What does Jupiter do?
The company is entirely in the railways business, or for the most part. They have started diversifying into non-railways businesses but they are an extremely small part of the overall scheme of things.

• Railway Wagons for carrying freight from one part of the company to another.

• Special components for tracks crossing each other

• Containers (some non railway parts)

• Wheels and Wheel Sets

• Axles

• Brake Discs

• EV components (non railway business)

• BESS

Railway wagons remain the single largest component of the business. At more than 75%. However, this business has declined over the last couple of years, reducing overall revenue.

Why has wagon business not grown?
This is simply because Jupiter (and all other wagon manufacturers) are at the mercy of the bureaucrats that comprise Indian Railways. They have been slower than tortoises in tendering wagons. As such, companies like Jupiter have not got enough business from them. And hence, the business has remained static.

Is tendering activity supposed to happen?
Yes. However, nobody knows when. Even if it happens, there are two possibilities. The stock will get a pop on the tender day and then stabilize. Or this possibility is already priced into the stock’s valuation and probably nothing will happen. The tendering activity has been so muted that now everyone expects wagon tenders to come out, hence ideally it should be already priced in.

Management has flagged that FY27 wagon volumes will remain muted, with recovery expected only in FY27-28 once IR tenders arrive.

And hence, the diversification from railway into non-railway businesses.

• Reducing dependence on Indian Railways (IR): The MD&A flags this as a key risk — “As Indian Railways (IR) is a major customer for wagons, any adverse impact on its budget allocation will affect the order flow. The Company has partially mitigated this risk by developing wagons for private operators.”

• Cyclicality mitigation: CV load bodies and private wagons reduce exposure to IR’s order cycle.

• Forward/backward integration: Products like brake systems (Stone India), wheelsets, and batteries improve control over cost, quality, and delivery.

• Higher-margin, value-accretive segments: Management notes the “wheel business and container business along with the crossing business would have the maximum margins.”

Wheelsets:
The wheels required for wagons are fully imported right now. Jupiter has made an acquisition to make these in-house for captive consumption and this will potentially boost margins in the future. However, this is also dependent on new wagon tenders.

Non-Railway businesses:
The non railway businesses are primarily on EV production and containers and both are at a nascent stage. However, the key point is these are margin dilutive currently.
Higher margins will start accruing as they scale which will only happen with time.

The Numbers:
With no growth numbers present, no tendering activity visible, ROE/ROCE of less than 10% and PE of more than 65, it is an expensive stock.

To management’s credit, they are trying their best to navigate this headwind and reduce dependence on Indian Railways. However, this will take at least 3 years to play out well.

Subscribe now

Disclaimer : This is not an investment recommendation. This post should be used for informational and educational purposes only and should not be construed as a call to action. Please consult a financial advisor before making any financial decisions. I am not a SEBI RIA.

So, how does it earn money?
Well, central and state governments incur capital expenditure on infrastructure. They make roads, dams, bridges, offices, etc and spend money on the same.

They then outsource these projects to third party contractors like NCC, L&T and others.

These companies in turn then buy equipment to build these projects.

One equipment is a concrete mixer as you can see above. And this, is what AJAX makes.

So, it follows that if central and state governments increase their capital expenditure, firms like AJAX will benefit.

As such, it becomes highly dependent on governments indirectly.

Wait, but is infrastructure not booming?
Yes and no. Central infrastructure spending is at 98% of budget. This is highly efficient.

But state spending is at only 52%. Why? Because of all “sisters” getting money everywhere in the country.

Schemes like Laadli Behen, Lakhpati Didi, etc have consumed a massive amount of money. And this money needs to flow from somewhere. That somewhere is state infrastructure spending.

Subscribe now

And because of all this free money being given to everyone in this country, major capex programmes are stalling. This impacts a company like AJAX.

Massively.

Because upstream contractors will just rely on the mixers they already have. Why will they buy new equipment?

The Numbers
But the numbers are pretty decent.
Revenue has doubled in the last 3 years to 2100 CR
Profits have gone up by around 70% as well (over 3 years) to around 300 CR

Working Capital cycle is stellar at less than 20 days.

Cash flow / EBITDA is at 140%!! This is brilliant.

However, markets price in the future.

The management managed to hike prices by a measly 2%. When your biggest achievement in the quarter is a 2% price hike and you keep on talking about it, there is definitely something wrong with the underlying business.

Management has also guided that infrastructure spending is lower than the budgeted estimate. (Sisters are happy).

Oh, and the cash - they have around 1100 Crores of cash lying on the books. No major plans for acquisition (what will you anyway acquire in construction equipment sector).

No major plans for expansion (the industry has slowed down)

Shouldn’t you give some of that money back?

Summary
At a valuation of 27x, zero revenue growth visibility, massive government dependency, I feel it is too heavily valued at the present moment.

That too for an industry that is cyclical and non-innovative (cement mixers, c’mon).

Share

Disclaimer : This is not an investment recommendation. This post should be used for informational and educational purposes only and should not be construed as a call to action. Please consult a financial advisor before making any financial decisions. I am not a SEBI RIA.

What actually is KPIT?
KPIT is in tech, but it is not a traditional IT company. It services the auto sector, but it is not an auto or auto-ancillary company.

KPIT specializes in building core embedded software, AI, and digital platforms that help car and truck makers transition toward autonomous, clean, and connected vehicles.

In short, it creates software that helps vehicles become smarter.

Core Focus Areas:
Autonomous Driving & ADAS: Building software for self-driving features and advanced driver-assistance systems.

• Electric Vehicles (EVs): Developing powertrain and battery management systems for clean mobility.

• Connected Vehicles: Creating cloud and digital infrastructure to link vehicles with smart ecosystems.

• Vehicle Architecture: Designing middleware, electronic/electrical (E/E) architectures, and vehicle operating systems.

Who are its customers?
All major auto-makers. Honda. GM. Jaguar. Stellantis. BMW.
KPIT helps them create software and features that makes the car more user-friendly.
KPIT firmly believes in the future of autonomous mobility and hence, it entirely focuses on the same.

Where does it earn money from?
This is a very important question. Most of its revenue comes from the West. US and Europe. This is typical of any tech or engineering company. It has recently started increasing its revenue from Japan, Korea and China.

However, India is a minuscule part of the total revenue and it is safe to say that KPIT has not been very successful in India.

What has been the historical growth?
The below snapshot is from the PPT they released in May 2026. This was a constant feature in all their PPT’s. However, in July, they quietly removed this slide from their presentation. This is an extremely telling point that KPIT is facing major headwinds.

As you can see, up until end of last fiscal year, KPIT was showing decent growth.

So, what’s changed?
KPIT's growth trajectory took a sharp turn in Q1 FY27 (April-June 2026) driven by a cascading crisis across European automakers – which form the backbone of KPIT's revenue base (~50%+ from Europe).

Loss of China market share > European OEMs used to make their highest margins in China. That market share has been decimated by Chinese EV makers.

Chinese competition entering Europe > Chinese OEMs are now winning share in the European OEMs' own home turf.

Restricted US market > Tariffs and geopolitical uncertainties have made the US market difficult for European players.

So, if your largest market is facing so many uncertainties globally, especially with Chinese EV dumping across Europe.

THIS IS A MAJOR CRISIS and it has a follow down effect on KPIT’s performance.

So, what are European automakers saying?
Multiple profit warnings by European automakers
Massive job cuts and ongoing restructuring
Pay cuts and longer work hours
Big write-offs — including one Japanese OEM (Honda) that took a $15 billion one-time hit and cancelled all new platform programs
R&D budget tightening across the board

KPIT's Q1 FY27 revenue was impacted by two major program losses:
European OEM SDV program > Planned ramp-down (reached start of production, naturally ending)
Japanese OEM program (Honda) > Abrupt cancellation — the OEM scrapped all new EV platform programs

Because of this, KPIT is significantly impacted.


Will it bounce back?
In the near term of 2-3 years, looks unlikely. The downturn is European car makers is more structural, not cyclical.

Before (FY23-FY25): OEMs were investing heavily in SDV programs, new architectures, and next-gen software features.

Share

Now (FY26-FY27): Conversations are about cost reduction — how KPIT can help OEMs reduce their product cost by 30-40% and production cost by similar margins.

“The conversation has shifted towards how does KPIT help these OEMs reduce the cost of their product by at least 30% to 40%. And secondly, how do we help them reduce the cost of their production. If they are able to bring these price points down by 30% to 40%, they become competitive again.” — Sachin Tikekar

What does this mean?
That KPIT, for revenue growth, will have to start compromising on its margins. This will further drag down growth.

This opinion is backed by numbers. EBITDA margin was 21% in Q1FY26 and one year later it is 17.2%. I expect this to go down further.

If your biggest client base is facing structural headwinds, you will suffer as well.

What are the numbers saying?
Between 2021 and 2024, the stock went up almost 6 times. I remember seeing 3-4% daily up moves. However, since then, the stock has lost nearly 60%.

This was because, KPIT was showing stellar growth between those three years and now growth is limited. More likely, there will be de-growth.

Revenue increased from 2036 CR in FY 2021 to 4872 CR in FY 2024. Growth in INR terms for 40% in FY23 and FY24. Similarly, profits grew from 147 CR to 600 CR by FY 2024.

Even now for FY 2026, revenue has increased but the pace has drastically dropped from 40% to 10%. And, profits have actually declined due to margin pressures.

The FY2026 PAT decline of 24% is not purely operational — it includes a ₹597 Cr exceptional charge from the new Labour Codes (notified Nov 2025). Stripping that out, FY2026 profit would have been closer to flat/slightly down rather than sharply down. The real operational deterioration is visible in Q1 FY27, where the European OEM actions caused the margin to drop to 17.2% — the lowest level since FY2021.

FY2020-FY2024 = hyper-growth powered by European SDV programs (₹2,156 Cr → ₹4,872 Cr, ~4.5x PAT growth). FY2025 onwards = the same European OEMs that fueled the growth hit a structural crisis (Chinese competition, tariffs, profit warnings), causing growth to collapse from +45% to flat, margins to compress from 21% to 17.2%, and PAT to decline for the first time.

This is evident in multiple compression. At one time, KPIT used to command a PE of 90. Now it is less than 30.

Summary
Headwinds, margin compression, clients losing business all have combined, and will continue to combine, to pressure KPIT. There is ZERO visibility on any turnarounds.

Are the products good? Absolutely
Is the company good? Of course.
Is the management good? Yes
But, is the business good? No
Is the stock good? Not for now.

Disclaimer : This is not an investment recommendation. This post should be used for informational and educational purposes only and should not be construed as a call to action. Please consult a financial advisor before making any financial decisions. I am not a SEBI RIA.

Subscribe now

What does the company do?
Gravita India recycles lead, copper and plastic. It has a deep-rooted procurement network which it utilizes to procure scrap metal at various locations across the world and then makes value-added products to re-sell them.

Where are its plants?
Most of them are in India and Africa, which is the right thing to do as cost of operations becomes low. However, they source scrap from almost continents in the world.

With this, they serve around 400 customers globally across 37 countries at a 50% utilization rate (which is low tbh).

However, the share of value added products, which command higher realizations is at 63%. This is pretty impressive.

The numbers:
The company has scored decent enough revenue growth over the last few years and good profit growth as well. However, the guidance is what matters.

They are guiding for sustained ROIC at 25%. Currently they are already near that.
Profit growth guided is 30-35% on sustained basis.

But current PE is near this level already. So all these numbers are already baked into the price.

Plus this is a commodity business. Which should inherently command a much lower multiple than the current 30x levels. While there doesn’t seem to be massive downside, the upside has been baked into it and there doesn’t seem to be a case for any explosive growth here.

Disclaimer : This is not an investment recommendation. This post should be used for informational and educational purposes only and should not be construed as a call to action. Please consult a financial advisor before making any financial decisions. I am not a SEBI RIA.

So, what does Affle do?
Put it simply, it targets you. It targets your wallet. It targets your spend. It targets your psychology.

Using AI and ML, Affle has proprietary technology (with multiple patents) that provides ads whenever a user is browsing something. If you want golf sticks, you will get an ad. If you want to go to Bali, you will get an ad. If you want to play Candy Crush, you will get an ad.

It is a global AI & ML-driven mobile advertising technology company. It enables brands to acquire, engage, and convert users through personalized, measurable mobile advertising — with a unique "pay only for results" model.

How does it earn money?
It has a unique outcome-based revenue model. Advertisers pay only when a real user completes a desired action (install, purchase, registration, etc.). This is called the CPCU (Cost Per Converted User) model. Almost 100% of its revenue comes from this.

How is CPCU calculated?
Simple.

What is the conversion in volume? Get the number of actual transactions (install, purchase, registration, etc).

What is the average rate charged in CPCU per transaction?

Multiply the two.

You get the company’s revenue.

So, as you can see, if the company is able to have more conversions or increase the CPCU rate, its revenue can skyrocket.

But can it increase both or either of the two?

Both levers have moved in the right direction every single year.

Lever 1 - Conversions
Affle’s majority business is still from India and EM. Only about 30% comes from developed markets. There is significant room for expansion. Plus, with e-commerce skyrocketing, gaming going up and new Gen AI apps becoming potential advertisers, Affle is uniquely positions to increase conversions.

Lever 2 - CPCU rate
Affle is positioning itself as an AI/ML player that targets serious and high-intent consumers. It targets premium positioning with iOS devices and deeply integrates with the advertiser’s data, justifying higher CPCU.

This dual-lever model is precisely why CPCU revenue has compounded at a strong rate — and management expects both levers to keep contributing.

The numbers
Profit and revenue have grown by 25% CAGR over 5 years.
CFO/Ebitda is consistently above 80% implying high cash conversion.

However, valuation is at 47x PE.

Guidance is 25% profit growth, which implies a PEG near 2x.

This is not an undervalued stock. However, it will continue to command premium multiples as there are very few other peers.

Affle trades at a premium valuation (P/E ~48x, EV/EBITDA ~35x) reflecting its high-quality metrics: asset-light model, near-debt-free balance sheet, strong cash conversion, consistent margin expansion, and a credible 20%+ organic growth runway with optionality from M&A.

The stock price is down by about 20% since 1 year and up by only 60% over 3 years. Watch out for the conversions. If they increase, Affle may jump.

At the very least, it is a steady compounder and worth having on your watchlist.

Disclaimer : This is not an investment recommendation. This post should be used for informational and educational purposes only and should not be construed as a call to action. Please consult a financial advisor before making any financial decisions. I am not a SEBI RIA.

Why has Gold Fallen?
This is essentially forced liquidity. Gold is a safe haven asset. When you have issues in your life or need hard cash, you sell your assets. The Iran war bared how countries started scrambling for oil, and money. They had to sell gold.

However, in this I tried to understand if the long term thesis remains intact or not. I found 7 bullish signals, one of them no less from the US Fed.

Signal 1 - China
China has been importing gold left, right and centre. Look at the chart below.
China has not only been buying gold, but increasing the tonnage.

Every. Single. Month. For the past 5 months.

Chinese customs data revealed that China imported a staggering 163 tons of gold in April. This represents a 10% month over month increase and a 40% year over year surge.

China is not backing down.

Signal 2 - Singapore
Singapore Exchange (SGX) is launching a brand-new Over-The-Counter (OTC) gold clearing system.

Instead of relying on Western infrastructure or fragmented regional trades, SGX will centralize the clearing and settlement of over-the-counter physical gold trades.

Six banks have already signed the MoU : JPM, DB, ICBC, DBS, OCBC, UOB.

Singapore will also gold vault services for foreign central banks and sovereigns. Being neutral, Singapore wants to be perceived as a safe haven for physical storage.

It is also removing 5% tax on bullion. This allows family offices and mega-funds to aggressively load up on physical gold without tax friction.

Signal 3 - CIPS
China's Cross-Border Interbank Payment System (CIPS) and gold are twin pillars of Beijing’s long-term geopolitical strategy.

The biggest functional overlap between CIPS and gold occurs in commodity trade—specifically with heavily sanctioned or de-dollarizing trade partners like Russia, Iran, and parts of Africa.

Foreign partners are often hesitant to hold large reserves of Chinese Yuan because the currency is not fully convertible.

They can sell oil or minerals to China via CIPS, receive Yuan, and exit the currency into a globally recognized physical asset (gold) without ever touching a U.S. dollar, Western clearing houses, or the SWIFT network.

Signal 4 - The US Fed Anomaly
The US Fed recently announced a one-ounce gold coin to be sold at $19600.

Read that again. It is not a typo. Current gold price is $4100 per ounce.

This is almost a 5x premium.

Yes, the coin is limited. Yes, it celebrates something. Yes, it has a numismatic value.

But, 5x?

Is this a signal for an impending revaluation?

Signal 5 - The US Gold Bonds
Judy Shelton is a prominent conservative monetary economist known for her advocacy of the gold standard, and her specific proposal for gold-backed U.S. Treasury bonds.

Because she has frequently been considered for high-level monetary roles—including past nominations to the Federal Reserve Board—her policy proposals are treated as a blueprint for how a future administration could theoretically reform the financial system.

Shelton proposes that the U.S. Treasury issue a new type of zero-coupon instrument, which she has dubbed "Treasury Trust Bonds" or the "Solidus".

The bonds would be directly collateralized by the nation's massive official gold reserves.

Signal 6 - The US Balance Sheet
A revaluation of the Treasury’s gold to $20,000 or even half would unlock trillions in new balance sheet equity plus provide solid collateral for new bonds, plus also help with the massive debt overhang US faces.

Signal 7 - The Supply
Global miners hit an initial record production of 3,672 tonnes in a recent full-year stretch, and production has continued to inch up. However, mining experts note that output is growing very slowly because easily accessible deposits have largely been tapped, making new projects more difficult and expensive to launch.

While I still feel gold might fall a little bit from here on, the long term thesis seems to be upwards; based on these breadcrumbs.

Disclaimer : This is not an investment recommendation. This post should be used for informational and educational purposes only and should not be construed as a call to action. Please consult a financial advisor before making any financial decisions. I am not a RIA.

What does the company do?
At its core, Nebius is an AI powerhouse. They build massive server farms filled with thousands of the world's most powerful AI computer chips (GPUs) and rent out this computing power to businesses, developers, and researchers so they can build, train, and run advanced AI apps.

So, Nebius buys chips from Nvidia » Creates server farms » Rents out compute.

Their Moat: What Makes Them Special?

1. AI-Native Engineering: Traditional cloud giants were built to host websites and corporate apps. Nebius was built completely from the ground up strictly for AI. This means their software and hardware networks (InfiniBand) connect the chips much faster, allowing AI models to learn and "think" without lag.

2. Cheaper Prices & Better Hardware Access: They are a preferred partner of Nvidia (who makes the world's best AI chips), giving Nebius incredibly early access to the newest models. Because they design their own servers and cooling systems (instead of relying on third parties), they offer significant cost savings.

The Pivot : Cursor + SpacEx + Nebius
The convergence of Nebius, Cursor, and SpaceX revolves around a major tech milestone: SpaceX’s massive $60 billion acquisition of Cursor (Anysphere).

While Cursor will now be backed by SpaceX’s powerhouse infrastructure (including the xAI Colossus supercomputer), Cursor remains a confirmed customer of ClickHouse Cloud, an enterprise database giant where Nebius holds a commanding 28% equity stake.

This interconnected relationship unlocks two massive ecosystems and strategic advantages for Nebius:

1. Massive Financial Upside: As Cursor scales globally under the SpaceX banner, its data and infrastructure needs will skyrocket. This directly drives Annual Recurring Revenue (ARR) growth for ClickHouse, significantly boosting the value of Nebius’s multi-billion dollar equity stake.

2. SpaceX intends to deploy Cursor's agentic AI tools across its entire engineering stack—meaning every rocket, Starlink satellite upgrade, and lunar exploration software patch will be built using AI-assisted coding. This opens up a highly lucrative Physical AI, Simulation, and Robotics ecosystem for Nebius.

The Numbers!
FY25 revenue increased from $92M previous year to $530M.

This is expected to go 6X at at $3B in 2026.

Gross margins highly lucrative at 74%

Price to Sales is elevated at around 20X but compresses for forward P/S.

FCF is negative.

Capex is high

The biggest risk is a funding gap. Nebius generated a record $2.3 billion in Operating Cash Flow in Q1 2026, largely driven by upfront customer prepayments. They also possess a $9.3 billion cash war chest built from a $4.3 billion convertible debt round and NVIDIA's $2 billion equity injection.

Total cash plus full-year projected EBITDA (~$1.2 billion to $1.3 billion) leaves an estimated $6 billion to $8 billion funding deficit to reach the top end of their CapEx target of $25B.

Nebius must close this gap through asset-backed debt or by tapping its At-The-Market (ATM) stock offering. Relying heavily on the ATM facility to raise billions in cash will result in shareholder dilution, putting downward pressure on the stock price despite exploding revenues.

Very interesting company to watch out for, but I believe this space is starting to get crowded with Nebius, Boost Run, Coreweave, Myscale, Northern Data, Applied Digital and of course, the hyperscalers.

Disclaimer : This is not an investment recommendation. This post should be used for informational and educational purposes only and should not be construed as a call to action. Please consult a financial advisor before making any financial decisions. I am not a RIA.

What does Boost Run do?
In layman's terms, Boost Run is an AI cloud service provider. They buy massive amounts of specialized AI computer chips (GPUs), wire them together, and rent them out to companies that cannot buy or manage the hardware themselves.

So essentially, companies can choose to rent the GPU’s and racks and servers and data centers from Boost Run rather than buying those GPU’s and racks and servers from providers like NVIDIA.

But wait. Why wouldn’t they directly buy the equipment from NVIDIA and bypass Boost Run?

1. Instant Access : Buying AI hardware is incredibly expensive and has long waitlists. Customers partner with Boost Run to get instant access to thousands of GPUs via a flexible rental model.

2. Infrastructure Headaches : Data centers require an immense amount of electricity, complex cooling systems, and highly specialized networking. Boost Run takes care of all the physical operations so the customer's engineers can focus purely on coding.

3. Massive Capital Expense (CapEx): Buying an individual enterprise server cluster can cost millions of dollars upfront. Boost Run absorbs that cost, allowing customers to use an Operational Expense (OpEx) model—paying a predictable subscription or hourly fee.

4. NVIDIA sells components, not rentals : If a customer goes to NVIDIA, they are buying physical silicon chips, circuit boards, and server racks. NVIDIA does not rent out time on chips by the hour; they sell hardware that you must plug into a physical building.

What Being an NVIDIA "Preferred Partner" Means
Being an NVIDIA Preferred Cloud Service Provider (CSP) means Boost Run has been vetted and granted VIP status within the NVIDIA ecosystem. It provides significant operational advantages

1. Guaranteed Supply Allocation: NVIDIA prioritizes its preferred partners during hardware rollouts. Boost Run gets earlier, more reliable access to scarce, next-generation architectures (like the NVIDIA Blackwell platform) than the general market.

2. Exemplar Cloud Certification: Boost Run achieved NVIDIA's elite Exemplar Cloud status. This technical milestone certifies that Boost Run's infrastructure operates within 5% of NVIDIA's strict engineering reference targets for real-world AI processing. It puts Boost Run into an exclusive technical tier alongside giants like Microsoft Azure, Oracle Cloud, and CoreWeave.

3. Direct Engineering Support: When Boost Run encounters complex architecture challenges, NVIDIA’s internal engineers help optimize the infrastructure setup, ensuring maximum performance for the end-user.

Key Question : how does the money move? How does the business actually work?

Step 1: Securing the Allocation First (The NVIDIA Relationship)
Boost Run cannot sell what it doesn’t have. Because Boost Run is an NVIDIA Preferred Partner with Exemplar status, NVIDIA gives them an upfront, guaranteed allotment of upcoming chips (like the new B300 and Blackwell GPUs). Boost Run places massive wholesale orders with hardware integrators long before individual customers sign up.

Step 2: The Financing Engine (The Dell & DFS Partnership)
To pay for these multi-billion-dollar chip orders without draining their own bank account, Boost Run utilizes Vendor Financing.

• Boost Run signed a massive $1.44 billion hardware purchase agreement with Dell Technologies.

• Dell acts as the middleman builder: they take the NVIDIA chips, build them into server racks, and supply them to Boost Run.

• To fund this, Boost Run uses Dell Financial Services (DFS). Dell essentially lends Boost Run the money (via credit lines and lease structures) to build out the data centers

Step 3: Landing the Customer & Closing the Loop

Once the capacity is secured or deployment timelines are locked in with Dell, Boost Run signs enterprise clients to massive, long-term contracts. A perfect example of this modality is Boost Run’s $471.7 million, 36-month contract with Thinking Machines Lab to deploy 5,000 NVIDIA B300 GPUs.

The critical operational terms of these customer contracts include:

• Non-Cancelable Commitments: The customer signs a binding agreement for 2 to 3 years.

• Take-or-Pay Structure: The customer must pay the fixed monthly fee regardless of whether they actually use the GPUs that month.

• No Upfront Cash Needed for Hardware: The customer doesn’t hand over $470 million on Day 1. Instead, they pay a predictable monthly or quarterly rental fee.

Step 4: The Cash Flow

Once the customer signs the contract, that contract becomes a highly valuable financial asset called Contracted Revenue Backlog (which sits at roughly $940 million for Boost Run).

Boost Run takes these ironclad customer contracts back to their lenders (like Dell Financial Services) to prove they have guaranteed income. This allows them to unlock even more credit, order more next-generation chips, and continue scaling up the flywheel.

The Numbers!
Revenue is tinier then your pinky toe.

Only $30M in 2025.

But Market Cap is upwards of $2B already.

Which means Price to Sales is a massive 75X!

Does it justify this premium?

Boost Run is expecting ARR to be $400M in 2026.

That is a 13X jump in revenue!

Which means on a forward price to sales basis, the multiple compresses to 5X only!

And contracted revenue backlog is expected to go to $1.4B by year end.

FCF margins expected to be 15% at least.

So, not wildly expensive on forward basis. And also will generate cash.

Risks
Of course there are risks.
This remains a highly speculative small cap company. There might come in another company that does this better and cheaper. Or simply, Boost Run might not find enough customers at all.

It doesn’t have quite a lot of marquee names on it customer list that provide anchor.

Boost Run's auditors have previously flagged a "going concern" warning, emphasizing the company's dependency on continuous outside financing to survive.

Disclaimer : This is not an investment recommendation. This post should be used for informational and educational purposes only and should not be construed as a call to action. Please consult a financial advisor before making any financial decisions. I am not a RIA.

First what is a Transformer?
An electrical transformer is a static piece of equipment that transfers electrical energy. Period. Transformers come in various capacities - 11KV, 33KV, 220KV, 400KV, etc.

What does Shilchar do?
Well, Shilchar manufactures these transformers for domestic and export consumption. But they manufacture only up to 132KV, not more. Manufacturing upto 132KV means they are a medium power supplier.

What are they doing next?
However, they have a firm expansion plan to manufacture and scale up to 220KV.

They are expanding almost double from 7500 MVA to 14000 MVA. (Technical terms - no need to get into details). Key word is doubling.

This expansion plan is supposed to complete by August 2027. In two months. A couple of quarters later, the new plant will have stabilized and will start contributing sizeable revenue.

What does expansion into 220KV mean?
Moving from 132 kV to 220 kV shifts Shilchar from being primarily a distribution and medium-power supplier into a high-voltage utility transmission supplier. It allows them to bid on much larger government, state electricity board (SEB), and massive global infrastructure tenders that they were previously unqualified to target.

Why not go higher to 400KV plus?
India’s grid utilizes ultra-high-voltage networks like 400 kV and 765 kV to transmit massive power across the country.

Sounds like a nice opportunity. But Shilchar specifically, at the moment, wants to focus on 220KV. Why?

Because, India is focusing mostly on renewable energy. Most massive solar and wind farms in India inject power into the regional grid at the 33 kV, 132 kV, or 220 kV levels. Building 400 kV or 765 kV equipment shifts them into heavy state-utility grids, facing stiff competition from massive legacy players like BHEL.

This means that Shilchar has carved out a highly profitable niche in renewable energy evacuation (solar and wind power plants).

Subscribe now

How are they funding this expansion?
Internal accruals only. 120 CR.

Zero debt.

Zero Equity Dilution,

Where are they expanding?
They have an existing 17 acre land parcel. They are expanding right next to their current plant ON THE EXISTING LAND PARCEL.

This means land cost is zero. Debt and equity raise is zero.

And this is not it. Even after the current expansion, they will have utilized only 40% of their land.

60% still remaining for expansion.
Future optionality to start more expansions.
And you can see the amount is negligible for expansions. Almost free.



What are the numbers?
Revenue has grown by 38% CAGR over 4 years.
180 CR —> 652 CR in FY 26

PAT has grown by 83% CAGR over 4 years
14 CR —> 158 CR in FY 26

This means there has been operating leverage.

EBITDA Margins increased from 11% to 29% in FY 26.

CFO increased from -6 CR to 192 CR in FY 26.

Why did the ROE and ROCE reduce?
ROE at 32% (Reduced from 42%)
—> ROE = PAT ÷ Shareholders' Funds
The equity base jumped by ~₹144 Cr in a single year — far outpacing profits.
ROE optically appears low.

ROCE at 42% (Reduced from 56%)
—> ROCE = EBIT ÷ Capital Employed
The company generated a massive ₹192 Cr of operating cash flow in FY26, but a significant portion remained as idle cash rather than being deployed into high-return operations. This cash earns modest interest income (not operating profit), dragging down the ROCE denominator without a corresponding numerator boost.

This is a temporary, transitional phenomenon typical of a company in an aggressive capacity build-out phase. The ROE/ROCE compression is not due to operational deterioration.

How has the stock performed?
Over the last 5 years, it has given a massive 72X return.
But it has tested patience in the last 1 year definitely.

PE at 30X. Forward guidance 20-25%. PEG is undervalued at 1.5X.

However, with the massive tailwinds in electrification, further bump up in margins, higher cash coming in, increase in margins and ROE, another inflection seems to be imminent in the business.

The key risk is lack of demand and competition heating up. But the electrification pie, globally, is very large.

Subscribe now

Disclaimer : This is not an investment recommendation. This post should be used for informational and educational purposes only and should not be construed as a call to action. Please consult a financial advisor before making any financial decisions. I am not a RIA.

What does Kilburn do?
Without going into technical details, it designs and manufactures products and provides solutions that go into factories.

These products are drying products. Think about how a clothes dryer, a hair dryer, or a food dehydrator works at home—they use heat and air to turn something wet into something dry. Kilburn does exactly that, but instead of clothes, their machines dry tons of raw chemicals, sugar, tea leaves, coal, and even municipal sewage sludge every single hour

They design specific equipment according to CUSTOMIZATIONS, rather than sell something off the shelf from an assembly line.

This becomes their IP.

The ME ENERGY Acquisition
This company was acquired in 2024 and the benefits are visible now. The payment was only 100 CR with around 75% paid in cash.
1. Why > ME Energy specialized in thermal engineering & waste heat recovery (WHR) systems — a perfect complement to Kilburn's drying systems. Both deal with heat/thermal processes, but from different angles.

2. TAM > Tap into each other's client bases. Kilburn had ~40% common customers with ME Energy; the other 60% of each company's customer base represented new addressable markets.

3. Operating leverage > ME Energy's Pune factory could serve as a second manufacturing facility for Kilburn. Product swapping between factories based on equipment type (heavier equipment at Kilburn's Kalyan plant, bulkier but lighter at Pune).

4. New Sectors > ME Energy brought access to steel, cement, food & beverage, and incineration/WtE sectors — entirely new verticals for Kilburn.

5. The current benefits > ME Energy has grown from 45 Cr pre acquisition to 100 CR Revenue today. Order book has gone 3X. ME Energy didn’t have the balance sheet, only capability. Kilburn brought in the balance sheet.

"Before we took over the company, they had working capital issues... they did not have the credibility or financial backing for customers to give large orders to them. With Kilburn acquiring ME Energy, two things have happened: the credibility of them executing large value projects is ascertained, and because of our network, we are able to open doors." - Management, Nov 2025.

The Monga Strayfield Acquisition
Paid 123 Cr, 100 Cr in Cash, acquired in Jan 2025.
We have established that Kilburn is into drying products. Monga did exactly that, just a different kind of drying.

1. Why? > Monga Strayfield is a pioneer and sole manufacturer of industrial Radio Frequency (RF) heating and drying. This is an entirely different drying technology from Kilburn's conventional thermal drying.

2. TAM > Monga had a well-established customer base in the USA and Europe. This gave Kilburn immediate access to developed markets.

3. Margins > Monga had strong operating margins (~24-25%) and was sitting on ₹33 crore of cash on its balance sheet at acquisition — making it an immediately accretive acquisition.

4. The current benefits > The benefits are less than ME Energy, but they are there. Monga becomes a steady contributor without margin dilution and ready cash on the BS.

The consolidation of both these entities has played out. What is the management guiding?

The Numbers!
First, the Capex » Combined capex expected over next 2 quarters at 40 CR, expected peak revenue potential 200 CR. 5X.

Second, the revenue » Increased by 50% to 629 CR in FY 26. Expected to increase by 20-25% in FY 27.

Third, the margins » This is a disappointment. Margin is expected to reduce from around 25% to 20%.

But, why? The Iran war is affecting everybody, right? But it impacts Kilburn differently. While most companies have commodity cost increase or decrease pass through clauses in their contracts, Kilburn does not. It will have to take the impact on its books. And that is why, the margins reduce. This is a risk and has a near term impact.

Fourth, the setup» The current PE ratio is 27. Assuming 25% revenue growth and 15% profit growth, it seems at a PEG of 1.8, it is fairly valued.

The company can compound from here.

I do not see issues in Debt.

I do no see issues in revenue.

I do not see issues in demand.

I see issues in margins.

I would still want to wait for management commentary post Q1 and Q2 results for more clarity on the EBITDA margin, which has the biggest impact currently.

As Amritanshu Khaitan noted in the June 2026 concall: "Economies of scale has already been achieved by the company... we are stable at 25%-odd EBITDA margin. It's not that from 25%, we can jump to 30% or 35%... we are also building the company for the future, with investments being made in people, talent, and infrastructure to take this company to a ₹1,000 crore level in 2 years' time."

Subscribe now

Disclaimer : This is not an investment recommendation. This post should be used for informational and educational purposes only and should not be construed as a call to action. Please consult a financial advisor before making any financial decisions. I am not a SEBI RIA.

My final assessment is at the bottom of the article. There are so many things are not even mentioned. I have laid them in the form of questionnaires.

Every thing that is missed. Every thing that should have been there. Right at the end. Don’t skip anything.

Point 1.
The Islamic Republic of Iran and the United States, together with their allies in the current war, declare upon the signing of this Memorandum of Understanding an immediate and permanent end to the war on all fronts, including Lebanon, and undertake that from now on they will not launch any hostile action against each other, and will refrain from the threat or use of force against each other. The final agreement will confirm the provisions of this Article and the remaining Articles.

My Assessment : Hezbollah is not specifically named. Well, you cannot name a front that you have declared as terrorists anyway. Israel is not named (reference to “allies”), It calls for an “end” to the war. It does not mean Israel will stop occupying Southern Lebanon. In fact, Israeli officials have gone on record to say they will not stop occupying that area. What stops either Israel or Hezbollah from dragging Lebanon into a wider conflict, since they are both unnamed?

Point 2
The Islamic Republic of Iran and the United States undertake to respect each other’s sovereignty and territorial integrity, and to refrain from interfering in each other’s internal affairs.

My Assessment : The war was started in anticipation of the local Iranian population rising against the regime. That didn’t happen. US didn’t come to their rescue. It means that if the local population continues to face extreme measures, US will not intervene. The local population loses.

Point 3
The Islamic Republic of Iran and the United States undertake to negotiate and reach a final agreement within a maximum period of 60 days, extendable by mutual consent.

My assessment : Both countries have already realized that a 60 day term might not be enough. Hence, it is “extendable”. By “mutual consent”. In fact, President Trump has already stated that he might just bomb Iran if they don’t act in good faith. How do you define good faith. The brevity of this point with no mention of what constitutes “final agreement”, is a minefield.

Point 4
Immediately upon the signing of this Memorandum of Understanding, the United States lift the naval blockade and prevent any interference or obstruction against the Islamic Republic of Iran, and restore traffic within a maximum of 30 days to its full capacity; the traffic of ships shall be proportional to the pre-war volume of traffic on the part of the Islamic Republic of Iran. The United States also undertakes to withdraw its forces from the surrounding areas within 30 days after the final agreement.

My assessment : This is already visible. Iranian ships have reportedly been crossing the US blockade already. It is working. The definition of “surrounding areas” is vague. Does it mean near the ports? In the Strait? US Bases in the Gulf? What defines withdrawal? Iran may well claim no withdrawal has happened. Again, open to interpretation.

Subscribe now

Point 5
Upon signing this Memorandum of Understanding, the Islamic Republic of Iran will immediately take steps to ensure that the movement of merchant ships from the Persian Gulf to the Sea of Oman and vice versa is resumed within 30 days to the pre-war volume, taking into account the need for the removal of technical obstacles and the neutralization of mines by Iran.

My assessment : What if the Strait cannot be de-mined within 30 days? What if Iran doesn’t “take steps” to remove mines? There is no definition of which specific steps to be taken and this seems to be more of good faith.

Point 6
The United States undertakes, together with its regional partners, to create a comprehensive plan agreed upon by both parties for the rehabilitation and economic development of the Islamic Republic of Iran, while ensuring financing of at least $300 billion. The implementation mechanism of this plan, as part of the final agreement, will be formulated within 60 days.

My assessment : Are you serious? What defines regional partners? Common sense dictates that if your neighbours attack you, you will NOT pay them money. Do we expect regional partners to pay $300 billion to Iran? There has be nothing on public record from any regional partner on this agreement. The wording is vague as well. The US AND Regional Partners will “create a plan”. Not necessarily pay. There is a high chance of this being nixed.

Point 7
The United States commits to ending, on a schedule to be agreed upon as part of the final agreement, all types of sanctions currently facing the Islamic Republic of Iran, including resolutions of the United Nations Security Council and the Board of Governors of the International Atomic Energy Agency (IAEA), and all unilateral U.S. sanctions, both primary and secondary.

My assessment : This is highly beneficial for Iran. Since all sanctions go away, Iran is free to sell its oil immediately. Something that was not possible before the war. They will get payments immediately. Something that was not possible before the war. This is the biggest win for Iran.

Point 8
The Islamic Republic of Iran reiterates that it will never produce nuclear weapons. The Islamic Republic of Iran and the United States have agreed that the fate of enriched material and the fate of all other mutually agreed nuclear-related issues, including Iran’s nuclear needs, will be adequately addressed in a final agreement; the final agreement will confirm the provisions of this Article.

My assessment: Iran will not produce nuclear “weapons”. What about nuclear “energy” for civilian purposes? Iran already was claiming pre-war that its nuclear programme is for civilian purposes. What is stopping them from starting it again?
It HAS to be addressed in the final agreement, as they say. This is of supreme importance. Otherwise, the US is back to square one.

Point 9
The Islamic Republic of Iran and the United States agree that, pending a final agreement, they will maintain the status quo: Iran will maintain the status quo on its nuclear program, and the United States will not impose new sanctions on Iran or strengthen its forces in the region.

My assessment : This directly contradicts Point 3 where they have to strive for an agreement within 60 days, mutually extendable. Here, the interpretation becomes “indefinite”, meaning forget about it. The key win for Iran here is that US will “not strengthen its forces in the region”. It will continue to supply weapons, but CENTCOM may not be able to increase troops. GCC will have to diversity its defense shield from other partners, including, but not limited to, Ukraine.

Subscribe now

Point 10
The United States undertakes that immediately after the signing of this Memorandum of Understanding, and until the date of the lifting of sanctions, the United States Treasury Department will issue waivers for exports of Iranian crude oil, petrochemical products and their derivatives, and all related services, including banking, insurance, transportation, and the like.

My assessment : Direct agreement to point 7 on lifting sanctions. Essentially, Iran is ready for business.

Point 11
The United States undertakes that, in light of the progress of negotiations towards a final agreement, frozen or restricted funds and assets of the Islamic Republic of Iran will be released and made fully available. These funds, whether held in the master account or transferred, will be used for any final beneficiary payment determined by the Central Bank of the Islamic Republic of Iran and will be fully available for use. The United States undertakes to issue all necessary permits and licenses on this basis.

My assessment : How much? From whom? When? What is the payment schedule? The key point - why is US not dictating the final purposes and usages of these funds? What if they are used for civilian nuclear energy?

Point 12
The Islamic Republic of Iran and the United States agree that an implementation mechanism will be established to oversee the successful implementation of and future commitment to the Final Agreement.

Point 13
Following the signing of this Memorandum of Understanding, and upon receipt of assurances regarding the commencement of implementation of Articles 4, 5, 10, and 11 of this Memorandum of Understanding, and the continued implementation of these steps, the Islamic Republic of Iran and the United States will enter into negotiations for a Final Agreement solely with respect to the remaining Articles.

Point 14
The final agreement will be approved through a binding resolution of the UN Security Council.

Now, what is not being mentioned?

1. What is IRGC’s say in all this? Why is there no public statement from the military branch?

2. What happens to Iran’s missile programme? What is the limit that will be agreed on? How many ballistic missiles can they keep or destroy?

3. What happens to the “fees”? Will Iran be stopped from collecting admin tolls and fees?

4. What happens if the proxies, Hamas and Hezbollah, do something? Who is responsible?

5. Why is Israel, who also started this war, not mentioned?

   As I said in my heading, there are a lot of moving parts, each in friction with the next. The heat is still on.
   
   

What does Veeva do?
At the core of it, Veeva is a software. But it is a very specialized and niche software.

It is tailored specifically for the global life sciences industry.

It serves as the digital backbone for pharmaceutical, biotech, and medical device companies, managing the entire drug development, regulatory compliance, and commercialization lifecycle.

Read that again. It is not just a CRM for sales management. It is not just a CRM for accounting.

It is a software for critical drug development.

Why is this software needed?

1. Pharma companies have to manage clinical trials of drugs. A software speeds up study setups and data.

2. The systems need to have lab information and Good Manufacturing Practice (GMP) data.

3. Global registrations across multiple health authorities need to be managed.

4. Pharma companies need to organize drug submissions.

You cannot do all of this manually on excel. You cannot do this without a software that specializes in the critical aspects of

1. Drug validation

2. Health registrations

3. Clinical data

4. Lab info.

And that’s why, a software like Veeva is needed.

But, what about sales pipeline and order management?
Well, you can use another software (like Salesforce) for this. But that data will NOT TALK to the data that is core to a pharma company.

And hence, Veeva provides a CRM tool as well, again designed for pharma companies.

So Sales CRM + Pharma Software = Veeva. All data points talk to each other.

What about AI? Can a company not build its own tool?
Yes they can. But I believe, AI can be an accelerant to Veeva’s growth.

Imagine you are a pharma company. You have used Veeva for years and years. All your data is already embedded in it. Your drug trials, drug pipeline, medical reviews, etc.

Do you want to migrate that core, private data to some other portal or would you like Veeva’s AI tool to help you get information quicker?

Would you not want to use Veeva’s AI and software to get patient data and analytics to map out treatment and advertising?

Why would you want to migrate?

The MOAT - Why Veeva is hard to replace.
Life Sciences Specificity: Unlike generic enterprise software (like Salesforce or Microsoft), Veeva is uniquely built to address pharmaceutical regulations, such as FDA 21 CFR Part 11 and EU Annex 11.

Built-In Validation: Veeva continually validates its own software for regulatory compliance (IQ/OQ), saving biotech companies millions in compliance overhead.

High Switching Costs: Once a life science company uploads decades of clinical trial and regulatory data to the Vault platform, ripping and replacing the software is nearly impossible.

The Network Effect: As a public benefit corporation, nearly every major global pharmaceutical company (and thousands of trial sites and vendors) uses Veeva. This makes sharing information, managing clinical trial data, and collaborating on supply chains seamless across the industry.

So, why has the stock fallen?
Veeva dominates life sciences, but a major structural shift in the industry has fractured the market.

Veeva was built on Salesforce infrastructure. And now, they have announced an end to that partnership. So companies have a choice. A rare window of opportunity.

To migrate to Veeva’s infra in the backend. To migrate to Salesforce. Or to choose something else.

Many pharma majors are choosing this window to launch RFP’s.

This. And AI. Threats of AI disruption have also led to a sell-off.

But, as I noted earlier, AI is an enabler and can help Veeva with more market share. Not a disruptor.

In March, we acquired Ostro, the leader in conversational AI for brands to provide patients and doctors with immediate, compliant answers through an easy-to-use chat experience. Ostro operates as a startup within Veeva and is now an important part of our Commercial Cloud. Things are going well, revenue and pipeline are growing as anticipated, and we have an ambitious product roadmap. We believe Ostro can be a significant revenue driver for Veeva and transformative for the industry, fundamentally changing how patients and doctors get information. - Management

The Lever
Since Veeva was built on Salesforce, they were obviously paying a certain amount to Salesforce. With that gone, imagine what it could do to margins!

The Numbers!
Revenue has increased from $1.8B in 2022 to $3.2B today.
Gross Profit Margins are high at 75% ish.
Operating Cash Flow has increased from $764M in 2022 to $1.4B today.

Forward guidance is 15% growth.

PE is at 29, so the PEG becomes still a bit higher at 2X.

So while there is still a great moat in the software, recurring subscription revenue is increasing, it is still a bit expensive!

However, worth having on watchlist.

Subscribe now

Disclaimer : This is not an investment recommendation. This post should be used for informational and educational purposes only and should not be construed as a call to action. Please consult a financial advisor before making any financial decisions. I am not a RIA.

The Copper Deficit
Everything needs copper. Your phone. Your laptop. Electricity. Transmission. Data Centers. AI.

Since AI and Data centers are in a structural tailwind globally, enormous quantities of copper are needed to build them and transmit electricity.

However, copper mining has not kept pace with demand. Prices have increased by 40-50% over the last 3 years depending on the benchmark.

So, how do we fill this deficit. By salvaging the copper already on the surface.

Copper Scrap Metal — How is it useful?
1. Smelting bottleneck - Mined copper must go through highly energy-intensive smelting and refining phases that are currently crippled by high acid costs and low processing capacity. In contrast, high-grade scrap—such as "Bare Bright" copper (clean, unalloyed wire)—bypasses the smelting bottleneck altogether.

2. Loss of performance - This is Zero. Recycled copper and new copper have zero performance difference.

Enter BIL - Bhagyanagar Industries Limited
BIL, at its core, sources scrap copper from across the world and manufactures various copper products / solutions. It is a simple business model and nothing complicated.

So, why are we talking about this?
3 Reasons:
1. EPR > Extended Producer Responsibility Framework : Original Equipment Manufacturers (OEMs) must ensure a specific percentage of their products use domestically recycled metal (escalating to a 20% recycled content mandate for copper by 2031–32). BIL sits here.

2. Value Added Products : If you simply buy copper scrap and make minor modifications to sell it off, you will not earn much. Anyone can do that. But BIL is specifically shifting its strategy to value added products. These contribute 62% of total sales (from 43% 2 years back). Higher share of VAP leads to better margins.

3. Data Centers : They consume and need massive quantities of copper.

The Numbers!
BIL’s margins are super low.
Operating margin of only 4% although improved from 2%.
This is classical of commodity companies. However notice that as the share of value added products increased, the margin doubled.

ROCE increased from 8% to 22%. In one year.

Management has given strong guidance as well on revenue numbers.

PE at 20X. Fairly valued.

The Risks
AI Data center build out may considerably slow down due to physical constraints.

BIL sources most of its copper from the US. Any increase in freight costs will impact margins. Why would they not source more from Africa or the Gulf is beyond me.

There is absolutely zero moat in this business. It does not take long for another existing recycler or a big conglomerate to set this up for captive consumption.

This looks to be more of a narrative play. But don’t discount the power of narratives.

Disclaimer : This is not an investment recommendation. This post should be used for informational and educational purposes only and should not be construed as a call to action. Please consult a financial advisor before making any financial decisions. I am not a SEBI RIA.

For decades, people in India have been stealing electricity for a few hundred rupees.

Every. Single. Day. If officials try to do something about it, they are beaten. Or politicians would not let a change happen. For vote-bank.

How does this theft happen?
If there is an overhead electricity wire, a hook would be attached to it before the meter to bypass the meter entirely. Or the meter itself would be tampered with.

What did this lead to?
Massive losses. Up to 50,000 CR.
Every. Single. Year.
State utilities couched it in soft language called AT&C Losses. But let’s call it what it is.

Theft. Loot.

This lost money meant grid didn’t get the needed upgrades.

It meant losses of state utilities spiralled.

It meant state utilities entered a debt trap. For example, Rajasthan itself carries 96000 CR of debt. UP carries around 62000 CR.

It meant frequent bailouts from the Centre.

It meant delayed payments to suppliers. Which in turn meant higher working capital.

It meant transformer and equipment failures. Costly replacement.

This became a Structural Vicious Cycle.

Enter RDSS! Revamped Distribution Sector Scheme
The Centre got really annoyed and frustrated with the inability of the state utilities to stop this theft. They stopped bailing them out. And instead provided them financial assistance ONLY IF the met strict loss reduction targets.

Put in 25 crore SMART METERS across India.
Smart meter manufactures build and operate the system and then recover their investments through a monthly fee, a subscription model sort of.

What does a Smart Meter do? How does it stop theft?
A typical smart meter acts as an intelligent computer for the electrical grid. Unlike traditional meters that only count total units (kWh) using a physical dial or simple digital screen, a smart meter continuously monitors, logs, and communicates power data.

1. Two-Way Communication: It essentially uses the internet to send power data to the utility every 15-30 minutes. So no human being needs to visit the premises to log in data.

2. Prepaid Billing Support: It features a built-in remote switch allowing it to function like a prepaid mobile SIM card, automatically shutting off power if the user’s balance hits zero.

3. Power Quality Monitoring: It tracks real-time electrical metrics like voltage fluctuations, power factor, and current load variations. If there is an uncalled for surge, the kill switch is activated.

4. Magnets : Traditional meters can be slowed down using magnets. Smart meters have a magnetic shield.

5. Higher Current : If there are two wires attached to a meter, a traditional meter cannot detect which has higher current. A smart meter can and bill accordingly.

6. High Voltage Protection : In traditional meters, people can use an electrical stun gun to fry the circuity board. Smart meters have surge - arrestor shielding.

7. Sensors : If anyone tries to break a smart meter, the power goes off.

Subscribe now

Genus : Police in the House
Genus serves as a primary manufacturing and execution backbone for the RDSS initiative.

It commands a 25-30% market share. Read that again. 25-30%.

And this is the market share where it builds and operates the smart meters.

However, Genus has so much capacity that it also supplies to other manufacturers if they manage to win tenders.

So the actual market share is far bigger.

Genus’ client base

The Financial Genius of Genus
The Special Purpose Vehicle (SPV) framework created by Genus and Singapore’s sovereign wealth fund, GIC, functions as a separate financial entity. GIC holds a 74% stake, while Genus holds a 26% stake, committing to a massive initial capital outlay of $2 billion.

The SPV bids for the projects, wins it, Genus acts as the sole supplier to the SPV, makes the meters and operates them.

Bulk of the financing is taken up by the SPV and Genus is the one that earns significant rewards.

High amount of debt is off Genus’ books. Risk of equity dilution becomes zero.

And this is visible in THE EXPLOSION in numbers.

The numbers!
Because the SPV helps to keep equity dilution zero and debt off the books, there is a visible improvement in ROE and ROCE.

Revenue up from 808 CR to 4750 CR over 3 years. 6X.
Net PAT up from 45 CR to 770 CR over 3 years. 17X.
Ebitda Margin up from 10% to 19%. 2X.
ROE up from 3.6% to 30% over 3 years. 10X
ROCE up from 6.14% to 26% over 3 years. 4X
Debt has also gone up but not by much. 0.35X to 0.94X, comfortable below 1.

How is the market rewarding the stock?
This is where things get a bit iffy.
Over 3 years, the stock has moved up only 3X. Where profits and ROE have exploded.
Over 1 year, the stock is down 10%.
PE multiple is at 16X.

Expected profit growth is around 30%.

PEG is very favourable at 0.5X

So, has the market missed this setup?
No. Mr Market is always right. This is primarily because in FY26, there were no big orders. In fact, Genus’ order book reduced from 28000 CR to 25000 CR.

If there were new tenders, it should have won.

So, are there new tenders coming?

The Market with the HUUGE UPSIDE
25 crores smart meters were part of the initial RDSS scheme. The reality would be closer to 31 crores.

Out of 31 crores, only half have been tendered. Out of of the tendered, only half have been installed.

The tendering AND installation pipeline is tremendous.

Source : Screener AI

As of FY26, Genus had installed around 1 Crore smart meters cumulatively.

They aim to install another 1 crore meters in FY2027 only.

But the biggest push might come from West Bengal, where both the Center and the State are aligned.

Genus will not win all these orders. But remember, it has a dominant market share plus it also supplies the meters to other providers.

We are looking at a multi year opportunity.

The Biggest Risk is vote-bank politics. And UP is going for elections early 2027.

Disclaimer : This is not an investment recommendation. This post should be used for informational and educational purposes only and should not be construed as a call to action. Please consult a financial advisor before making any financial decisions. I am not a SEBI RIA.

You have been using Claude, Gemini, OpenAI, etc for simple tasks like writing birthday messages, replying to emails and creating images. That is on your screens. In front of your eyes.

What if AI can help with manufacturing products better, placing them better, controlling people and even driving cars? AI has to flow from intangible screens to tangible, real-world, physical stuff. Ouster is at the fulcrum of a technology shift that works in the shadows of the physical world.

What does Ouster actually do?
Think of Ouster as a company that build “digital eyes” for machines. These eyes on machines are called sensors. Specifically, sensors that detect light and, on detection of light, range those objects on distance, movement, velocity, etc. This technology of sensors is called LIDAR (Light Detection and Ranging).

How is this technology useful?
Take out your phone and take a selfie. It is a flat 2D picture. But a LIDAR sensor shoots out millions of invisible laser beams EVERY SECOND in ALL directions.
Those beams bounce back to that sensor if the come across an object. And then, the sensor creates a highly accurate, real-time 3D map of the environment. This allows robots, vehicles and smart infrastructure to precisely understand how far away objects are, shapes, movements and velocity even in pitch black darkness or blinding sunlight.

Subscribe now

Yeah, so what? Where does AI fit in?
Raw LIDAR sensor output is just millions of dots and beams. A human or standard computer system cannot inherently infer a meaning. Ouster’s primary software, called Ouster Gemini, instantly translates those beams into clear, actionable intelligence. The sensors are the eyes; the software is the brain.

Ouster Gemini, their proprietary software (key term), uses AI powered Deep Neural Networks trained on millions of real-world physical objects to process data in real-time.

What does the software do?
1. Instantly identifies what an object such as a cyclist, a pedestrian, a car, a pole, etc

2. Tracks it in 3D with velocity, centimeter level spotting.

3. If someone steps into a restricted zone at night, the software can trigger an alarm. So, if a car stops inside a “Traffic Box”, it signals the traffic light to change.

4. Sensor Stitching : If multiple LIDAR sensors are mounted across a large area, a single 3D bird eye-view digital twin can be created.

Who would be end customers?
As you can guess, robots in dark factories, warehouse forklifts, unmanned machineries for agriculture, port cranes and pretty much anything that requires a human being to operate it.

Vehicles for smart tracking and autonomous driving.

Smart city traffic infra.

Retails - real time occupancy trends, user journey, dwell time.

What are the product advantages?
Ouster uses fewer parts. Which means constantly vibrating heavy machinery performs far better.

Since their tech is semiconductor based, as the technology improves, the sensors get higher resolution without mechanical complexity.

Share

Competitors?
Of course, there are competitors, but the real potential moat for Ouster is the software - Ouster Gemini.

The Potential Moat
Ouster acquired a company called Stereolabs which is knows for its techonology in this field. But with that, came a massive software ecosystem as well.

If you are a customer, common sense dictates that you will want a one-stop shop and will not want to “stitch” together various applications. Many clients don’t want to buy a raw sensor and hire engineers to build an interpetation platform.

By offering Ouster Gemini, which is AI-based, they make it incredibly easy to buy the ecosystem. A customer buys the hardware because the software is easy to deploy.

And this creates stickiness. If a major logistics firm maps its entire warehouse grid using Ouster’s software, it becomes incredibly difficult to switch as they will have to rewrite their entire stack and hence, customers get locked.

What is Rev8?
It is an internal project name for their latest chip architecture which allows their sensors to see twice as far and much darker objects at the same amount of power.
If power usage remains the same and output is better, it potentially helps with increasing gross margins.

Numbers
Right now, most of the revenue is from sensors but the aim is to change that mix and lock in more recurring, subscription based revenues. That will give incredibly better margins as operating leverage kicks in.

Revenue was $49M in Q1 2026, up 49% YoY.
GAAP gross margin at 43%
Still in loss, still bleeding cash, but still with around $175 million of cash in bank.
Aspiration of 30-50% revenue growth and margins in the 35-40% range.

The guidance range is very wide, which ideally means that there is some time to go for profitability.

They have cash, but I believe they will need to raise as well.

Since it is in negative and still in product development and capex phase, price to sales is the right metric to look at.

The price to sales number is around 13X right now, market cap of around $2.5B. Not terribly mad sky-high for a new AI-tech company, but not cheap either. (Spacex is at 100X).

Summary : The tech is exciting for sure and it does sound promising. Hardware + Software is a potent combination. 2027 and 2028 can be inflection years and this is new space, still not much on the “sensor” of investors.

Subscribe now

Disclaimer : This is not an investment recommendation. This post should be used for informational and educational purposes only and should not be construed as a call to action. Please consult a financial advisor before making any financial decisions. I am not a registered investment advisor.

So what is Spacex? Well, for starters, it is the largest IPO to be launched. Not in 2026, not in this decade. Ever.

The largest. With an expected valuation of $1.7T. T. Not B.

Subscribe now

The Business
Spacex has Three Distinct Business.

1. Rocket Launchers —> They are definitely an engineering marvel. Building reusable rockets can significantly lower the cost of space exploration and mission critical projects and the Falcon series from Spacex have exactly done that. This segment contributed around $4B in revenue but is unprofitable due to the heavy R&D required.

2. Internet —> Via Starlink. It generates $11.4B of revenue and is profitable at $4.4B. While there is nothing wow about the product, the delivery mechanism and the coverage area is unique.

3. AI —> Through xAI or Grok. Nominal revenues, with losses of $6B.

As you can see, combined revenue is around $18B, with the company still being in losses.

So, how should we value this business?
As a cable company? As an internet company? As a space - tech? As an AI company?
At the current revenue and valuation, the Price to Sales comes at 96X. Digest that. 96 times sales.

As an AI, maybe it would be difficult to swallow but still possible. As a space-tech with the heavy capex? No. As an internet company? Absolutely not.

And the bulk of the business right now is Internet.

Competition
Amazon is going to launch Kuiper this year. It is its own answer to Musk’s Starlink.
Common sense dictates that Starlink and Kuiper will enter into a price war.
In fact, the S1 acknowledge declining ARPU (Average Revenue per User - incidentally, a metric used in the telecom industry not in aerospace).

Similarly, xAI will continue to face stiff competition from other AI models and companies for end users (Gemini, Chat GPT, Claude, etc). Monetization, for now, will face headwinds.

In rockets, there is competition from RocketLabs and Blue Origin (Bezos). It does not reduce prices per se as the goal right now is to get technically proficient in launches and work in volumes, but this competition does provide food for thought as all three companies will have to carve their own niches on the kind of payloads they will cater to.

Subscribe now

Voting Rights
This is a slammer. Elon Musk will retain 85% of the voting control, despite holding only an estimated 42% equity. This effectively nullifies standard proxy votes and minority shareholder influence.

So, if tomorrow, X (or like me, if you still call it Twitter) becomes part of Spacex, your only option is exit in the public markets if you don’t want it.

What I like about it?
The SpaceX IPO is doing one thing most favourably. It is putting aerospace, space exploration, LEO right in the front of innovation. The next generation of warfare, and technological advancement will come from our universe and SpaceX, at the very least, is starting that race.

There are so many companies in space-tech : LUNR, RDW, ISTS, RXLB > but more on that later this week.

Share

Disclaimer : This is not an investment recommendation. This post should be used for informational and educational purposes only and should not be construed as a call to action. Please consult a financial advisor before making any financial decisions. I am not a registered investment advisor.

First, why I am writing about this?
Simply because for the better part of 2 years, the stock did absolutely nothing. And then (seemingly) out of nowhere it attached to itself a rocket booster.

5X in 1 year

3.5X YTD

What happened? And more importantly, how do we look at it from now?

What does this company do?
MTAR Technologies is a precision engineering company providing components for high-tech sectors —> Clean Energy, Aerospace, Nuclear, Space, etc.

Its revenue was 876 CR in FY2026 and, out of it, Clean Energy, dominated.

With around 70%, i.e., 615 CR.

Hence, for the purposes of this write-up we will stick to this segment only since it covers the majority revenue of 70%.

So, what does it do in clean energy, specifically?
Well, it does not assemble windmills, solar panels and all those things. It does something highly technical, highly skilled, very un-replicable. What is that thing?

To know that, you have to know what Bloom Energy does

Bloom Energy is based in California. It is a “green-tech” company that manufactures something called as Solid Oxide Fuel Cells, aka, Bloom Energy Servers (in photo).

These servers essentially convert an input like natural gas or hydrogen into electricity. They do this without any burning, hence the system generates electricity with virtually no emissions. Also, if they use natural gas today, they can “switch” to hydrogen seamlessly in the future.

And more importantly, these Bloom Energy Servers, this technology, serves as “always-on”, on-site power plants once deployed.

Subscribe now

Who are Bloom’s customers?
AI Data centers and hyperscalers that require tremendous amount of uninterrupted power as also power grids that just cannot have downtime. AI is the fastest growing segment for Bloom (obviously).

How does MTAR fit in?
The defining element of Bloom servers is something called as a “Hot Box”. And MTAR makes 60-70% of these for Bloom. That - is almost exclusivity.

MTAR manufactures “Hot Boxes”, the core energy units —>Bloom then integrates the tech and sells contracts —> Data centers / hyperscalers buy the green power to run AI.

And since Bloom has had a massive order backlog recently, MTAR has seen that stock price surge!!
(Bloom has won orders from Oracle, Nebius, Brookfield and others - order backlog worth $20B)

So, looks brilliant, right? Near exclusive partnership, AI tailwinds, sounds too good to be true.

Or, is it?

1. Iffy Management Guidance - I have been following MTAR since over 3 years now and the management is known for upgrading or downgrading the guidance. This is usual and expected behaviour, but the frequency is not. Also, MTAR management always give a “conservative” guidance, but even with that, they had two downward revisions in the same year with the final sales number even lesser than those two revisions. Check below:

Source - Screener AI
And hence, because of so many “frequent” revisions, it is very difficult to fully go by the guidance. What else?

Share

2. Customer concentration - Until a few years back, Bloom was their only dominant customer. They realized the risk and started diversifying by signing contracts with other companies. However, clean energy, as we have seen, contributes 70% of the total revenue. It is a reasonable to assume a significant portion of it coming from Bloom.
This, is single biggest risk factor to MTAR. See below:

Source - Screener AI

PS - Yuma and Santa Cruz are two versions of hot boxes (the latter being the latest one)

Source - Company PPT

3. Valuation - The PE ratio as of today is 247X!! This is not a typo. Now let that sink in.

PE > 247 X
1 year return > 5X
You might assume ROCE and ROE would be high considering MTAR has a niche technology and near exclusive deals with Bloom. But no.

ROCE = 17% ROE=12% (Risen, but still low).
Debt Equity has actually nearly doubled from 0.24 to 0.45.
Ebitda Margin has improved by 180 bps compared to FY25, but still lower by 730 bps compared to FY23.

Working capital days, although improved, stay elevated.

So while MTAR is working on a niche technology with high barriers to execution, the numbers are absolutely priced to perfection with less than zero margin for error.

A sky high valuation always results in either a significant time correction or a violent price correction.

Oh! And What about Bloom’s numbers?
1 year return = 14.5X
ROE = 0.8%
ROCE = 2.48%
D/E = 3.2x
PE = 1282X

So, well, MTAR, the supplier to Bloom, looks like a steal by comparison :) But, is it?

Disclaimer : This is not an investment recommendation. This post should be used for informational and educational purposes only and should not be construed as a call to action. Please consult a financial advisor before making any financial decisions. I am not a SEBI RIA.

The Background:
The Indian government has set up a massive, gigantic target for solar installations. The industry is booming, and rightly so. Right from captive industry and utility-scale to residential rooftop solar and (even) solar water pumps, there is huge demand for those glass modules that trap the incoming rays of our star.

For context, India’s current TOTAL electricity generation capacity is 520 GW and out of it, renewable is 270 GW.

This is already installed.

The target for Renewable Energy out of the total energy generation is 500 GW by 2030.
From the current 270 GW.

And out of that, solar energy would be roughly 280 GW. Which is 56%! Sounds big, right?

But, how much solar power capacity is already installed? Roughly 150 GW.

So that leaves around 130 GW only to be installed over approximately 3-4 years. Roughly 40-45 GW per year.

Why does this 130 GW matter?
Because we have to look at the overall solar module manufacturing capacity currently installed and coming up. Is it more than the projected demand of 130 GW? Less? Same?

What goes into a solar panel?
A solar panel consists of something called modules.
Modules are made up of cells.
Cells are made up of ingots and wafers.
So, Ingots —> Wafers—> Cells —> Modules —> Finished Product

Module Capacities - The Structural Surplus
India’s current module capacity is already at 210 GW!
And the planned capacities over the next couple of years may top this up to 280 GW.

So, pending solar power demand = 130 GW (over few years). Module supply = 210 GW (yearly)

If you assume that the demand of 130 GW over 3-4 years will be translated to around 40-45 GW yearly, we are looking at a massive overcapacity of approx 175 GW per year already today. That is a structural surplus of ~4x.

This is for modules.

What about the three other things in the value chain that go into making the modules (Ingots —> Wafers—> Cells —> Modules)?

Majority is being imported. Which means that solar module manufacturers in India are paying a premium to import the raw materials (well, somebody else is there to earn profits, right?).

What does this mean?
If you have to sell 3 sofas and the demand is only for 1, what will you do? Reduce prices or start finding new markets.

Reducing prices will lead to margin compression. To avoid margin compression, you build the raw material factories internally and consume cells, wafers, ingots rather than importing. Which is exactly what is happening.

Once these Raw material factories are up, the module makers will stop paying the profits to others and save costs.

However, that only protects margins. What about the finished product? Where does that flow, considering the surplus?

Subscribe now

Exports
There is this whole China+1 thing happening across the Western world, where they want not only modules, but the ingredients (wafers, ingots, cells) also to be of non-Chinese origin. But this rests on a key assumption - that Chinese imports will continue to be strictly regulated.

Why that assumption is shaky
If your house catches fire once, what will you do? Buy fire extinguishers, make it fire proof, essentially spend money to ensure you are well protected next time. This is your first line of action.

Common sense dictates that.

With the Russia - Ukraine war, budget deficits across the Western world are going to increase. Defence spend is going to increase. That money has to come from somewhere.

While I am not claiming that that money will come from solar sector, but if a Chinese module is coming up at significantly lower cost, there is only so long that you can hold off in the face of massive spending increases.

In any case, Indian module makers might all scramble to export to the same market and flood those markets undercutting their own margins.

Share

Summary
The sector has some tailwinds right now and this is reflected in the financial performance of the listed companies in the sector (below), but will it continue? Time will tell. Reasonable P/E, reasonable D/E, high ROCE, super high revenue growth being shown by most companies.

Below companies being used for representation.

But what about the future? There are quite a few unknowns. But today, one thing is known : Overcapacity is building up.

Disclaimer : This is not an investment recommendation. This post should be used for informational and educational purposes only and should not be construed as a call to action. Please consult a financial advisor before making any financial decisions. I am not a SEBI RIA.