I bought Western Digital at $463 at the market opening

Read this disclaimer

The business continues to progress, but the share price has fallen big. Western Digital closed at $519.17 on August 5.

I bought the shares at $463 after the earnings release.

This price was about 11% below the previous closing price.

The shares also reached an intraday high of $799.87 on June 18.

My purchase price is about 42% below this high. For me, this is a large sell-off.

However, the latest financial results, customer demand, financial guidance and product roadmap still support the investment case. Revenue increased 44%. Non-GAAP earnings per share increased 109%, and free cash flow reached $1.28 billion.

Management also expects the next quarter to produce higher revenue, a higher gross margin and higher earnings per share. The business did not become weaker.

But the market expected another extra large increase in estimates. It received a strong quarter instead. The problem with the current market is that it just need proof the AI trade is not over yet to support current valuations, but in the same time, want massive quater as if we were at the starting point.

Western Digital was one of the strongest AI infrastructure stocks during 2026. The market now views it as an expensive storage company with weak share-price momentum and a later HAMR roadmap than Seagate.

This type of situation usually gets my attention.

Why now

The share price increased too quickly before the recent decline. Western Digital more than tripled during 2026 before the earnings report. The shares also increased much faster than the semiconductor index and the wider market.

The market did not only expect strong growth but continuous acceleration. Which is not possible. WD has to take another wave of demand before adding revenue perspectives.

Western Digital reported revenue and earnings above published analyst estimates. Management also gave next-quarter revenue guidance above the market consensus.

But still that was not enough.

Investors started to question whether current pricing growth and margin expansion can continue at the same speed. The market also compared Western Digital with Seagate.

Seagate already has HAMR products in volume production with hyperscale customers. Western Digital expects to start the production ramp of its HAMR drives during 2027.

This difference is real but Western Digital also shipped 231 exabytes during the quarter, compared with 190 exabytes one year earlier. This is growth of about 22%.

Some investors compared this result with Western Digital’s long-term storage growth target of more than 25%. However, this target applies across several years. It is not a target for each quarter.

The company also expects a non-GAAP gross margin of between 55% and 56% during the next quarter. The midpoint is above the 54.4% margin reported during Q4, but the market wanted a larger increase.

These concerns are valid. However, they do not show that Western Digital has a demand problem. In my opinion, they show that expectations were too high. They also show that Seagate currently has an earlier position in the HAMR transition.

The market sold this relative weakness as if Western Digital had reported an absolute business problem.

I think the market sold too much, too fast.

Fundamental analysis

Western Digital supplies storage infrastructure for data centers. The company completed the separation of its Flash business in February 2025. Sandisk is now a separate company, and Western Digital is now mainly a hard disk drive business.

Western Digital sells high-capacity hard disk drives and storage platforms to hyperscalers, cloud service providers and large companies. In simple terms, Western Digital helps customers store very large quantities of data at a low cost.

This function is important for AI.

GPUs perform calculations. Flash storage provides fast access to active data. Hard disk drives store the much larger quantity of data that companies must keep for long periods.

AI systems use training data. They also create inference logs, model outputs, images, video, embeddings and user records.

This data continues to accumulate even after the calculation.

Western Digital states that flash storage can have a cost premium of six to ten times compared with hard disk drives for some large-capacity workloads. Hard disk drives therefore remain important when customers need storage at a very large scale.

Cloud customers produced 89% of Western Digital’s latest quarterly revenue. Client products produced 6%, and Consumer products produced 5%.

Cloud revenue is not the same as direct AI revenue. Cloud customers also store data for many other applications.

However, the company is now mainly connected to data centers. This is not the old Western Digital investment case. The business no longer depends mainly on personal computers and consumer storage products.

The latest financial results are strong.

• Fourth-quarter revenue reached $3.747 billion. This was an increase of 44% year over year and 12% from the previous quarter.

• Non-GAAP gross margin reached 54.4%, compared with 41.3% one year earlier.

• Non-GAAP operating margin reached 44.2%, compared with 28.1% one year earlier.

• Non-GAAP earnings per share increased from $1.70 to $3.56.

Cash generation was also strong.

• Operating cash flow reached $1.389 billion during the quarter. Capital expenditure was $108 million. Free cash flow therefore reached $1.281 billion.

• This was a quarterly free cash flow margin of about 34%.

• The full-year results show the same direction. Fiscal 2026 revenue increased 36% to $12.919 billion. Non-GAAP operating income increased 107% to $4.817 billion.

• Non-GAAP earnings per share increased 104% to $10.22. Free cash flow increased from $1.432 billion to $3.511 billion.

This is not weak operating performance.

Revenue is growing. Margins are expanding. Earnings and cash flow are growing faster than revenue.

Management also gave strong guidance.

Western Digital expects first-quarter revenue of $4.1 billion, plus or minus $100 million. This represents expected growth of between 42% and 49% year over year.

Management also expects a non-GAAP gross margin of between 55% and 56%. Non-GAAP earnings per share are expected to reach $4, plus or minus $0.15.

The next quarter is therefore expected to be stronger than the latest quarter.

And the market sold the stock anyway.

Western Digital also has a stronger balance sheet. The company finished fiscal 2026 with $1.579 billion of cash and $1.052 billion of debt. It therefore had more cash than debt.

Western Digital spent $2.592 billion on share repurchases during fiscal 2026. It also paid $184 million of dividends.

This is not a company that needs external capital to support normal operations. The business produces enough cash to invest, repay financial obligations and return capital to shareholders.

However, the GAAP earnings figure requires attention.

Western Digital reported quarterly GAAP net income of $3.195 billion. This figure included a $2.05 billion gain from its former investment in Sandisk.

This gain did not come from the normal hard disk business. For this reason, I focus on non-GAAP operating income, non-GAAP earnings and free cash flow.

These measures also increased sharply.

Market sentiment

Market sentiment has moved from strong optimism to fear.

Your fear is my greed.

Western Digital became one of the strongest AI infrastructure stocks in the market. The shares more than tripled during 2026 before the latest results.

Investors focused on strong cloud demand, limited storage supply, higher prices and rapid margin expansion. This created very high expectations.

Western Digital then reported 44% revenue growth, a 54.4% non-GAAP gross margin and $1.28 billion of quarterly free cash flow.

These results were strong. But the market had already priced in more.

After the market opened on August 6, the shares fell sharply. The decline showed the very high standard that the market now applies to companies connected to AI spending.

This decline was not limited to Western Digital. Sandisk, Seagate, Micron and other semiconductor companies also fell. Investors reduced risk across the storage and semiconductor sector.

The wider demand signal did not disappear. AI data-center component demand remains strong. Investors are mainly concerned about the possible normalization of growth and pricing. The end of the cycle.

The market is not saying that storage demand has ended. The market is saying that the rate of improvement cannot continue forever.

I agree with this point.

Storage is a cyclical industry. Prices, margins and market expectations will not increase in a straight line.

But I did not buy near $800. I bought at $463 after expectations started to collapse.

The fear is visible again.

Momentum

The business has strong momentum.

Quarterly revenue increased from $2.605 billion to $2.818 billion, then to $3.017 billion, $3.337 billion and $3.747 billion.

During the same period, non-GAAP gross margin increased from 41.3% to 43.9%, then to 46.1%, 50.5% and 54.4%.

This is strong operating momentum.

Total exabyte shipments increased from 190 to 204, then to 215, 222 and 231 exabytes.

Nearline exabyte shipments increased from 170 to 183, then to 192, 199 and 209 exabytes. Nearline drives are the high-capacity drives that cloud customers use in large data centers.

The product roadmap is also moving forward.

• Western Digital’s 40-terabyte UltraSMR ePMR drive is in qualification with two hyperscale customers. The company expects volume production during the second half of 2026.

• Western Digital is also qualifying HAMR drives with two hyperscale customers. Management expects the HAMR production ramp to start in 2027.

• The company plans to extend ePMR capacity to 60 terabytes. It also plans to increase HAMR capacity to more than 100 terabytes by 2029.

• This dual roadmap is important.

Western Digital does not need to stop selling ePMR products while it completes HAMR development. Customers can use the current architecture and move to HAMR on their own schedules.

Cloud customers do not select a drive only because it has the highest capacity. They also consider reliability, power use, qualification time, software compatibility, supply stability and total cost.

Western Digital already has long customer relationships and approved product platforms. This can give the company time to complete its HAMR transition.

However, Seagate has a real lead.

Seagate announced that its 44-terabyte HAMR drives were already in production with two hyperscale cloud providers. Western Digital must prove that its later production schedule will not create a permanent cost or capacity disadvantage.

The share-price momentum is now weak.

The stock fell from almost $800 to my purchase price of $463. It then traded lower during the same session.

This does not prove that the operating business has become weaker. It shows that investors want more proof and less narrative.

Risks

The first risk is the valuation.

At $463 and about 388 million diluted shares, Western Digital has an implied equity value near $180 billion. The stock trades at about 45 times fiscal 2026 non-GAAP earnings per share.

This is not a cheap multiple.

The share-price decline made the valuation more reasonable than it was near $800. It did not make Western Digital a deep-value stock.

The company must continue to produce strong revenue growth, high margins and high free cash flow.

The second risk is the storage cycle.

The current market has strong cloud demand and limited high-capacity supply. These conditions support prices and margins.

A future increase in production capacity can change this situation. Lower prices can reduce earnings and free cash flow quickly.

The third risk is customer concentration.

A small number of hyperscalers produce a large part of Western Digital’s revenue. These customers have strong purchasing power.

One customer can delay orders, change suppliers or demand lower prices. These decisions can have a material effect on Western Digital.

The fourth risk is the dependence on Cloud.

Cloud produced 89% of the latest quarterly revenue. This gives Western Digital strong exposure to data-center growth, but it also creates dependence on cloud investment.

A slowdown in hyperscaler spending would damage the investment case.

The fifth risk is HAMR execution.

Seagate is already producing HAMR drives at scale with hyperscale customers. Western Digital must complete qualification and increase production without large delays, poor yields or high manufacturing costs.

The final risk is expectations.

The stock remains much higher than it was at the start of 2026. A 42% decline from the high does not automatically make the stock cheap.

If revenue growth slows or gross margin stops increasing, the market can reduce the valuation again.

This is why I bought after the reset at $463. I was not interested near $750 or even $600.

Why I am buying

I am buying because the main investment case remains intact.

AI does not only need more computing power. AI also creates more data, and this data must be stored.

Western Digital supplies one of the lowest-cost ways to store very large quantities of data. The company now produces 89% of its revenue from Cloud.

Revenue increased 44%, and non-GAAP gross margin reached 54.4%. Non-GAAP operating income increased 126%, and free cash flow reached $1.28 billion in one quarter.

Management expects the next quarter to produce approximately $4.1 billion of revenue, a gross margin of up to 56% and non-GAAP earnings per share of approximately $4.

These are not the results of a broken company.

The market is correct about some points.

Western Digital is behind Seagate in the commercial HAMR transition. The valuation is still high, and the stock was extremely crowded before the decline.

The storage market is also cyclical.

But the market sold a relative technology problem and a valuation problem as if Western Digital had an absolute demand problem.

Western Digital did not report lower demand. It reported higher revenue, higher margins, higher earnings and higher free cash flow.

The company is qualifying 40-terabyte drives and its future HAMR products with hyperscale customers.

The price at $463 includes more fear and lower expectations.

GPUs perform the calculations. Western Digital stores part of the data that these calculations use and create.

More AI use creates more data. More data creates more storage demand.

In any case, storage is needed.

The market sold too much.

The research is complete because the fear is visible and the business continues to grow.

Disclaimer

This article is for information and educational purposes only. It does not constitute financial, investment, legal or tax advice.

The author owns shares of Western Digital Corporation ($WDC) and bought the security at approximately $463 per share. The author may buy or sell the security at any time without notice.

All investments involve risk, including the possible loss of capital. The information in this article is based on sources considered reliable at the time of publication, but its accuracy and completeness cannot be guaranteed.

Readers must conduct their own research and make their own investment decisions. When necessary, they should consult a qualified financial adviser.

I am buying Lemonade at this price, at en entry trigger of $49 per share.

Read this disclaimer

The business continues to grow, but the share price has fallen sharply. Lemonade closed at $78.70 on July 6. The shares now trade near $56. This is a decline of about 29% in less than one month.

The current price is also about 44% below the 52-week high of $99.90. This is more than a small market correction.

However, Lemonade has not reported a new operating problem. The latest official results show faster premium growth, better insurance performance, higher gross profit, and lower operating losses.

In Force Premium increased 32%. Revenue increased 71%. Gross profit increased 159%, and the adjusted EBITDA loss improved by 64%.

The company also increased its full-year guidance. Management continues to expect its first positive adjusted EBITDA quarter in Q4 2026.

The market previously viewed Lemonade as a popular AI insurance company. The market now views it as an expensive insurer that is still losing money.

This type of situation usually gets my attention.

Why now

The share price has fallen in several large moves.

• On July 6, the shares increased 10% and closed at $78.70.

• On July 8, they fell 8.5%.

• On July 15, the shares fell another 5.8%.

• On July 22, they fell 7.6%, and they lost another 4.3% on July 23.

The market activity in July gives a clear message. Investors first increased their exposure after strong operating results. They then reduced this exposure when the valuation became too high.

The business did not change during these few weeks. The market price changed.

Investors are also waiting for the next quarterly results. Lemonade must prove that its strong growth can produce real profitability and standard free cash flow.

The market now wants proof.

This is reasonable. But the shares already include much less optimism than they did near $79.

Fundamental analysis

Lemonade supplies insurance through a digital platform.

Its products include:

• Renters insurance

• Home insurance

• Pet insurance

• Car insurance

• Life insurance

Lemonade uses artificial intelligence and software to sell policies, calculate prices, support customers, and process claims.

In simple terms, Lemonade wants software to complete work that traditional insurers complete with large operating teams.

This is important because insurance includes many repeated tasks. The insurer must collect customer information, calculate a price, create a policy, answer questions, identify fraud, and process claims.

A company that automates these tasks can add customers without increasing employee numbers at the same speed.

The latest financial results support this idea.

In Force Premium reached $1.33 billion in the first quarter. This represented 32% growth and the tenth consecutive quarter of faster In Force Premium growth.

Lemonade also reached 3.14 million customers. Customer numbers increased 23%, while premium per customer increased 7% to $424.

Annual dollar retention reached 85%. Lemonade is therefore adding customers while also increasing the average value of each customer relationship.

Gross earned premium increased 31% to $306.2 million. Revenue increased 71% to $258 million.

The difference between these two growth rates is important.

Revenue increased faster because Lemonade now transfers less premium to reinsurance companies. The company keeps more premium and more insurance risk on its own income statement.

The 71% revenue growth rate does not mean that the underlying insurance book increased 71%. In Force Premium and gross earned premium are better measures of underlying growth.

These measures increased 32% and 31%.

This is still strong growth.

Gross profit increased from $38.6 million to $100.1 million. Gross profit margin increased from 26% to 39%.

The adjusted EBITDA loss improved from $47 million to $17.1 million. Net loss improved from $62.4 million to $35.8 million.

Lemonade is still losing money.

However, the loss is becoming smaller while the business continues to grow.

Insurance performance

An insurance company can increase revenue by accepting bad risks or by charging prices that are too low.

Revenue growth alone is not enough.

The loss ratio shows how much premium the insurer uses to pay claims and related costs. A lower ratio normally means better insurance performance.

Lemonade’s gross loss ratio improved from 78% to 62%. Its net loss ratio improved from 82% to 63%.

The trailing twelve-month gross loss ratio also reached 61%. This is a much better level than the company reported during its earlier growth period.

This result is important because Lemonade is not producing faster growth through weaker underwriting.

Homeowners insurance had a gross loss ratio of 49%. Pet had a ratio of 69%, and Car reported 74%.

Europe remained weaker, with a gross loss ratio of 85%. This part of the business still requires improvement.

Insurance results can change from one quarter to the next. Weather events and reserve adjustments can create large movements.

But the wider trend is clear.

The insurance book is growing, while the long-term loss ratio is moving in the correct direction.

The AI advantage

Many companies now use the word AI in their investor presentations.

For Lemonade, AI is not only a product feature. It is the operating system of the company.

Lemonade uses AI for customer acquisition, pricing, support, fraud detection, and claims processing. The objective is to increase the number of policies without adding the same number of employees.

The company finished the first quarter with more than $1 million of In Force Premium per employee.

Since the fourth quarter of 2022, Lemonade has more than doubled its In Force Premium while reducing its employee count by 6%.

This is real operating leverage.

It does not mean that every AI model works perfectly. It does not mean that Lemonade has solved insurance.

It means that the company can support more premium with a similar operating structure.

Lemonade also increased its growth spending by about 200% since Q1 2023 while keeping its estimated lifetime value to customer acquisition cost ratio near or above three times.

The company believes its AI pricing and marketing systems allow it to increase spending without reducing acquisition quality.

This must continue.

If customer acquisition becomes more expensive or if the new customers produce weak insurance results, the advantage will disappear.

For now, the operating evidence remains strong.

Pet and Car

Pet is now Lemonade’s largest insurance product.

The business passed $500 million of In Force Premium early in the second quarter. Lemonade also describes itself as the fourth-largest U.S. pet insurance carrier by written premium.

The company reports a claims adjustment expense ratio near 4% for Pet. Lemonade believes automation gives it a cost advantage against traditional insurers.

Car is smaller, but it has faster momentum.

Car In Force Premium increased 60% year over year. This was the strongest growth rate since Lemonade launched the product.

The Car gross loss ratio reached 74%. This remains high, but it is much better than the 88% level reported one year earlier.

Lemonade also started to sell insurance for autonomous driving.

The company reported that customer conversion for its autonomous product was about 70% higher than for its comparable traditional Car product.

This product is still small. However, it shows how Lemonade can use direct vehicle data to offer different prices for autonomous and human-driven miles.

The long-term opportunity is important.

If autonomous vehicles produce fewer accidents, insurers with direct access to vehicle data can price this risk more accurately.

Market sentiment

Market sentiment has moved from strong optimism to fear.

Your fear is my greed.

The shares moved from below $59 on June 26 to almost $79 on July 6. Investors focused on AI, faster premium growth, better insurance performance, and the expected move toward adjusted EBITDA profitability.

The share price then lost most of this increase.

Investors now focus on the net loss, the use of non-GAAP financial measures, the cost of growth financing, and the absence of standard free cash flow.

These concerns are real.

Lemonade has not reached GAAP profitability. It also remains difficult to value because a large part of its value depends on future growth and future margins.

However, the operating progress is also real.

Customer growth remains above 20%. In Force Premium growth has accelerated for ten consecutive quarters. Loss ratios have improved, and the adjusted EBITDA loss is becoming smaller.

The market moved from excessive optimism to caution.

For me, the current setup is more attractive than it was near $79.

Momentum

The business has strong momentum, but the share price has weak momentum.

Management expects second-quarter In Force Premium of between $1.428 billion and $1.433 billion. It expects revenue of between $287 million and $290 million.

Management also expects a second-quarter adjusted EBITDA loss of between $19 million and $23 million.

For the full year, management expects In Force Premium of between $1.632 billion and $1.639 billion. This would represent growth of about 33%.

Full-year revenue is expected to reach between $1.197 billion and $1.203 billion.

Management expects a full-year adjusted EBITDA loss of between $47 million and $51 million. It continues to expect positive adjusted EBITDA during the fourth quarter.

This will be the next important test.

A positive quarter will not make Lemonade fully profitable. But it will prove that the company can produce operating leverage while growth remains high.

The share-price momentum is different.

The stock has fallen about 29% from its July high. It also trades about 44% below its 52-week high.

The chart is weak.

This does not prove that the operating business has become weaker.

It shows that investors want more proof.

Risks

The investment has important risks.

Lemonade is still unprofitable. The company reported a net loss of $35.8 million during the first quarter.

Adjusted EBITDA does not include stock-based compensation, interest, depreciation, and some other costs.

Positive adjusted EBITDA will not automatically mean positive net income.

Cash flow also requires careful analysis.

Lemonade reported adjusted free cash flow of $17.4 million. However, the company adds net borrowings under its financing agreement to this measure.

Standard operating cash flow was negative by $0.6 million. Capital expenditure was $3.5 million.

Standard free cash flow was therefore negative by $4.1 million.

The positive adjusted figure included $21.5 million of net new financing. This is not cash generated by the operating business.

This point is important.

The company must eventually finance customer growth with its own operating cash flow.

Lemonade had $179.6 million of borrowings under its existing growth-financing agreement at the end of the first quarter.

The company also signed a new agreement with Hannover Re. This agreement can provide up to $250 million of capital for growth spending during 2027 and 2028.

This financing can reduce the need for immediate equity issuance.

But it is not free capital.

The financed customers must produce enough future value to pay the financing cost and create a return for Lemonade shareholders.

Stock-based compensation is another risk.

Lemonade reported $21.2 million of stock-based compensation during the quarter, compared with $10.3 million one year earlier.

Insurance risk also remains important.

Hurricanes, wildfires, and other large events can increase claims. Car and Pet can also produce large losses if Lemonade sets prices incorrectly.

The valuation is not very low.

Near $56, Lemonade has a basic market value of about $4.3 billion. This represents approximately 3.6 times the midpoint of management’s 2026 revenue guidance.

This valuation requires continued growth.

Lemonade must also improve insurance margins, reduce operating losses, and produce standard free cash flow.

A delay in the Q4 adjusted EBITDA target would damage the investment case.

Why I am buying

I am buying because the main investment case remains intact.

Lemonade is adding customers. Premium per customer is increasing, and In Force Premium growth has accelerated for ten consecutive quarters.

Revenue and gross profit are growing faster than operating expenses. Insurance loss ratios are improving, and the adjusted EBITDA loss is becoming smaller.

The company is also showing real operating benefits from AI.

Lemonade has more than doubled its In Force Premium since the end of 2022 while reducing its employee count.

This is not an insurance company that only added an AI chatbot to an old system.

Lemonade built its complete business around software, automation, and data.

More customers create more data. More data can improve pricing, fraud detection, and claims decisions.

Better decisions can reduce losses and improve customer economics.

The company still has important problems.

It does not produce positive standard free cash flow. It uses external capital to finance part of its growth, and it continues to report a GAAP loss.

But the direction is correct.

Growth is accelerating. Insurance performance is improving, and operating leverage is becoming visible.

The shares have fallen from almost $79 to near $56 without a new weak earnings report.

The market sold the valuation.

It did not sell a weaker business.

The price now includes more fear and lower expectations.

The next quarterly results must confirm the guidance. Lemonade must also deliver positive adjusted EBITDA in the fourth quarter.

If it does, the market will have to reconsider the profitability thesis.

Market sold too much.

The research is complete because the fear is visible and the business continues to grow.

We are literally buying fear

Disclaimer

This article is for information and educational purposes only. It does not constitute financial, investment, legal or tax advice.

The author own shares of Mirion Technologies, Inc. ($MIR) and may buy or sell the security at any time without notice.

All investments involve risk, including the possible loss of capital. The information in this article is based on sources considered reliable at the time of publication, but its accuracy and completeness cannot be guaranteed.

Readers must conduct their own research and make their own investment decisions. When necessary, they should consult a qualified financial adviser.

I am buying Mirion at this price, near $15.

Read this disclaimer

The world needs more electricity, and AI data centers are increasing this need. The electrical grid cannot add enough stable power quickly, so nuclear energy is becoming part of the solution.

Mirion does not build nuclear reactors but supplies the detection, monitoring and safety systems that reactors need to operate.

Its products are present in more than 98% of nuclear power plants worldwide. More than 80% of its nuclear power revenue comes from reactors that already operate today.

The stock traded above $30 during the last year. It now trades near $15, which is a decline of about 50%.

For me, this is more than a normal correction.

The market expected the nuclear theme to produce rapid revenue and cash growth. The orders arrived, but these orders are taking more time to become revenue and cash.

The market sold this delay as if it were a demand problem.

I think it is a timing problem.

Why now

The nuclear sector became very popular during 2025. Investors expected AI electricity demand, government support and small modular reactors to produce fast financial results.

Mirion moved above $30 during this period. The story became more popular than the financial results.

The company then reported moderate organic growth and slow cash conversion. Investors reduced their exposure because the results did not move as quickly as the nuclear narrative. And I insist, narrative.

But nuclear projects are slow by design.

A reactor project requires engineering, equipment qualification, regulatory approval, construction and testing. An order can enter Mirion’s backlog today, while the related revenue can arrive over several years.

This delay does not mean that customer demand has disappeared.

It means that the market must wait.

The current price shows that many investors do not want to wait anymore.

Fundamental analysis

Mirion supplies important infrastructure for the nuclear industry.

Its products include radiation monitoring systems, reactor protection systems, neutron monitoring equipment, dosimetry products, safety systems, software, replacement parts and technical services.

In simple terms, Mirion helps nuclear facilities detect radiation, monitor reactor conditions and protect workers.

These products are not optional, nuclear operator cannot remove safety and monitoring equipment to reduce costs. The systems must meet strict technical and regulatory requirements.

This makes Mirion very sticky.

A power plant cannot replace a Mirion system in the same way that a normal company replaces a computer. A new system can require tests, documents, approvals, integration work and employee training.

A cheaper product can therefore create a much larger total cost.

Mirion already has approved systems, long customer relationships and a large installed base. Its products are present in more than 98% of nuclear plants worldwide.

This position is difficult to reproduce.

The latest financial results remain strong.

• Second-quarter revenue increased 19.7% to $266.8 million.

• Adjusted EBITDA increased 27.5% to $65.3 million,

• and the adjusted EBITDA margin reached 24.5%.

• Orders also continued to grow.

Orders increased 10% when Mirion excluded the Paragon and Certrec acquisitions. Total orders increased 40% when the company included these acquisitions.

• Backlog reached approximately $1.14 billion.

Management also maintained its full-year guidance.

• It expects total revenue growth of 22% to 24%, organic growth of 5% to 7%, and adjusted EBITDA of between $285 million and $300 million.

Management also expects adjusted free cash flow of between $155 million and $175 million.

These figures do not show a business with lower demand.

They show a business that must convert its demand into revenue and cash.

Mirion does not have an order problem, it has a conversion problem.

Subscribe now

The installed-base advantage

Most investors focus on new reactors (the narrative) and small modular reactors but the current nuclear fleet is already a large opportunity for Mirion.

Nuclear plants can operate for several decades. During this period, they require replacement detectors, monitoring systems, dosimetry equipment, software, maintenance and technical services.

Approximately 80% of Mirion’s nuclear power revenue already comes from operating reactors.

This means that the thesis does not depend only on future reactor construction.

The current fleet supports the base business and new reactors create additional growth.

The existing fleet is also becoming more valuable.

Governments want to extend the life of current reactors and they also want to increase reactor output and restart closed nuclear plants. Each of these projects can require new safety systems, monitoring equipment, replacement parts and regulatory work.

Mirion already has the customer relationships because it is already inside the plants.

The new nuclear opportunity

The next nuclear cycle includes large reactors, small modular reactors and microreactors. Mirion does not need to identify the reactor developer that will become the final winner. It can supply safety and monitoring systems to different reactor designs.

The company is already involved in advanced reactor projects.

Mirion and Paragon have supplied reactor protection systems, neutron monitoring systems, radiation monitoring equipment and other safety products for projects connected to the U.S. Department of Energy.

This is important because Mirion can benefit before a reactor starts to produce electricity. The company can receive orders during the design, construction, testing and approval stages. It can then supply replacement equipment and services during the operating life of the reactor. It can also supply products during decommissioning. That way, Mirion has exposure to the full nuclear life cycle.

Market sentiment

Market sentiment has moved from strong optimism to fear.

Your fear is my greed.

Investors previously viewed Mirion as an obvious winner from the nuclear expansion. AI data centers needed electricity, governments supported nuclear energy, and Mirion was present in almost every nuclear plant.

The market then started to ask a different question.

Where is the cash? And the answer is that it is arriving more slowly than investors expected.

The market now gives less value to the backlog because it does not know when the cash will arrive. This concern is reasonable. However, the market has moved from excessive optimism to excessive impatience.

The orders are still growing. The backlog remains high, and management has maintained its guidance. The nuclear need has not become weaker, the valuation has though.

Momentum

The business has better momentum than the share price.

Governments continue to support reactor restarts, life extensions, power increases and advanced nuclear projects. Electricity demand from data centers also continues to increase.

Mirion is positioned in both parts of the market.

The company benefits from the existing reactor fleet, which requires continuous maintenance and replacement equipment. It can also benefit from new reactor projects before these reactors become operational.

Orders increased during the first and second quarters.

Backlog reached $1.14 billion. Revenue, adjusted EBITDA and margins also increased.

The share-price momentum remains weak though.

The stock has fallen from above $30 to approximately $15. It can continue to fall if organic growth or cash conversion remains below market expectations. Weak share-price momentum does not prove that the operating business has become weaker. It shows that investors want faster proof. But it doesn’t work like that, which makes it a good entry point.

Risks

Organic revenue growth remains moderate for now. A large part of the reported growth comes from the Paragon and Certrec acquisitions. Mirion must prove that it can integrate these businesses. It must also show that acquisition growth can support future organic growth.

Cash conversion is another risk as the company generated $77.4 million of operating cash flow during the first half of 2026. It must produce a stronger second half to reach its full-year adjusted free cash flow guidance.

The balance sheet also requires attention because Mirion had approximately $419 million of cash at the end of the second quarter. It also had approximately $1.2 billion of term debt and convertible debt.

They will say ‘Backlog is not guaranteed revenue’ as customers can delay projects, change budgets or cancel orders. Nuclear projects can also face regulatory, political and construction delays. But everything is going in the opposite way: America needs energy, and it needs it ASAP.

The valuation is also not extremely low. At the current price, the market still gives Mirion credit for its nuclear position, now the company must deliver organic growth, margin improvement and stronger free cash flow.

Subscribe now

Why I am buying

I am buying because the main investment case remains intact. The world needs more electricity, and nuclear power will supply part of this electricity.

The existing reactor fleet needs maintenance, upgrades and replacement equipment. New reactors need detection, control and safety systems.

Mirion is present in more than 98% of nuclear power plants worldwide.

More than 80% of its nuclear power revenue comes from reactors that already operate. This gives the company an existing business before we include the future nuclear cycle. The company has real products, approved systems, long customer relationships, growing orders and a $1.14 billion backlog.

The problem is the speed of conversion. Orders are not becoming revenue and cash quickly enough for the market. But this is also the problem of the full nuclear industry. Approvals require time. Construction requires time. Reactor restarts, life extensions and new projects also require time.

The demand does not become false because the conversion is slow.

The market sold a timing problem as if it were a demand problem.

The world needs more power. Governments are supporting nuclear energy, utilities want to extend the life of their reactors, and advanced reactor projects are moving forward. Mirion is already inside the current fleet. It is also supplying the future fleet.

The research is complete because the fear is visible and the business continues to grow. Sometimes, you must simply buy.

Disclaimer

This article is for information and educational purposes only. It does not constitute financial, investment, legal or tax advice.

The author own shares of Mirion Technologies, Inc. ($MIR) and may buy or sell the security at any time without notice.

All investments involve risk, including the possible loss of capital. The information in this article is based on sources considered reliable at the time of publication, but its accuracy and completeness cannot be guaranteed.

Readers must conduct their own research and make their own investment decisions. When necessary, they should consult a qualified financial adviser.

I am buying Marvell at this price (163$)

Read this disclaimer

The business continues to grow, but the share price has fallen sharply. Marvell closed at $297.89 on June 30 and at $174 on July 28. This is a decline of about 41%.

On July 29, the shares traded near $170 during the day. This price was almost 49% below the June high of $329

This is more than a normal market correction, for me it’s a large urrational sell-off.

However, the latest company filings, financial guidance, product launches, and customer demand still support the investment case.

The market previously viewed Marvell as a popular AI stock. The market now views it as a strong company with weak share-price momentum.

This type of situation usually gets my attention.

Why now

The share price has fallen in several large moves.

• On July 1, the shares fell 8.7%.

• On July 7, the shares fell another 7.5%.

• On July 15, the shares fell to $206 after an analyst downgrade.

• On July 16, the shares closed at $188.

The market activity in July gives a clear message. Investors have reduced their exposure to Marvell and other semiconductor stocks.

The broader semiconductor sector also became weaker. Investors started to question the future growth of AI spending. They also became concerned about high valuations and crowded positions.

Reuters reported concerns about slower capital spending growth from hyperscaler cloud companies.

Marvell has been part of this wider market decline.

Subscribe now

Fundamental analysis

Marvell supplies important infrastructure for AI data centers.

Its products include:

• Custom processors

• Optical interconnect products

• Ethernet switches

• PCIe and CXL connectivity products

• Silicon photonics products

In simple terms, Marvell helps AI systems move large amounts of data at high speed.

This is important because expensive processors must receive data continuously. Slow data movement can reduce the performance of the full AI system.

Marvell’s data-center portfolio includes custom computing, interconnect, network switching, and storage products.

The company estimates that its total data-center market opportunity can reach $94 billion by 2028.

The recent financial results are strong.

• Fiscal 2026 revenue reached a record $8.195 billion.

• Data-center revenue was $6.1 billion. This represented 74% of total company revenue.

• In the first quarter of fiscal 2027, revenue reached $2.418 billion.

• GAAP gross margin was 52.1%.

• Non-GAAP gross margin was 58.9%.

• Operating cash flow reached a record $638.8 million.

• Management guided second-quarter revenue to approximately $2.7 billion at the midpoint.

Management now expects fiscal 2027 revenue to reach almost $11.5 billion.

It also expects fiscal 2028 revenue to reach $16.5 billion.

Market sentiment

Market sentiment has moved from strong optimism to fear.

Your fear is my greed.

In June, Reuters reported that the share price increased after Marvell joined the S&P 500.

The shares also increased after NVIDIA CEO Jensen Huang described Marvell as a possible future trillion-dollar company.

Retail investor activity also increased sharply.

Reports based on Stocktwits data showed that message volume increased by 435%. Sentiment reached the “extremely bullish” level.

That market mood has now disappeared.

Short interest is approximately 32.48 million shares. This represents 3.74% of the public float.

This level does not suggest a normal short-squeeze reversal.

The decline appears to come mainly from long-term investors and institutional holders who reduced risk after a period of excessive optimism.

Subscribe now

Momentum

The business has strong momentum, but the share price has weak momentum.

During the first-quarter earnings call, management reported strong demand and exceptional bookings across the data-center portfolio.

Management expects revenue to grow by at least 10% from one quarter to the next in the second, third, and fourth quarters.

Management also expects quarterly revenue to reach $3 billion in the third quarter.

Marvell recently launched the Teralynx T100.

The Teralynx T100 is a 102.4-terabit-per-second Ethernet switch for large AI computing clusters.

Customer sampling is expected to start during the current quarter.

Options trading is active around the $170 and $180 call options. Activity is also high around the $170, $185, and $200 put options.

This activity suggests that traders expect continued volatility.

It does not prove that the operating business has become weaker.

Risks

The investment has important risks.

• One direct customer represented 14% of fiscal 2026 revenue.

• One distributor represented 37% of revenue.

• The ten largest customers represented 82% of total revenue.

• Total debt was approximately $5.0 billion at the end of the first quarter.

• Company insiders also continued to sell shares during July.

However, the sales by the CEO and by Sandeep Bharathi were completed under pre-arranged Rule 10b5-1 trading plans.

Why I am buying

I am buying because the main investment case remains intact.

The share price now includes more fear and lower expectations.

Marvell has real products, strong demand, growing cash flow, and higher financial guidance.

The company also has a stronger relationship with NVIDIA through NVLink Fusion.

Marvell operates in an important part of the AI infrastructure market.

Wherever AI is going, Marvell benefits from Jevon’s Paradox.

GPUs perform the calculations. Marvell helps move the data between processors, memory, storage systems, and network equipment. In any case it’s needed.

This function becomes more important as AI clusters become larger.

Market sold too much.

The research is complete because the fear is visible and the business continues to grow.

Sometimes, you must simply buy.

Disclaimer

This article is for information and educational purposes only. It does not constitute financial, investment, legal, or tax advice.

The author may own shares of Marvell Technology, Inc. ($MRVL) and may buy or sell the security at any time without notice.

All investments involve risk, including the possible loss of capital. The information in this article is based on sources considered reliable at the time of publication, but its accuracy and completeness cannot be guaranteed.

Readers must conduct their own research and make their own investment decisions. When necessary, they should consult a qualified financial adviser.