] https://wise.com/imaginary-v2/images/915eb3b394df2a0159500f72b8034ff2-2023%20Remittances%20Report%20-%20FINAL.pdf#:~:text=While%20there%20is%20evidence%20that,traditional%20banking%20associated%20with%20higher

[11]https://remittanceprices.worldbank.org/sites/default/files/rpw_main_report_and_annex_q224.pdf

[12] https://www.saveonsend.com/bank-money-transfer/?utm_source=chatgpt.com

[13] https://wise.com/imaginary-v2/images/a743105a05bf55701e3690fab53bc43d-WiseFY25AnalystPresentation.pdf

[14] https://d18rn0p25nwr6d.cloudfront.net/CIK-0001365135/0ec0aec2-5987-44ee-8629-d6e6e0ce56a1.pdf

[15] https://wise.com/imaginary-v2/images/7225a78f5d177b9bba2c8152e664ee7e-Wiseplc-FY2025.pdf

[16] https://www.saveonsend.com/transferwise-money-transfer/

[17] vimeo

[18] https://www.saveonsend.com/bank-money-transfer/?utm_source=chatgpt.com

[19] https://www.congress.gov/bill/119th-congress/house-bill/1/text

[20] https://www.airwallex.com/newsroom/airwallex-raises-usd300-million-at-a-usd6-2-billion-valuation-to-build-the-future-of-global-banking

[21] YouTube

[22] https://wise.com/gb/blog/transferwise-debit-card-launch?utm_source=chatgpt.com

[23] https://wise.com/imaginary-v2/FY23-Mar-23-Analyst-Presentation-(June-27-2023).pdf

[24] https://newsroom.wise.com/en-CAS/243185-standard-chartered-partners-with-wise-platform-to-transform-international-payments-experience?utm_source=chatgpt.com

[25] https://wise.com/gb/blog/wise-morgan-stanley?utm_source=chatgpt.com

[26] https://wise.com/gb/blog/raiffeisen-bank-international-and-wise-platform

[27] https://wise.com/imaginary-v2/images/7225a78f5d177b9bba2c8152e664ee7e-Wiseplc-FY2025.pdf

[28] https://www.bankingdive.com/news/coinbase-muscles-in-on-popular-remittance-corridor/618968/?utm_source=chatgpt.com

[29] https://www.reuters.com/business/crypto-exchange-kraken-debuts-peer-to-peer-payments-app-krak-2025-06-26/?utm_source=chatgpt.com

[30] https://www.reuters.com/article/business/blockchain-firm-ripple-to-end-partnership-with-moneygram-idUSKBN2B02IM/?utm_source=chatgpt.com

[31] https://www.saveonsend.com/bitcoin-blockchain-money-transfer/

[32] https://wise.com/gb/blog/unicredit-chooses-wise-platform

[33] Bloomberg