—————————————————————————————————————————————————————

A few months ago, I talked about Status quo and the fundamental shift that is required to realize 30-50X increase in Power density within AI data centres.

Let me do a quick recap

Problem:

• A single NVIDIA B200 or AMD MI300X consumes 0.9kW – 1.4 kW. Dont even ask me about Ruben.
• A “pod” of 8 such accelerators + CPUs approaches 12–15 kW
• An AI training cluster is hundreds of pods – multi-megawatt

At 12 V distribution you would need kilo-ampere bus-bars thicker than a wrist; power loss alone would exceed the chips’ TDP

• On top, Every 1 cm² of copper busbar you delete saves you north of $20/kg and a small argument with the procurement team.

Hear about it directly from the Horse’s mouth

NVIDIA’s own whitepaper claims ~45% less copper in an 800 V plant.

So First principles first, whats actually happening?

Here is a nice infographics to follow along the transition. Power distribution is changing from AC into DC microgrids.

Optics is forcing this transition quickly. And now that the retail knows what happened to Optics stocks, they are coming swinging for this sector too.

Here is what you need to understand about the Power Semis

Voltage is the cheat code. Crank it up early and keep it up for as long as possible. Then step it down ONLY inches from silicon, and you can unlock power dense, greener, cheaper AI factories.

If you need a refresher, go read this 6 months old post first.
On this post, VICR did 5X since this post.

If you are confused about SiC, vs GaN etc. Here is a great infographics that shows where does each of them thrive in the Data center value chain

Now that the market is interested in this deeply, lets go through the 3 small cap names that might do a few multi baggers for us. I will assess them and rank them.

Read more

In my last message on 16 Mar , I talked about how I am raising cash in my folio since a few weeks and that bounces were to be sold. This bounce seems no other, especially because Trump yesterday said nothing new in his speech.

Trump TACO’d 3 times and markets simply ignored until 2 trading days ago. And now this reaction is getting awashed too. Remember that bullish news + bearish action = bearish

This rally has been so good so far, but On a dead cat or oversold bounce, the QQQ is going to rebound aggressively to overbought conditions, stop there, and then crash back down to earth.

So we gotta stay vigilant and watch the reaction today and tomorrow, SPY at 610-615 is gonna be very juicy if there is a last leg downwards.


My entire strategy is to find misunderstood names that are overly negative and get a good entry point. Timing the macro markets perfectly is not where the real money is made, but rather buying when everybody is scared in strategic SPY and QQQ levels like 610-620. Thats when I get in heavy with the cash I raised.

This has allowed me to do well as well as stay calm in these moments of volatility. My opinion is that the markets don’t bottom when Oil is still making all time Highs. Lets see

That being said, here are some of my favorite names to play right now. I usually talk about small caps, however, in downturns, you wanna get into famous big names so you can hold with conviction and for the strong bounce.

RDDT

It is getting lumped into a software basket. Too much negativity for a product that has real network effects and real moats. Yes there is some scare around META’s lawsuit and being liable to creating addictive software, but it is worth to look at

It’s showing early signs of Operating leverage. Expenses are not lifting off as fast as the profits are. This is a sign of a great business. Ofcourse the margins and cash flow is seeing a lift as well

I am super pleased to see the uplift in ARPU in the US as well as the global average revenue. Seems like they have been able to grow their Daily active users as well as the revenue they generate from them

Stockdrifts is working on a feature to identify if growth is accelerating or decelerating. Market is punishing deceleration, so this feature is important. In the meantime here is a snapshot of how growth is slightly accelerating in the US and Internationaly

Buy Zone: $90-$110 is a great accumulation zone

MU

All the news about DRAM prices dropping is peak stupidity.
90% of DRAM is sold as contracts and the contract prices is what actually matters.

This is what reputable research house are talking about. 60% QoQ increase in 2026. HOLD MY BEER!!

There are plenty of X posts talking about how turboquant is a year old innovation and how it is not going to relieve the pressure on tight Memory supply

This correction is a gift. Just look at how this revenue is actually accelerating. Not growing, but accelerating.

Another reason why MU is a better bet than SK Hynix and Samsung because MU has better access to North American helium-linked supply chains

U.S. helium logistics can be more secure during the Iran and US war

Helium prices have roughly doubled since the Iran war began, according to Fitch Ratings. Severe helium disruption is actually bullish for premium memory pricing And that usually helps Micron compared to its Asian competitors

Ofcourse, MU has done 3X suddenly from the last base, so you gotta be careful about your entry points. I am eyeing the 260-290 level. Fingers crossed.

Buy Zone: $260-290

Rest articles covers 2 more names I am excited about

Read more

What Happened & Why the Stock is Down 20%?

For paid subscribers, I have my analysis of T1 and TOYO’s earnings and the arguments about the Ethiopian TOYO’s strategies. Here is a snapshot, TOYO is making a lot of EBITDA vs TE

Read more

This is a conversation space exclusively for subscribers—kind of like a group chat or live hangout. I’ll post questions and updates that come my way, and you can jump into the discussion.

Join chat

How to get started

1. Get the Substack app by clicking this link or the button below. New chat threads won’t be sent sent via email, so turn on push notifications so you don’t miss conversation as it happens. You can also access chat on the web.

Get app

2. Open the app and tap the Chat icon. It looks like two bubbles in the bottom bar, and you’ll see a row for my chat inside.

3. That’s it! Jump into my thread to say hi, and if you have any issues, check out Substack’s FAQ.

They hold a lot of memory patents and is crucial for HBM, MRDIMM memory stack. All the big ones SK, MU, Samsung as well as Google will have to pay them substantial amounts. Litigations going on for over 15 years. Will reach a conclusion this year. District court already ruled in favour with $1B. Should be more actually

There might be partnership deal with SK Hynix coming very very soon. SK Hynix 5 year partnership deal is expiring. The space is too valuable for SK Hynix to NOT renew.

But leaving the litigations aside, their product revenue is ramping up massively. Here is a snapshot from Earnings from a few weeks ago.

There are every indication that Q1 2026 might be better as well.

Here is some commentary from the CEO. I find that very very interesting

“Sales volume for Lightning, Netlist’s overclocked and low latency DDR5 line of products, is ramping up nicely in the system integrator market segment. Adding to this, we’ve completed qualifications at a global server OEM, and they, in turn, are currently in testing with many of their end customers.”

— Chuck Hong, Q4 2025 Earnings Call (prepared remarks)

So their products are coming along well and they are not just relying on litigation and IP fees.

Also,

“A ruling in Netlist’s favor would direct U.S. Customs and Border Protection to stop Samsung memory products that infringe these patents from entering the U.S.”

— Chuck Hong, Q4 2025 Earnings Call (ITC discussion)

Secondly, do you think for a such an important product like memory, Samsung would diddy around and NOT settle? They have already won this case 3 times in district court. ITC is just going to force a fast resolution.

Also, I thought it was interesting to point out that cost of borrowing is at 20%. And now suddenly overnight 8.5M shares become available.

I expect manipulation or some kind of funny business soon. So if it comes to $1-1.2, I might be looking to averaging upwards.

I was the ONLY analyst covering NLST on X, but I can assure that there are plenty of cult retailers on this name over the years.

I am raising some money in my account to profit from the opportunity IF it arrives due to a prolonged war.

I dont think the market’s really realized the risks/chances that Fed might not cut or might actually hike. Market’s been expecting Fed’s control by Trump and this might not materialize in mid terms.

In any case, all you can do is get into misunderstood stocks, at a good entry point and wait for your thesis to play out.

In the next post, I want to share some exciting ways I have been able to build alpha pipelines. Will send the next post soon

Spoiler Alert: $TE will do just fine!!

Module assembly capacity has surged to ~65 GW annually, HOWEVER, the supply chain remains imbalanced with only ~8.3 GW of wafer capacity and ~3.2 GW of c-Si cell capacity domestically

Okay while we scratch our heads over that, there is an S&J’s interview with TE’s VP this Thursday. Definitely don’t miss

Okay, lets look at the supply chain that exists. 5 steps to get to that sweet sweet Solar Panel

Supply chain capacities across the world

This chart allows you to appreciate the mismatch and how the dream of full vertical integration is still far far away.

STAGE 1: POLYSILICON

Combined US nameplate: ~50-55K MT/yr. Actual production ~33K MT. If all of that was converted into Solar cells it would be equivalent to ~13-16 GW of solar modules.

Keep in mind that Polysilicon is also required for semi conductors. Albeit the grade gotta be much purer than solar. 98% of Polysilicon is used for Solar

The real issue is that the 45X Credit: $3/kg only and the gap between Western polysilicon and Chinese is around 12$. So that really is a big gap. Pure, US supply chain is going to come at a short term cost. Not to mention there is simply no capacity for Polysilicon in the US

I think that TE, TOYO and eventually Tesla might do the most minimum possible to satisfy FEOC. They are eating glass or paying 3 times the price to get US Polysilicon.

Atleast the Good news is that we can see Corning promising to increase the production massively. These infrastructure companies do not simply increase capacity. They do it when there is massive shortage.

It is also an indication that the demand is through the roof for Polysilicon if GLW is planning to exponentially increase their Solar revenue.

Lastly, I spun up Claude code Python to calculate the difference. When you source from Non China Polysilicon, it suddenly becomes 1/3 your total costs.

I think there is a lot you can do to mitigate this. For eg. Credits might help and some more financial sorcery.

Lets see if this really comes through in 2027-2028. However, it did happen that a lot of Chinese supply shut down and Polysilicon prices bounced up ~30-50% up.

Also P.S. cell-level cost is deceptively low; a company’s success ultimately depends on the quality of its cell manufacturing process. Because a better manufacturing process means higher efficiency for the panels. And even a little bit more efficiency over its lifetime of a few decades can compound quickly.

Have a look at how Solar cells are produced.

I am sure Elon will delete a LOT of steps and rethink the whole industry from fundamental point of view, however the reality of things are still going to be that having American Solar supply chain is not a short term target.

I am planning to give away all of my Solar watchlist to Annual subscribers. My favorite small caps or best multibagger opportunities with liquidity in the space.


I have been sitting on it for far too long. Will also release interesting watchlist in other themes as well.

If you have been looking to invest in alternative grid scale Batteries and never got around to EOSE, here is a really interesting stock <$1B market cap

Before I talk about this stock, lets set the stage right. After the disastrous earnings of FLNC with suppressed margins and having to move around Battery supply chains through 3 different US facilities in order to chase them 45X credits, it’s time to look at something new

Read more

I have huge respect for Daniel, S&J, and DVB and value their insights on TE massively. Cherish them as fellow investors. However, I think there are a few angles they might be overlooking. Let’s unpack them

Here Daniel mentioned that having whole domestic US supply chain is important.

TOYO’s been alleged to be trade loop hole extractor

DeepValueBagger is assuming that TOYO isn’t US production.

However, Even as seasoned/knowledgable these investors are on the Solar topic, here are my assessment and rebuttal on it

INSIGHT 1: TE literally doing the same as TOYO. Shipping their US polysilicon to Asia, converting it into cells and shipping it back

First lets start with what TE/T1 is actually doing with their OWN cells

- T1 Energy CEO Daniel Barcelo in Nov 2025 explicitly stated:

“Until we have our solar cell plant up and running, we currently are sending our polysilicon to Southeast Asia in order to make wafers and cells and then bring those cells back to the United States.”

Here is a timestamped video for when he said that.

So TE/T1 is literally doing what TOYO is doing and getting a lot of shit for it. Since TOYO has now shifted to Ethiopia, travel time there is still lesser than the South east partners of TE/T1

In the show it’s talked about how there are literally 1 or 2 GW of Solar cell capacity in the entire US. Just TE’s G1 capacity is 5GW of modules.

Recall how Solar modules are finally produced

Also, Remember that Tree Oaks collaboration told us that TE/T1 is aiming to be 70% US supply chain by end of 2027. That’s still 2 years away.

Conclusion: Even the American Solar TOP DAWG is doing exactly what TOYO is doing and aiming to be only 70% American by the end of 2027

I wrote 3 TE/T1 deep dives back in start of October. I have spent countless hours diving deep into the industry

INSIGHT 2:

Read more

Just hours later, court rulings removed the biggest risk to this small cap with prejudice. Meaning this small cap has irrevocably won the court rulings. It cannot be appealed again. It’s final

No I am not an insider on this stock, just hundreds of hours research deep into Solar

The stock’s already up about 25% since my post. But I still think this is an incredible buy point here for the following reasons

- The official PR piece is still not yet released

- Market’s yet to realize the implications of this

- Schwab and Fidelity still restrict it due to low liquidity—most investors haven't arrived yet

Here’s the original article if you missed it.

Subscribers have had this name since Friday morning 9.49AM EST.

I’ll publish the full write-up for the wider audience within 24-48 hours since the train might be leaving the station soon

For now, lets dive further on this name

Read more

• “Made-in-USA” Pivot & Supply chain strategy in the Solar sector

• It’s under $300M Enterprise Value ($250M Mcap + $35M bank loans)

• Insider’s own >80% of the stock and Float is under 20%

• Experienced growth of 300% in 2024, and then >100% in 2025. Revenues in 2025 are $350M+

• 2027 Expected EPS is 50% of the current share price (Insane!!). Trading at 2X times 2027 Expected EPS

• Virtually no Analyst coverage

This stock is fairly illiquid so please don’t hit buy button immediately. Assess the risk of chasing and be responsible

The stock is $TOYO

This company figured out how to combine:

1. Cheap African hydro-power
2. US Polysilicon (to keep Uncle Sam happy)
3. Texas final assembly (to get the tax credits $0.07/watt)
4. Vertically integrated Solar supply chain

The result? A Vertically integrated Solar stock trading at under $300M Enterprise Value with insider ownership so high it looks like a dictatorship.

Investment Thesis: Asymmetric risk/reward with 2025 guidance implying ~$375-400M revenue (with 4.2- 4.4 GW) and $39-45M net income at a market cap of only ~$235M (P/E ~5-6x)

INSIGHT 1: THE MATH IS NOT MATHING

Enterprise Value (EV) < $300M
2025 Revenues ≈ Guided for $360-$400M
2027 Projected EPS ≈ $3.61

Wait, pause.
The stock is trading around $7-8. The projected 2027 earnings are $3.61.That means this thing is trading at roughly 2x 2027 Earnings.

How much would you pay for a stock in hot sector growing massively?

INSIGHT 2: THE SUPPLY CHAIN HACK

Trump slapped a 42069% tarriff on Vietnam. The Giga chad Junsei Rye (CEO of TOYO) pivoted the supply from Vietnam’s plant (4GW) for non US. Also ventured into onshoring jobs to America and build Solar supply in the US.

Toyo had acquired Solar module facility from Solar cell in Nov 2024. It’s planned to scale to 1.5GW by 2026, 2.5GW by 2027,then finally to ramp to 6.5GW by 2029. They acquired it cheaply because there was no money in Solar back then and everyone was bankrupt

The Texas facility only began commercial deliveries in mid-October 2025. So it’s ramping up only now

TOYO’S GAMEPLAN:
• Step 1: Buy Polysilicon in the US (Signed a deal Jan 7, 2026). This keeps the “Forced Labor” police (UFLPA) happy.
• Step 2: Ship it to Ethiopia. Why? Because Ethiopia has massive hydro-power dams (cheap electricity) and labor costs that make China look expensive. Also, no AD/CVD tariffs.
• Step 3: Turn Poly into Cells in Ethiopia.
• Step 4: Ship Cells to Houston, Texas.
• Step 5: Assemble Modules in Houston.

Result: You get a “Made in USA” module that qualifies for the 45X tax credits ($0.07/watt), but your cost basis is pretty good.

They are essentially laundering US polysilicon through cheap Ethiopian electricity and bringing it back home as a value-added product. It’s geopolitical arbitrage at its finest.

Have a look.

When compared to TE, they don’t get $0.04/Watt for producing the solar cells yet. However, think of them as G1 facility in the US.

Here is a great slide to understand what the current incentives are. Read my first deep dive on TE to get a clearer picture on credits

So even though they might miss out on $0.04/watt for cells, they can easily make it up for lower land and labor cost for the solar cells in Ethiopia

Now they signed a sales contract with a U.S. polysilicon manufacturer, to secure domestically sourced critical raw materials for TOYO’s solar manufacturing operations. This agreement adds U.S.-sourced polysilicon to TOYO’s existing non-Foreign Entity of Concern (“FEOC”) overseas supply

The company claims that they will be eligible for 45X. Let’s pray and hope that they are right. And that their customers/credit buyers believe it too.

“During the trial production phase, the Company has received strong demand from its U.S. clients for the modules manufactured at the Houston Facility. As operations scale, TOYO anticipates robust demand from utility-scale developers seeking reliable and domestically manufactured modules that qualify for favorable tax incentives.”

“Expected to qualify for tax incentives under Section 45X of the Internal Revenue Code, providing up to $0.07 per watt in production credits through 2030. This strengthens TOYO’s cost competitiveness while supporting broader goals for domestic energy manufacturing.”

INSIGHT 3: Growth and Demand is through the RooF!!!

They shipped as much Solar cells in H1 2025 as they did in it’s entire 2024 Fiscal year of operation. These babies can really scale!!

They just coalesced with sister company VSUN. I will probably make a deep dive 2 to get into the spaghetti of subsidiaries. However for now, focus on the growth and distribution of VSUN.

INSIGHT 4: 82% INSIDER owned with Low Float

One man, Junsei Ryu, controls 67.6% of the company. The public only trades the remaining 32.4% (about 12.2 million shares). This is a highly centralized "low float" stock.

Some credentials on Junsei Ryu. I couldn’t find any interviews of him online, but he seems like a Solar Giga Chad to me

INSIGHT 5:RISKS

a. Jinko Solar Litigated TOYO on IP infringement. The company does not comment on what the damages could be.

However, keep in mind that everybody is suing everybody else for Topcon Patents. It’s an entire Suing-fest. OoooOO

b. Company claims FEOC compliance, but we have to see something real to fully de-risk the fact that the company setting is so convoluted.

CONCLUSION: It’s an Asymmetric Opportunity

If 2025 guidance is achieved ($375-400M revenue, $39-45M net income):

• Forward P/E: ~5-6x (vs. First Solar at ~15x, JinkoSolar at ~5x)

• Revenue growth: ~112-126% YoY

• Price/Sales: ~0.6x on forward revenue

Bull Case

• Tariff arbitrage proves durable and TOYO captures meaningful share of $30-40B U.S. solar module market

• Section 45X credits at full capacity = ~$175M+ annual benefit by 2029

• JinkoSolar patent suit settled or dismissed; TOYO continues growth

• AI-driven energy demand accelerates utility-scale solar buildout

• Target: 3-10x upside quickly (~$18-50/share)

Bear Case

• JinkoSolar wins injunction, blocking U.S. sales of TOPCon products

• Ethiopia loses tariff exemption or faces new trade restrictions

• IRA modifications reduce or eliminate Section 45X benefits

• Module price collapse compresses margins to unprofitability

• Target: >50% downside (~$1.50-3.00/share)

P.S. Neils bought this stock to my attention. He is usually early and gets into very volatile names

DISCLAIMER:

I invested in this stock and is about 1/2 the current TE holdings. As I discover or get more information, I will get in bigger

Serenity has been going viral in 2026 with some posts getting millions of impressions. This account is super respectable

So I had to check out and validate their claims

Debt is low. No crazy dilution. Ofcourse lossmaking. But the revenues is steady since a few quarter. All right, interesting, lets dive deeper.

It’s an Israeli company so they are not required to file 10Q. But I found the relevant file quickly. I fact checked claims and formed opinion about the company in this article

You all know how deep in fundamentals I like to get

The inventories claim actually check out. It is indeed only $10.9M. Lets fact check all important claims

1. Claim: “$93.5M in Cash, 0 Debt.”

• Verdict: TRUE.

• Evidence: The Balance Sheet shows Cash and cash equivalents ($31,588) + Short-term deposits ($61,961) = $93,549 (thousands).

• Debt: There is no line item for “Long-term debt” or “Bank loans.” The only liabilities are trade payables, accrued compensation, and lease liabilities (which are operational, not interest-bearing debt in the traditional sense)

2. Claim: “~$11M inventory.”

• Verdict: TRUE.

• Evidence: The Balance Sheet lists Inventories at $10,974 (thousands).

3. Claim: “High gross margins ~69.1% (CIB/ProAV) margins, 43.2% automotive.”

• Verdict: MOSTLY TRUE (Automotive is actually higher).

• Evidence: Note 14 (Segment Info) for 9 months ended Sept 30, 2025:

  • CIB: Gross Profit $25,966 / Revenue $37,802 = 68.7% (Very close to 69.1%).

  • Automotive: Gross Profit $6,373 / Revenue $13,420 = 47.5%.

  • Note: The tweet actually understates the Automotive margins (43.2% vs actual 47.5%), perhaps relying on Q3 specific data (which was 43.2%), but the 9-month trend is strong.

4. Claim: “Inventory of US$82 million... is a typo.”

• Verdict: TRUE (The math supports the typo theory).

• Evidence: Total Assets are $136,691 (thousands).

  • Cash/Deposits: ~$93,549.

  • Remaining assets allowed: $136,691 - $93,549 = $43,142.

  • It is mathematically impossible for the company to have $82M in inventory because $93M (Cash) + $82M (Inventory) = $175M, which exceeds the Total Assets reported ($136M). The inventory is definitively ~$11M.

5. Claim: “New CES 2026 info... pivot to robotics.”

• Verdict: TRUE

• Evidence: The acquisition of Acroname (described in Note 3 as “specializing in advanced automation and control technologies”) and the massive growth in the CIB segment (which includes machine vision/industrial) financially supports the narrative that they are pivoting to robotics.

6. Claim: “Extremely heavy dilution at $11.5 Strike from warrants.”

• Verdict: TRUE.

• Evidence: Note 12 lists “Private Warrants: 3,330,000” and “Public Warrants: 5,750,000” that were excluded from EPS calculations because they are anti-dilutive (meaning the stock price is currently below the strike price). While the filing doesn’t explicitly list the “$11.50” strike price in the text, this is the standard SPAC warrant strike price, and the presence of ~9M warrants is confirmed in the notes.

Following are some of my insights on the company that might help you decide if you want to jump in with Serenity

1. The “Growth” Story is Lopsided (Automotive is actually shrinking)

While the company is often pitched as an Automotive AI play, the filing reveals that Automotive revenue actually declined year-over-year for the 9-month period (from $16.6M in 2024 to $13.4M in 2025). The entire revenue growth of the company is currently being driven by the Cross Industry Business (CIB) segment, which grew back ~53% ($24.6M to $37.8M).

Could this be the start of CIB recovery? I love the thesis that Robotics play will be able to be validated quickly (~6 months)

2. Aggressive Buybacks Despite Operating Losses

It is highly unintuitive for a company reporting a Net Loss of $(22.8M) to spend more cash on share repurchases ($23.4M) than it lost in operations. Usually, loss-making growth companies hoard cash. Valens is aggressively returning capital to shareholders, behaving more like a mature cash-cow than a loss-making semi growth stock.

3. The “Acroname” Acquisition Masked Organic Burn

The acquisition of Acroname (Note 3) added ~$4.1M in revenue for the 9-month period. Without this acquisition, the company’s revenue growth would look less impressive. Furthermore, the acquisition brought in $1.3M in cash, slightly padding the cash flow statement against the purchase price.

4. Inventory Composition Shift (Finished Goods vs. WIP)

While total inventory rose slightly ($10.1M to $10.9M), the mix shifted drastically. Work in Process (WIP) dropped by ~21%, while Finished Goods spiked by ~32%. This suggests the company has converted raw materials into sellable chips but hasn’t shipped them yet. This could indicate either preparation for a big order or a sudden slowdown in demand where production outpaced sales

5. Acroname might be hitting ambitious targets

The company has a liability of ~$2.5M for “Earnout” payments to Acroname shareholders (Note 8). The fair value of this liability increased (an expense of $81k), which unintuitively is a “good” sign—it implies the acquired company (Acroname) is likely hitting the performance targets required to trigger those payments.

6. They filed important design wins in 2026

This a great long-term opportunity and it is definitely real, but near-term catalysts for technical products and growth remain 12-18 months away.

7. Subsequent Event: CEO Change Signals Pivot

There is a new CEO appointed in Nov 2025. Combined with the shrinking Automotive revenue and the massive options grant (4M options) to the new CEO, this suggests the Board was dissatisfied with the previous trajectory. It validates the “pivot” narrative

MY VERDICT

This is indeed mispriced and might still be substantially cheap after a 50% move on Friday. Fair value seems to put the stock price at $4-7. I am keeping this in my watchlist. As for the future of the product it requires more due diligence

RISKS

1. One of the risks not mentioned in his post, is that there is considerable Insider sales. CEO has been periodically selling,

Founder Ex CEO Dror has sold about 10% of his holdings in the last month (Total only 1% of the outstanding shares). Maybe it’s not such a bad thing since he stepped down as CEO in Nov 2025

2. One-Time “Insurance” Profit Boosted Q3
In Q3 2025, the General and Administrative (G&A) expenses were unusually low ($2.2M vs $5.8M in Q3 2024). Note 5 reveals this was due to an insurance recovery asset of $1.476M recognized regarding a customer claim. This is a non-recurring event that artificially made the Q3 operating loss look significantly better than it actually was operationally

3. High Stock-Based Compensation (SBC) Relative to Revenue
   The company incurred $11.7M in stock-based compensation on $51.2M in revenue (9 months). This means ~23 cents of every dollar of revenue is effectively paid out in equity to employees. While this preserves cash, it is a significant dilution driver that is excluded from “Adjusted EBITDA” metrics often used by analysts.

I promise to provide more alpha and insightful posts this year

Back in May/June, the battery sector was dead money. Nobody wanted to touch it. But contrarian investing is about identifying value before the crowd arrives. The market has finally understood the importance of storage. Most names in my watchlist delivered 4-6X returns in just a few months

Just look at this insane stock prices.

Now Solar is shifting gears.

The sector is catching a bid, charts are bottoming out, and we are seeing breakouts from long bases. This is the ‘sweet spot’ for entry.

Allright, let me show you exhibits of why this sector is going to nail it in the next months/years

EXHIBIT 1

Hate him or love him, but nobody’s made money SHORTING him

EXHIBIT 2

TAN, the Solar ETF broke out of it’s 4-5 year trend.

EXHIBIT 3

The economics and the markets are clear on this one

EXHIBIT 4

Most new additional capacity added was Solar. Yes, even under Trump Regime.

EXHIBIT 5

You can now pack your Solar power in a lunch box and eat it when you please. Pack me some Sun please!!

EXHIBIT 6

It’s one of the fastest and cheapest way to bring power online quick

EXHIBIT 7

Big companies are signing very long renewable contracts. Microsoft recently signed 20 year contract. And now Google and META. These companies are Pro’s and they know what they are doing

EXHIBIT 8

Regardless of unclear policies in the coming years, Solar and Batteries are the biggest interconnection request in ERCOT

In the Following posts, I will be detailing my favorite names. Some with higher risks and some with the best R/R. There’s a large number of stocks to choose from for different risk spectrums

Stay tuned for the next one. I will do a deep dive into a name I am currently accumulating, and hope would become my next TE like conviction

IF there is over allotment then $21M for stocks and $18M for convertible debt can be further bought.

The total amount raised could reach $299 million if the underwriters exercise their full over-allotment options

We will probably know within 24-48 hours if this raise was successful or not.

They claim that it will be used for the following

• FEOC compliance by Dec 31
• Build-out of G2 Austin (first 2.1 GW)
• Working capital + debt repayment

Trina initially received a 9.9% stake (approx. 15.4 million shares) for it’s factory and then additional 12M shares. Total is about 27M shares bringing the total to 16.6% in the last filings

However, Trina’s ownership is nowhere close to <25% required. So that’s not it.

Then I looked into their latest 10Q and found that Trina does own a lot of T1/TE’s debt.

Or here is an easier to read summary. And Bingo, here in lies the issue. Trina holds 67% of TE’s debt.

Under OBBBA, an entity is designated a “Foreign Influenced Entity” (FIE) and ineligible for these credits if:

• 15% or more of its total debt is held, in the aggregate, by “Specified Foreign Entities” (SFEs).

So this needs to be sorted out. The unfortunate thing is that 220M or 1/3 of TE’s debt was near zero. We gotta look at the new terms for it’s debt.

The important thing to remember here is that there was no mention of the exact price at which convertible note will convert into shares. So we don’t exactly know how much dilution from that one.

Here are my assumptions. Stock offering could be at at $5, then the dilution due to it would be 28M shares so approx 12%. Convertible note will have much less dilution. So overall I would put this dilution at 15-18%. (Rough calculation since the full terms are not public yet)

All of this to get 300M cash injected. This is not a bad deal at all.

Remember that just 1.5 months or so ago, they raised with stock offerings at $3.25. That price looks so good today already doesn’t it? It sent the stock crashing down 18% or so from $4.6 to $3.5

EOSE and ASTS have done the same in the past and it’s stock price is at much higher price now. In the long term it will work out. Ofcourse it can linger for a few weeks or days.

Lastly, I must say that I am impressed by these announcement hitting perfectly at key levels. Last announcement at 1.6 fib levels. Previous raise was when the stock hit 200MA at $5.3 from below.

If you been following me on X, you wouldn’t have been suprised that Dilution hit and that it wasn’t priced in.

Techdrifts is now called Stockdrifts.
DISCLAIMER: Nothing on this post is FA. I own a big position in FLNC

For some odd reason, they call this earnings Q4 and end the FY2025 with this quarter. Wierdos!!

Contents:
- Deep Dive into Q4 Earnings report and FY 2025
- Multiple Revenue streams for BESS (Battery Energy Storage Systems)
- Parallels between BESS and hedge funds
- Key Performance Metrics for BESS

I’ve been saying it since late May/early June that FLNC will be a major AI beneficiary

If you are interested in reading the full thesis on FLNC, when I posted it first, check it out here. Literally got the bottom and avg. of $4.5.

They are finding ways to be FEOC compliant and be China proof. I predicted that it would happen and now you can see their slide from today.

There is massive institutional inflow. Ownership increased by 11% in the last quarter.

Okay, now lets look at the bad, ugly and good of what’s been announced. Stock’s up 13% AH.

• GAAP gross profit margin up 13.1% for fiscal year 2025 and approximately 13.7% for the fourth quarter, compared to 12.6% and 12.8% for the same periods in fiscal 2024, respectively. This improvement is really great to see. Meaning they are able to command better pricing.

• ARR up to $148M. Dissapointed with slowdown in ARR growth. Only 20% next year to $180M.

• Rev came at $1.0B vs $1.4B est, and lower end of $2.3B in 2025. But this year was already priced into the stock. It’s all about looking forward

This is the crucial part why the stock’s really excited. 30GWh Pipeline added for AI data centres in ±50 days.

This is so insane.

With a 40% pipeline jump in under 50 days, FY26 is hard to predict. I think they’re playing it safe by pointing to the 85% already secured by signed contracts.

Energy storage growth is explosive—they’re guiding to 50% growth, which is huge for an energy company. I suspect that’s conservative if the AI data centre build out continues.

Order intake hit a record $1.4B in Q4, the largest ever for Fluence, driving backlog to a record $5.3B, up 18% from last year.

- Q4 order intake $1.4B
- Contracted backlog grows to $5.3B, highest ever
- Contracted backlog covers ~85% of 2026 revenue guidance

If you missed it. Just read the %Change again. This is insane in my opinion. These are GW and GWh numbers, so no joke. Only TSLA had such high growth numbers for energy products earlier.

At the expense of being annoying, I predicted that Fluence will be the biggest beneficiary of European grid security. Europe has too much renewables and too much voltage fluctuations to take care of. Without even considering the AI data centres.

12 Days later, this announcement came

Not to anybody’s suprise, EU is leading the charge on Battery regulations.

On August 18, 2025, the EU Battery Regulation fully replaces the previous 2006 EU Battery Directive, marking a major legal shift in battery regulation within the EU market. What you need to understand that there is going to be more red tape and more insurance responsibility for Battery producers and developers.

This will make EU a high barrier of entry for smaller players due to compliance and risk increase. The fact that Fluence contracted the biggest battery in EU in this environment, is doubly impressive

Okay, so while we wait for the earnings call tomorrow, I have a couple of interesting technical insights that will help you make predictions like these.

Multiple Revenue streams for (BESS)

Now I would like to mention how these BESS batteries can have multiple revenue streams.

There are 7 ways a BESS projects can generate massive revenues and profits. I found a great table.

• Wholesale Energy Arbitrage

• Balancing Mechanism (BM)

• Ancillary / Frequency Services (e.g., DC, DR)

• Capacity Market

• Grid / Network Services

• Quick Reserve / Grid Stability Products

• Local Flexibility Markets

Depending on your geography, they could be growing, stable etc.

Optimizing Battery revenue is a lot like running a hedge fund.

Discharge = shorting power (sell now, expecting lower prices )
Charge = long power (buy now, expect prices to rise)

It’s all about timing the peaks and the troughs. This chart really drives home the point. Just look at the price swings during the day in Texas during peak summer load.

Recent ERCOT price spikes occur even during off-peak hours and weekends due to thermal outages, requiring sophisticated forecasting beyond traditional weather-based models

Interestingly, grid beyond is talking about Alpha generating trading algorithms, just like in the stock market.

Some more alpha strategies in CAISO

BESS operators face a dual challenge: maximize revenue while safeguarding safety and battery health.

• Failures and downtime hurt trading, so bid optimization must be integrated with OEM controls to protect longevity.

• Unlike solar or wind, every trading move accelerates or preserves battery degradation—demanding tight coordination between technical and commercial teams.

In many cases, the actual degradation of BESS is much higher than contractually agreed upon. Often BESS managers abuse the lifecycle of batteries for short term gain. However, it is a delicate balance in maximizing returns for asset owners.

More difficulty arises when battery don’t degrade at the same time. Imagine having to replace individual cells at different times. It’s a complete nightmare.

That said, Batteries can be operated to accept more degradation during high market price spikes while staying within warranty limits, optimizing returns.

Basic strategies like charging when the sun is out and discharging at sunset work but advanced optimization incorporating ancillary services and state-of-charge management can increase revenue by ~7% or more

Interestingly, Significant revenue leakage is not from technology failures but from operational inefficiencies. Operators can lose up to 20% of annual revenue through avoidable issues like data fragmentation and poor coordination.

Key Performance Metrics for (BESS)

These metrics collectively impact the performance, operational optimization, and profitability of battery energy storage systems. Understanding and accurately measuring them is essential for maximizing system value and longevity

1. State of Charge (SoC): Represents the estimated amount of energy currently stored in the battery, expressed as a percentage of its maximum capacity. It is a calculated parameter, not directly measured, can be very complex to determine accurately

   

   The long, flat voltage curve of LFP batteries makes it very difficult to determine SoC from voltage alone. A single voltage reading could correspond to a wide range of SoC values (e.g., 60% to 80%), leading to potential errors of 20-30% if relying solely on this method.
   
   It is common for a battery’s reported SoC to “jump” significantly after a calibration. An upward jump means missed revenue opportunities (the battery had more energy than thought), while a downward jump risks market non-compliance
   

   

   
   

2. State of Health (SoH): Indicates the maximum possible capacity of the battery relative to its original capacity at manufacturing. It reflects battery degradation over time, with a decrease in SoH meaning increased degradation
   

3. Cell Imbalance: Refers to differences in state of charge or capacity among individual cells within a battery module. Imbalance leads to stranded energy that cannot be accessed because the battery must stop discharging or charging when the weakest cell reaches its limit, impacting usable capacity

   
   

   

   

   
   

4. Round-trip Efficiency (RTE): Measures the ratio of energy output to energy input at the point of interconnect, reflecting how efficiently the battery charges and discharges. High RTE is desirable for maximizing energy and financial returns
   

5. Real-time Usable Energy: The actual amount of energy that can be discharged by the BESS at any moment, considering SoC, SoH, and cell imbalance. This metric is crucial for operational decisions and trading activities.

FLNC’s software stack is increasingly important to understand key BESS parameters and maximize performance

I might decide to do a deep dive Part 3 on FLNC after the earnings call tomorrow and incorporate the insights from the call. Let me know if you like me to get more technical on batteries or would you prefer I do more financial analysis.

TLDR:

• Polisilicon prices have bounced off it’s lows from June 2025. competetion in the TopCon tech has eased up a little bit in this quarter

• There might be a suprise announcement about SPAC merger with ALUSSA 2 for the remaining funding of G2 phase 1

• TE’s Topcon Tech vs, FSLR CdTE. The only 2 players in quest for Vertically integrated US solar production

HOUSEKEEPING:

I am keeping all my research free, so its not so nice to take these insights and paywall them in your own substack. It takes substantial amount of time to produce these.

Moving onwards, I am considering paywalling some content to deter this behavior

If you haven’t already, I suggest you start with Deep Dive 1 and Deep Dive 2

INSIGHT 1

There is a lot of fear that the 4 Chinese big-dawgs are selling Solar panels at a loss in H1 2025

News outlets including Reuters, Bloomberg, and PV Magazine reported that losses across the five biggest producers (including LONGi, Jinko, Trina, JA Solar, Tongwei) for H1 2025 ranged from CNY13 billion to CNY17.3 billion ($1.8–2.4 billion)

Notice the GW module shipments. And here we are, getting excited about 10GW in 2 years time frame for TE. The Chinese are playing a completely different ballgame.

Interestingly only $CSIQ and DMEGC made profits. This is obviously very scary fact.

- How will $TE be profitable then?
- How can they do what all the other Chinese mass producers couldn’t do?

To answer these questions, lets look at the reasons as to why this oversupply happened?

• Destructive price war: Cutthroat “involution” slashed prices to win share, pushing margins below zero.

• Tariffs bite: U.S. trade barriers piled on costs and choked some export routes, squeezing margins further.

• Demand dip at home: China’s 2025 subsidy reforms cooled installations, pressuring prices and volumes.

• Tech transition: Old PERC lines winding down, new high-efficiency lines not fully online—efficiency and capacity in limbo.

• Policy whiplash ahead: China’s planned end to the 13% VAT export rebate by Q4 2025 yanks a key price support, forcing repricing and investment resets.

This tells me that there were a lot of factors that contributed and the Chinese industry will consolidate soon enough and prices will rebound.

Well well, lets check out the prices of Polisilicon, an important material in Solar wafers.

There are two trends that plays in our favor. Polysilicon in China seems to have bottomed from July and has rebounded already 50-70% This means there is more demand and stabilization.

Secondly, Polysilicon outside china (which is 5X more costly) is slowly reducing. Both these trends combined, TE might have a little favorable time this quarter.

I love that it literally bottomed July, so that might not be appreciated by the markets yet. for this quarter.

INSIGHT 2

Remember Alussa I Energy corp in 2021? FREYR raised SPAC money ($850M) with them. Mr. Barcelo was the man behind it.

Now after 4 years, they filed a new SEC filings today. Literally half a day before earnings announcement tomorrow.

Daniel Barcelo, the current CEO of TE/T1 is the also the head of Alussa Energy. The timing is super suspicious.

The two filings from 2019 and 2025 are EXACTLY the same.

Original SEC filing from 2019: https://sec.gov/Archives/edgar/data/1781115/000121390019024877/f424b4112619_alussaenergy.htm

Now compare it to 2025 filings: https://sec.gov/ix?doc=/Archives/edgar/data/2041493/000121390025102171/ea0217506-21.htm

Exactly identical. It’s like as if they didn’t want to pay anymore lawyer fees because they know they were going to do the same thing again. This delay of 6Nov to 14Nov 2024 could be to accommodate the IPO filing of ALUSSA II.

MEANING: ALUSSA II might provide the remaining funding of $250M for G2 phase 1 for very little dilution.

Also interesting to note, Encompass Capital Advisors Holdings are the SPAC sponsors of Allusa. IF you look at their 13F filing. They are heavily focused on Energy and industrials.

The last spike in number of shares is due to the capital raise of $72M and $50M convertible debt about 3 weeks ago.

I spent more than an hour trying to understand how much shares do they own in total. I am still confused because it says in the SEC documents they own 74M for 19% stake. Also 61M number is thrown around.

But that would mean an outstanding shares count of 350M. This is not the case right now. So maybe someone smarter than me can clarify.

Here is the link to the most recent 13D filing

https://www.sec.gov/Archives/edgar/data/1541901/000092189525002886/0000921895-25-002886-index.html

In any case, Encompass has substantial interest in $TE’s success.

This insight was originally found by Ringdom. An awesome analyst on X. Super sharp. Give him a follow for alpha

INSIGHT 3

• Geographic variation matters significantly – TOPCon wins in most scenarios, but CdTe retains advantages in extreme heat + humidity environments and has better degradation profiles​

• LCOE is highly location and assumption-dependent – the same technology can have vastly different costs depending on financing, labor, land costs, and system design

Before we start, lets look at the 45X details. These incentives are amazing for a vertically integrated player.

$FSLR has absolutely crushed it this year with margins and shareholders return. Their EBIT margins stand at 30%. Woat???Ask the chinese how they feel about that.

First Solar (FSLR) has taken full advantage of the credits. Over the past two years, they’ve generated more than $2.1 billion in net profit—coincidentally, about the same amount they received from the 45X subsidy.

Given this, and after a closer look at the margin drivers, I’m hesitant to draw firm conclusions about margin comparisons at this stage. (Opinion changed from deep dive 2)

Their tech does obviate many steps and makes it easier to manufacture. So that’s kinda cool.

However, an important point to remember is that they are not as efficient as T1/TE’s tech.

Efficiency Records:

• Trina Solar(TE) TOPCon: 27.8% (record set Nov 2024), 23% in field​

• First Solar Series 7: 22.3% efficiency, 19% in reality​

• Efficiency gap: ~4.3 percentage points in TOPCon’s favor

The advantage is about 4% in real world. Now this spread across it’s entire lifetime means considerably less electrons generated per acre

On the flip side, one advantage that FSLR’s Cdte has is that it is more reliable for extremely hotter temperatures. TE’s Topcon can do bifacial capture meaning it can gather energy from light that is reflected from ground.

The Reality is that LCOE calculations are dependent on locations and each tech has their own strengths. I asked Claude to do some calculations in different locations where they these type of plants exists. see it in APPENDIX

Conclusion: Topcon for most cases are more suitable because of higher efficiency. There is a reason why 70-80% of the Solar market in the world is Topcon. CdTe tech is 1-3% market share currently.

P.S. I am writing so many posts at the same time (slow burners). I don’t want to sacrifice on the quality, so they are trickling slowly.

APPENDIX

Example A

Phoenix Result: TOPCon $42.6/MWh vs. CdTe $49.7/MWh Winner: TOPCon (-14% LCOE advantage) Key Driver: Bifacial gain (+10.7%) outweighs CdTe’s temperature advantage

Note: If using unsustainable $0.096/W TOPCon pricing → LCOE drops to $0.034/kWh, creating a -32% gap. This is artificially distorted by Chinese overproduction.

Example B

Seville Result: TOPCon $48.0/MWh vs. CdTe $53.8/MWh Winner: TOPCon (-11% LCOE advantage) Key Driver: Tracker + bifacial synergy (+30% energy vs. fixed monofacial)

EXAMPLE C

Hamburg Result: TOPCon $63.3/MWh vs. CdTe $87.0/MWh Winner: TOPCon (-27% LCOE advantage) Key Driver: CdTe’s low efficiency cripples performance in low-irradiance climates; TOPCon’s bifacial gain + higher efficiency dominate.

EXAMPLE D

Ohio Result: TOPCon $47.5/MWh vs. CdTe $56.6/MWh Winner: TOPCon (-16% LCOE advantage) Note: IF domestic content bonus (+10% ITC) applies, CdTe gains additional $0.08/W advantage → LCOE drops to $51.2/MWh (still loses to TOPCon).

DISCLAIMER: I initiated little at $14.7 to start tracking. You might get opportunity to buy again at $12-13. Earnings was a big miss.
Don’t ever FOMO. None of this is FA

TLDR:

Alberta’s “Texas of Canada” policy + TransAlta’s 9GW (bonkers) of energized power and sites are the fastest path to scale AI data centers, with material re-rating potential

It’s all about the power rush right now.

THESIS:

• Tailwind: Alberta state is the Texas of Canada. It wants to attract $100B in AI data centre investment. For perspective the population is only 4M

• They control 9,000 MW of gross installed capacity across 88 facilities in Canada, the United States, and Australia. Delievered $1.25 EBITDA in 2024. So not the usual no revenue shit Crypto miner play.

• They have so much excess capacity behind the grid.The quickest in my opinion in the North American subscontinent

• They also sit close to major fibre routes, which means data can move quickly to the US

• They already own >52,000 acres of land tied into the grid. Quick data centre greenlighting

• Canada hasn’t seen crazy AI data centre boom yet. Their government has already said, if you wanna build a data centre here, you bring your own energy. (Only now has the US started to do this with its policy posturing)

A quick Technical Analysis (TA) check, whether it’s worth our time to research.

1. Perfect retest of an important level.

2. Rising volume meaning high interest

3. Price staying above 10EMA
   
   Simplicity in life and in TA is what makes things Mama Mia!!
   

   

Okay so why did the stock crash 20% on Q3 earnings yesterday?

TransAlta (TAC) missed Q3 2025 earnings estimates by a wide margin, posting an EPS of -$0.01 to -$0.02 versus the $0.12 consensus and revenue of $615 million versus $849 million expected. The shortfall was due to lower Alberta power prices and unfavorable market conditions, despite strong operational availability and hedging efforts.


Also,the timelines around the Alberta data center are uncertain, which rattled investors. Investor Day is now pushed to Q1 2026, and a partner’s filing pointed to a 2030 “in-service date” that to spooked the market.

Management called 2030 a “worst-case” scenario on the call. They and their customer are still targeting 2027–2028, while the partner works through transmission bottleneck analysis.

The company is progressing toward a final MOU with partners, which suggests this is normal project work rather than a red flag.

Market over reacted. Relax guys. Work’s in progress

Okay seems worth our time, lets get deeper. Here’s a snapshot of all the assets in it’s fleet.

INSIGHT #1

There are multiple benefits of being in Alberta as an Energy company trying to build AI data centers

a. Cold air that significantly reduces your cooling costs. Estimated to be about 5-15% total costs.

b. Alberta’s AI/HPC Ambitions: The Alberta’s government aims to attract $100 billion in AI-driven data center investment over the next 5 years

c. Alberta is the TEXAS of Canada with policies favoring behind the meter generation. About two years ago, Alberta’s Premier Danielle Smith stated that data center developers must sign deals with generating companies to “bring your own electricity, bring your own generation.”

This article was 2 years ago!!!

According to an article in DCD published on June 05 this year:

The Alberta Electric System Operator (AESO) has launched an interim approach to large load connections. The new approach is expected to enable the connection of up to 1.2GW of large load projects, including data centers, between now and 2028.

AESO (Grid operator at Alberta) is seeing unprecedented Power requests with backlogs for years. Look at how it outstripping generation requests

TransAlta’s Competitive Advantages:

1. Behind-the-Meter Model: Offers 90% dedicated power directly from generation assets, bypassing transmission constraints

2. Turnkey Sites: Sundance 6 (401 MW) and other sites prepared with water, land, fiber, and substation infrastructure

This is such an Important point. It’s not the GPU’s that are the bottlenecks. It’s the powered shell, or real estate that is powered that is in extreme shortage.

Here is Corweave’s CEO Michael Intrator, talking about how Power shells/infrastrcuture to set up AI data centre is the biggest bottle neck in Q2 2025

At the end of the day, right now, it’s the powered shells that are the choke point that is causing the struggle to get enough infrastructure online for the demand signals that we are seeing, not just within our company. It’s the massive demand signals that you’re seeing across the industry.

He felt the need to reiterate

It is a market that is really working hard to try and balance it is it there are fundamental components at the powered shell, at the power in terms of the electrons moving through the grid, at the supply chains that exist within the GPUs the supply chains that exist within the mid-voltage transformers. There’s a lot of different pieces that are constrained. But, ultimately, the piece that is the most significant challenge right now is accessing powered shells that are capable of delivering the scale of infrastructure that our clients are requiring.

3. Flexible Supply: Can offer mix of baseload (hydro, cogen) + peaker backup for 99.999% uptime. They are a mammoth and have a lot of flexibility

4. Speed to Market: 18-24 months vs. 36-48 months for greenfield builds

INSIGHT #2

As you know, I love to do conservative napkin math.

TransAlta’s rough valuation is at $492k per MW of energized power , based on its $4.43 billion market cap and 9,000 MW capacity. Even if you take EV value ($8B) it’s still comfortably below <0.9M/MW

This is significantly undervalued.

Lets take some of their famous peers:

Neoclouds often hybridize, starting with co-location to fund expansions into full builds, driven by AI demand. Figures can vary by contract length (3–15 years), power efficiency (e.g., 4–7¢/kWh), and utilization (50–100%)

In a co-location setup, Neo Cloud may simply rent space in a data center, bring its own GPUs, and manage them independently.

Of course there are a lot nuances to think about. For eg. they obviously lack the expertise to build AI data centers themselves.

Here are some important aspects of contracts.

As an investor in $TAC, you gotta have some premonition about the following questions

• What would their partnerships look like?

• Are they going to provide co location and ONLY sell the powered shells?

• How are they going to bridge for new power since AESO claims they will only allow 1200MW until 2028, even though there is 16GW waiting for interconnection?

Look at how co location deals for their companies and their partners are valued at. I asked Perplexity to compile this data.

Now $TAC are exploring to convert their power into high margin data centre hubs. Contracts are still up in the air, but For sure, margins in the future will be much higher than it is now.

INSIGHT #3

I am not a modeling expert, but here are some of my estimations for what the business could be worth.

Economic Model:

• Land Lease: $50,000 - $100,000/acre/year (200-400 acres typical = $10-40M/yr)

• Power Lease: $1.0 - $1.5M per MW per year (equivalent to $114-171/MWh at 100% utilization)

• Ancillary Revenue: ±$85M, Grid balancing, backup power services

• Total Revenue Potential: $1.2 - $1.8B annually for 1,200 MW allocation

• Incremental EBITDA: $900M - $1.35B (assuming 75% EBITDA margin)

• Implied EBITDA Multiple:
  If valued at 10x (conservative) =$9-13.5B incremental EV
  
  $9-13B just for the Data centre business.
  
  

  

  It’s obviously an over assumption that they will get 1200MW of new power. But I wonder how much can they repurpose without needing AESO’s permissions. They have filed for 650MW of new power for Datacentres as of 6/25. So assumed 1200MW total plus or minus.
  

Timeline & Probability:

• Q4 2025: AESO executes Demand Transmission Service contracts (expected mid-September completion, now delayed to Q4

• Q1 2026: TransAlta signs definitive agreements with 1-2 hyperscalers (Stock crashed after earnings Q3 because this got delayed. I would say it will happen with 60% probability

• Q2-Q3 2026: Site prep and infrastructure build begins

• 2027-2028: First data centers operational (400-600 MW initially)

• 2028-2029: Full 1,200 MW build-out (if successful)

Probability-Weighted Value:

• 60% chance of securing 600 MW @ $1.2M/MW/yr = $720M annual revenue × 10x multiple = $7.2B EV × 60% = $4.3B

• 30% chance of securing 300 MW @ $1.0M/MW/yr = $300M annual revenue × 10x multiple = $3.0B EV × 30% = $0.9B

• 10% chance of no deal = $0

• Expected Value: $5.2B incremental EV so atleast a double from here

  
  

INSIGHT #4

Market is mispricing that AESO (Grid operator in Alberta, the ERCOT of Alberta) will provide much higher interconnection with the help of Batteries and reduce the grid instability during peak hours.

Alberta has about 16GW waiting interconnection approval. AESO came out and said they are planning for only 1.2GW until 2028

Woah? Boooring!!

Grid stability is their major concern.

But this will be rethought. Market’s skeptical but they will be able to get in much more GW in. This is not getting priced in that AESO will end up approving much more GW of interconnection.

This is how

• Stick a battery on the data-centre: charge it off-peak, dump it back during the evening rush, and your “new” load barely shows up on the grid’s peak meter.

• It’s plug-and-play: $TSLA Megablocks or $FLNC, or $EOSE are factory-built, already running near Grande Prairie, and can be on-site earning money well before the 2027-28 crunch.

Regulation hack: AESO’s 1,200 MW cap targets peak demand; if you bring your own battery and commit to peak-shaving/DR, you skate around that limit and unlock extra revenue streams.

For years they have had more generation than load. But this flipped a few a quarters ago. This is unprecedented.

They have a 400MW battery in interconnection study right now.
Batteries will have a big role to play to service all of that load requests

This source offers a great deal of information about AESO’s thinking. Just look at how Load demand is overtaking gen. This is insane.

INSIGHT #5

Their track record and Management execution is on point.

Speed to energize is comparable to a Bitcoin miner. Which in amazing for a traditional Utility company. Don’t forget about the Sundance 6 site which might be even online in 18 months.

Transalta seems like they do what they say they will do. Just Look at their track record

I don’t got no trust issues when I can see your past.
( I like wrong grammar. so forgive me)

These are historical Alberta’s Pool Prices:

• 2019-2020: $55/MWh (oversupply from new wind)

• 2021: $95/MWh (coal retirements, demand recovery)

• 2022: $135/MWh (natural gas price spike, Ukraine war)

• 2023: $88/MWh (normalization)

• 2024: $82/MWh (continued stability)

• 2025E: $75-85/MWh (new Heartland supply pressure)

EBITDA has shown some Sensitivity in the past to Alberta Pool Prices (based on 4,000 MW merchant exposure)

Still I am very surprised and happy to see that they squeezed much more $ revenue per MWh than the pool pricing in Alberta. They do this by employing a number of hedging strategies.

Even though, they missed expectations bigly, this gives me a signal that the management at TAC are A class and they know their energy business well. I mean, just look at this slide

Lets check further about the management.

Board & Executive Relationships:

• John Kousinioris (CEO): Former ATCO executive; well-connected to Alberta business elite. Look at that sauve look!!!

EDIT: John will be handing over the reigns to the CFO. Which makes sense since the company has a lot of financial engineering ahead of them. The stock might have not liked that, this quarter was a big miss too.

• Board Member: Lynn Ewing: Former AESO executive (2010-2018); direct insight into grid planning and maybe some influence too

• Board Member: Martha Wyrsch: Former VP at Suncor; oil & gas ties critical for Cogen partnerships

  Lastly, also valuable to look at their debt structure to know if they will have a cash crunch soon. This is important because AESO is very touchy about providing approvals to non stable developers. They have been multiple eg. of projects started but then abandoned.

Total Debt (as of Dec 31, 2024): $3.31 billion,
Current Cash: $162.8 million

This looks honestly not so bad. Clear and manageable.

RISKS

1. They are not able to make an agreement with a hyperscalers

2. META completes the deal with Capital Power (CPX). The market seems to think so.

3. They get blocked politically to open data centres. I found a reddit post of albertans super unhappy about Meta’s potential deal to place a data centre there

4. Red taping in Canada could be significant effort and it largely depends on the administration’s desire to will data centres in their jurisdictions. So wild card. Check the approval processes involved

5. Edmonton-based Capital Power also has two massive proposals in the AESO connection lineup, representing 1,500 megawatts of load for potential data centres to be co-located at its Genesee Generating Station.

   
   

Upside Case: If TransAlta converts 1GW to data center hosting at $100K/MW/month = $1.2B annual revenue vs. current $750M from same capacity = +$600M incremental revenue = ±$500M incremental EBITDA (75% margin).

This is at the core of the thesis for EBITDA almost doubling in a year.

REMARKS:

I didn’t go all in on this because I like to chase higher risk as well. But otherwise I think it’s a solid play with good returns especially after this 30-40% crash.

APPENDIX:

AESO Interconnection Process: Alberta moved to a cluster study approach in 2024 to manage queue backlog.

TransAlta’s Queue Positions:

• Cluster 1 (2022-2023): All projects withdrawn or completed (100% success for completed)

• Cluster 2 (2023-2024): 2 projects totaling 150 MW (battery) - 1 completed (100 MW), 1 in construction (50 MW, COD Q2 2026)

• Cluster 3 (2024-2025): 1 project, 400 MW battery storage (SASR submitted; system impact study Q1 2026)

Required Upgrades for Cluster 3 Battery Project. All of this is doable pretty well. Some more Napkin maths on the interconnection costs

• Transmission: $45M (new 240 kV line, 12 km) - 50% allocated to TransAlta, 50% to other cluster participants

• Substation: $18M (new breakers, transformers) - 100% TransAlta

• Total Interconnection Cost: $40M or $100,000/MW (typical for Alberta solar/battery)

• Fee Deposits Paid: $8M (Q3 2024)

If you have not read my part 1 on $TE, I would highly recommend you to read that Part 1. Here is the link

Part 2 explores further the bullish and the bearish cases. I cite certain accountings and important insights from the SEC filing.

This article is quite long since I found a lot of things and writing helps me build conviction on my investments. Hopefully it does the same for you.

TLDR:

T1/TE looks undervalued with a meaningful upside ahead, but its headline market cap of $700M is misleading. On a fully diluted basis, the market cap is roughly 50% higher, around $1.1B. Warrants and SBC might not convert so this is the most conservative calculations.

Also, most analyses seem to miss the potential value of the PowerShell Giga Arctic project. If that asset is realized, it could add several hundred million dollars beyond whats currently reflected.

Content:

- Bullish Points
- Bearish/Risk Reasonings
- Deep Dive into the history of T1/TE
- Major Thesis. Securing US domestic supply chain and Investments into Talon PV.
- Margins at FSLR. Why is it so good?

Appendix (For Paid subscribers for savvy details)
- Details of the debt and equity raise in detail. Full dissection.
- Insights about the Giga Arctic that could get purposed into an AI data centre.
- Details about the Chinese Solar companies profits.

In the last final PART 3, I will dive deeper into the PV tech. Topcon tech.

Now lets get into each of these points

Bullish:

1. Even with Trump administration’s anti renewable back drop, US is securing it’s energy independence. Solar has a big part to play here FSLR($25B) and TE($700M) are the ONLY pure play US Solar stocks in the block.

Since the passing of the IRA in 2022, Domestic annual nameplate capacity for Solar modules increased by over eightfold 800%

Joe Biden’s policies secured the path to America’s solar energy independence. Trump is continuing to do so by leaving the 45X incentives intact.

Now ramping up just the modules is not enough. The entire supply chain’s gotta be upgraded so that any weak links can be eliminated in the path to pure independence.

This is how the supply chain looks like:

Polysilicon → Wafers → Solar Cells → Modules → Installation

TE’s G2 factory is going to tackle cell production. It is trying to be vertically integrated so that it can stack it’s subsidies and control it’s own destiny.

US is STROOONG and independent!!

2. T1’s roadmap targets 70%+ U.S. content modules by 2027 (Hitting the 45X requirements) and $650M-$700M EBITDA run rate with fully operational G1 and G2 plant(Phase 1). This is an insane guidance in 1-2 years.

In the company’s Q1 2025 earnings release, T1 explicitly stated:​

“There are no changes to T1’s projected $650 - $700 million annual run-rate EBITDA estimate based on optimized production at G1 Dallas and G2 Austin.”

T1 reiterated the same target in its Q2 2025 (August 20, 2025) results:​

“There are no changes to T1’s projected $650 - $700 million annual run-rate EBITDA estimate...”

Most recently, in the October 22, 2025 preliminary Q3 results announcement, T1 maintained this guidance:​

“There are no changes to T1’s projected $650 - $700 million annual run-rate EBITDA estimate based on 5 GW of annual optimized and integrated production at G1_Dallas and G2_Austin.”

Gotten a little repetitive, but WE the vigilantes will hold the management accountable.

15-20X guidance on EBITDA in 2 years? Am I reading this right? If anyone thinks differently please let me know.

3. 2025 production sold out at low end of production at 2.6 GW. Remember: For every 1GW they sell, they get roughly 100-150M in credits

This means there is a lot of demand for their modules in light of 45X OBBB requirements.

Here they said this.

T1 generated total net sales for the third quarter of $200 - $210 million on modules sales of approximately 725 MW.
and has accrued $92 million of Section 45X credits year-to-date that the Company expects to monetize.

Also after reading this, my expectations are that H2 2025 should bring in significant revenue.

T1 expects a significant increase in sales related to the highest expected production year-to-date at G1_Dallas

4. No one is talking about this bullish point. Key important pumpamental factor

4. 
   They got an empty Power shell called Giga Arctic in Norway. A 120,000 square meter battery gigafactory in Mo i Rana, Norway that was formally commissioned in 2022. But it can be now pivoted into a AI Data centre facility
   
   Construction began on a planned 29 GWh battery cell factory, with FREYR sanctioning the project by June 2022 and securing over $1.6 billion in debt financing, including €400 million in loan guarantees from Norway’s export credit agency Eksfin.

   

   This image is from Sep 2022.
   
   Before you get too bullish, let me show you a quote from Corweave’s CEO Michael Intrator, talking about how Power shells/infrastrcuture to set up AI data centre is the biggest bottle neck in the previous earnings

   

   
   At the end of the day, right now, it’s the powered shells that are the choke point that is causing the struggle to get enough infrastructure online for the demand signals that we are seeing, not just within our company. It’s the massive demand signals that you’re seeing across the industry.

   
   He felt the need to reiterate

   
   It is a market that is really working hard to try and balance it is it there are fundamental components at the powered shell, at the power in terms of the electrons moving through the grid, at the supply chains that exist within the GPUs the supply chains that exist within the mid-voltage transformers. There’s a lot of different pieces that are constrained. But, ultimately, the piece that is the most significant challenge right now is accessing powered shells that are capable of delivering the scale of infrastructure that our clients are requiring.

   
   So, here is a Linkedin post from their official company account talking about the pivot
   https://www.linkedin.com/posts/t1energy_here-in-the-united-states-the-future-is-activity-7357035159608545282-ypZe/
   

   

   Although they are alleging that Statnett took away their promised 60MW power and that they only just need to turn the switches on. The factory is located right next to the 500MW hydro dam.

   

   There is still a few issues to consider which I discuss in Apendix.
   

5. Technically the stock looks like they almost bottomed

Bearish/Risks

1. Section 232 polysilicon tariffs could be clarified by ‘end of the month. This is unclarity for their current supply chain.

The proposal came from the Coalition for a Prosperous America (CPA), which suggested $0.20/W tariffs on solar modules, $0.10/W on cells, $0.07/W on wafers and $10/kg on polysilicon imports and a TRQ threshold of 30GW for silicon cells, wafers and ingots and 40,000MT of polysilicon from “trusted allied countries with secure, non-Chinese-controlled supply chains.”

Even though some argued that the tariffs are too high, it remains to be seen what the actual numbers turn out to be.

FSLR is completely immune to this, since they are producing cadmium telluride (CdTe) thin-film modules.

2. The Enterprise value and dilution is a lot higher than everyone thought or expects. They might need to raise a lot more in funding. G2 phase 1 would need $400M in investments. They have secured about $125M in last weeks raise. I will deep dive into it in the appendix

3. Margins are non existent. It doesn’t help that the top Chinese PV producers have been selling their products for losses in the last year. In comparison, FSLR’s margin is stupendously good.

Bears argue that first Solar manufactures high-margin CdTe modules with proprietary technology. T1 manufactures commodity TOPCon crystalline silicon modules—a technology not proprietary to T1 and competing against global manufacturers

4. Leading Chinese companies have made only losses in 2025. There might be a glut in the industry

The Complex History of FREYR/T1 Energy Saga:
A story of Pivots and Abandoned Plans

The relationship between FREYR Battery (now T1 Energy) and its European operations represents one of the most dramatic strategic pivots in history of pivots

The Original Vision: Norwegian Battery Champion (2018-2022)

FREYR Battery was founded in 2018 with ambitious plans to become Europe’s battery manufacturing powerhouse, using 24M Technologies’ SemiSolid battery tech. Adrenalin was high and they raised $704M via SPAC. Yes Yes the same ponzi Chamath started.

The main show was Giga Arctic in Mo i Rana, Norway. A planned 29 GWh battery cell factory with an estimated capital cost of $1.7 billion. By June 2022, FREYR had sanctioned construction with identified debt financing of over $1.6 billion, including €400 million in loan guarantees from Norway’s export credit agency Eksfin.


However, just like many dead cleantech projects back then, it wasn’t easy to raise money and then make money in this space

2023: First Cracks Appear

In November 2023, FREYR paused construction of Giga Arctic, citing the need for European government subsidies that remained out of reach and deciding to focus on the US market instead. The company had spent approximately $64 million in Q1 2023 alone, with 75% allocated to Giga Arctic.​

In August 2024, CEO Tom Einar Jensen told investors that cheap Chinese battery oversupply made it impossible to raise money for battery manufacturing. The company announced it would pursue “less capital intensive opportunities” including module and pack assembly rather than cell production.​

By September 2024, FREYR formally paused battery cell production in Norway and fired 700 workers from Giga Arctic alone

They tried to bring 24M MIT tech into production and failed. Realizing that its wayy too expensive and gave up.

A Big Pivot in Motion

By early 2025, FREYR underwent a dramatic transformation. Almost the entire management team was replaced, and the company reemerged under a new name — T1 Energy.

November 2024: The Solar Acquisition
FREYR completed acquisition of Trina Solar’s 5 GW solar module factory in Wilmer, Texas for $340 million. The facility employs over 1,000 people and began production in November 2024.

February 2025: Complete Transformation

• Relocated: Headquarters moved from Norway to Austin, Texas

• Strategic Reset: Cancelled the planned $2.6 billion Giga America battery factory in Georgia

• Asset Sale: Sold the 368-acre Georgia site for $50 million, netting about $22.5 million

• Reality Check: Despite the big 2022 announcement, construction never started

Current Status: What Remains in Europe?

Norway Operations:
T1 Energy still maintains a presence in Mo i Rana with the CQP facility, but production has been paused. The company is exploring converting Giga Arctic into a data center or AI hub rather than battery production. T1 had already invested a lot in Giga Arctic infrastructure plus the CQP facility. which they have written down by $312.9M in their SEC filings.

Finland CAM Project:
This is where the situation gets murky. While FREYR received the €122 million EU grant selection in October 2024, several critical factors remain:

1. Not Finalized: Selection was in Oct 2024; but still no confirmation by mid-2025).

2. Conditional Nature: Grant is conditional on partners/criteria.

3. Company Focus Shifted: Company shifted focus to U.S.; exploring options for EU assets.

4. Uncertain Partnership Status: Partnerships unclear: Aleees JV fizzled; FMG status uncertain.

5. Competing Project: FMG pursuing €774M CAM JV with Easpring (groundbreaking Apr 2025)

Can They Proceed Individually?

The critical question is whether FREYR/T1 can move forward with the Finland CAM project given their complete pivot to US solar manufacturing:

The €122 million EU grant likely requires demonstrating a committed partner structure and feasibility—difficult for a company that has abandoned battery operations

Theoretical Possibility:
The Previous CEO still has his job as European Battery operations and is not fired.

Also this Linkedin post seem to suggest that they have not fully given up on their Battery ambitions.

Securing the Supply Chain

TE/T1 Energy invested minority stake into Talon PV. I am almost fully quoting Will norman’s article here

Both T1 Energy and Talon PV are building solar cell manufacturing facilities in Texas. T1 Energy is building a 5GW site in Rockdale near Austin, and Talon PV is building a 4.8GW plant in Baytown.

T1 Energy did not disclose the stake it has acquired or the value of its investment. The companies have entered into a Simple Agreement for Future Equity (SAFE) and “remain in discussions to potentially deepen the strategic relationship in the future”.

“We are excited by the opportunity to invest in another American solar energy project that complements our G2_Austin development,” said Daniel Barcelo, T1 Energy’s CEO and chairman of the board. “Expanding the output of American solar cells is necessary to support energy security, achieve regulatory compliance and build an American solar industry based on advanced manufacturing and cutting-edge solar technology.”

However, Talon PV has acquired a TOPCon IP portfolio from cadmium telluride (CdTe) module manufacturer First Solar, which it told PV Tech Premium would clear the way for commercial TOPCon cell production in the US. It is unconfirmed whether the two companies plan to share intellectual property after the investment.

Also the deal with Corning to source US-made polysilicon and silicon wafers is significant They are big and talked about T1 in their earnings call quite a bit. Combined with its planned cell and module production facilities, the deal would see a fully US-made silicon solar supply chain realised and would mark a significant moment in the country’s PV manufacturing industry.

Margins at FSLR

Quite INSANE margins for a Solar producer. That too in the US.

FSLR’s manufacturing process is fully vertically integrated. Modules are manufactured from sheets of glass to a finished module, under one roof, in about 4.5 hours. Typical crystalline silicon modules are manufactured using the batch processing technique. One module can take up to three days for production, in pieces, across up to four factories (sometimes in different countries and/or continents) with less stringent oversight of the manufacturing process.

They clearly has fewer steps. However, their tech is not as efficient. Meaning they miss out about 4-5%. For big solar farms, this can amount to significant numbers.

I will be doing a deep dive into the PV tech in PART 3.

Okay so I recently discovered that they are planning to dilute shares further and raise the upper limit to 500M this shareholders vote. Also found certain bits of information quite appaling.

Secondly, I also dove deeper into the valuation of Giga Arctic.

APPENDIX

I posted about asking the shareholders for 50% further dilution on X.

Company asked to have the capability to dilute if the need arises from 355M to 500M. Asking for permission to have the authority to dilute by 50% Vote on 2nd of Dec 2025

It’s understandable ofcourse, but annoying none the less. They need to raise $800M total for G2 full facility. They have now secured 120M for phase 1 that requires $400M

You can see that cash is dwindling

But the working capital is good, and they will be entitled to around $200M worth of credit soon. So why ask for this much dilution already?

Okay lets look at the capital raise, they just did

Here is the Capital Raise Summary

• $72 million: Raised through a registered direct equity offering (selling common stock to institutional investors).

• $50 million: Expected to be raised via a convertible preferred stock issuance (a type of preferred stock that can convert into common stock), announced on October 22, 2025.

• Total capital expected soon: Approximately $122 million ($72M + $50M) to fund the first 2.1 GW phase of the G2_Austin project.

Here they are giving a little speech about how their book value is negative and that even after paying $3.25 per share, they are worth nothing immediately after investments.

After the Direct offering raise on 23 Oct, the total outstanding shares are 190M.

This is what I thought. However, there are about 98M shares more that are already given out. This is why they had to raise the ceiling so much sooner. So I would argue that total Fully diluted outstanding shares is already around ±300M.

Here they are clearly reiterating how they have pretty much exhausted the 295M of allowed outstanding shares. 190M already outstanding and they point the gun at us reinforcing about what they can actually still do. Lets account for that extra 98M shares

Warrants: 25M (These are NOT so concerning because stock gotta 3X before they are exercisable)

Stock based compensation: 7.8M + 6.6M + 28.2M = 42.6M

Not everything should be considered as outstanding since we don’t know what strike prices, however, it isimportant to consider them as the worst case.

I Asked Perplexity to make a table for conditions under which these extra shares can become common shares.

Okay now on the topic of Norwegian Assets. Here are the two facilities they built

• Giga Arctic (planned 120,000 m², Mo i Rana)

• Customer Qualification Plant/CQP (13,000 m², Mo i Rana)

This image is from Sep 2022 of Giga Arctic.

I looked at their 10K from 2024 to understand how the company is valuing their Norwegian facilities.

• The Giga Arctic and CQP facilities in Norway were written down by $312.9 million in 2024.

I am willing to bet that if they are able to successfully curry favor (restore power) with the Norwegian politicians, they might be able to sell it for higher price than just $43M in assets.

Just look at it’s location. It is right next to a lot of Hydro power stations. Getting access to power to this facility is just a matter of Turning it on and having the right rapport with the government.

Norway has an acute shortage of Data centre power. Just like in the US. Even though they are energy rich, there are grid interconnections issues as well.

Breaking it down: Statnett (Norway’s grid operator) has reserved 17 TWh for projects with secured capacity, while another 27 TWh sits in the queue, waiting for grid availability that simply doesn’t exist. Meanwhile, 14 new data centers are under construction

Giga Arctic has lost its interconnection queue for power. I dont think that it is right for TE/T1 energy to claim any wrongdoing by Stattnett (The Norwegian public utility company. They were the ones who pulled out and eagerly ran into the arms of Papa America for 45X. However, this is a bet that I am willing to make that they might find the partner and will get immediate access to power for the shell infrastructure.

Also recently, Stargate Norway and Aker Nscale announced 250MW facilities. These companies might be prime to form partnerships and lease/sell these power shells.

LASTLY,

Leading Chinese solar companies have faced significant profitability challenges in 2025, with major firms like LONGi, Jinko Solar, Trina Solar, and JA Solar reporting billions of dollars in combined net losses in the first half of the year.

China’s focus on solar as a strategic industry drives continued capacity buildout and subsidized expansion, leading to oversupply and thin margins intentionally or unintentionally.

The companies primarily use advanced crystalline silicon technology, focusing on large, high-efficiency wafers and cells, especially leveraging TOPCon (Tunnel Oxide Passivated Contact) technology, which dominates around 70-80% of the solar cell market share in 2025.


Still one should remember that Americans dont wanna be blue balled like the germans!!

Also, I love that there was a lot of reiterations of the 600-750M of EBITDA and that their capacity is sold out, but it all depends on how trump’s administration agrees on tarriffs for Chinese solar panels.

The long trend of de globalization has begun, so these are two opposing forces in action.

Global gas turbine demand has accelerated faster in 2025 than at any other time in the past decade.

Big industrial companies are saying these are unprecedented times with unprecedented demand for power for Data centres. Specifically Nat gas capacity is backlogged upto 4 years already.

Lets look at a few examples.

1. EXHIBIT A ($GE Vernova)

The Gas Boss at GE was asked how their Gas business was doing, Whether it is experiencing softening. He said that People are instead paying extra premium to cut lines for Nat Gas capabilities.

More booking with slot reservation at $GE, meaning they want to pay more to skip lines. Half the backlog is trying to pay extra to expedite. There is only one direction that the pricing for Gas and related industries can take.

2. EXBHIT 2 (Siemens)

3. EXHIBIT 3 (Mitsubishi Heavy industries)

Mitsubishi Heavy Industries reported that it MUST double production capacity within two years to handle current orders, describing a “massive backlog” of new contracts and unprecedented order flow tied to power needs from AI and data centers.

Do you remember anytime when Heavy industry producers talked about doubling anything in 2 years? Woah are you serious? World war 2 maybe?

4. EXHIBIT 4 (Argan Inc $AGX)

Argan Inc (An industrial unsexy construction utility company just did a 6X on their stocks in 1.4 years. What world are we in?) just said this

As if this shortage wasn’t enough, there is also retirement of 78GW power plants.

”A substantial portion of the nation’s natural gas infrastructure is aging out.” — David Watson from Argan inc

Okay enough of pumping. Lets get to the relevant small cap.

Broadwind Inc. is a manufacturing company that produces structures, equipment, and components for clean technology and other specialized applications. The company focuses on serving the power generation, wind energy, and infrastructure sectors.

It is priced at 0.3 Price to TTM sales. 50M market cap for 150M revenue. Undervalued

THe bussiness can be lumpy, but the increase in working capital is a very encouraging sign.

Broadwind’s product portfolio includes complex components such as wind towers, gearing products for natural gas turbines, and industrial fabrication solutions.

Since Wind energy faces some headwind (pun intended), they are pivoting to focus on Industrial solutions for Gas turbines OEM’s.

So what caught my eye to deep dive into this? This is what they said in their earnings call

“On Industrial Solutions, primarily services the large gas turbine markets, and those are over 100 megawatts. So far this year through Q2, the market sold 93 units versus 21 units last year at the same time. So that’s 4x up.”

They emphasize U.S.-based manufacturing and are strategically shifting towards higher-margin, stable recurring revenue streams.

Here is some of the commentary from their SEC 10Q filings.

“We received $20,956 in new orders in the second quarter, up from $18,372 in the second quarter of 2024. Industrial Solutions orders increased by over 200% compared to the prior year quarter primarily due to an increase in demand associated with new gas turbine and aftermarket gas turbine projects. Additionally, Gearing segment orders increased 45% versus the prior year period primarily due to improved demand from most markets served.”

Clearly they are getting some boost on their gas focused bussiness. On top they are selling the unprofitable Manitowoc facility for 13M additional cash on their sheet, also 8M less costs.

So Industrial Solutions, primarily services the large gas turbine markets, and those are over 100 megawatts. So far this year through Q2, the market sold 93 units versus 21 units last year at the same time. So that’s 4x up.

Assessment: Can They Execute the Nat Gas Pivot?

Positive Signs:

• Gas turbine orders growing 200%+ with multi-year contracts worth $8M+​

• Market growing 4x YoY with decade-long backlogs at major OEMs​

• Gas turbine margins (6-7%) better than wind (currently negative)​

• CEO Eric Blashford explicitly prioritizing “diversification” and reducing wind dependence​

• Working capital on the rise. This could mean they are closing sales and are gearing up to deliver. Ofcourse it could mean inventory build up or other things. Just the closing sales is more plausible

  

  
  

Critical Concerns:

• Still 52% dependent on wind revenue—unclear how quickly they can pivot​

• Margins collapsing across all segments including gas turbines​

  

• Currently posting net losses (-$1.4M in H1 2025)​

• Working capital requirements ballooning while cash remains constrained​. This one is significant, it either shows that they are building inventory or

The Bottom Line

The natural gas turbine opportunity is genuine and substantial. One of their segments grew 4x YoY with multi-decade tailwinds from AI infrastructure buildout.

BWEN might be a good overlooked play. But the rerating of the stock might not happen anytime soon, so get ready to hold it for some time. I haven’t bought so much of it because I do see a lot of opportunity everywhere in the market.

Unless we increase voltage, the copper wires in AI data halls will be as thick as Slytherin’s serpent

As AI models grow larger and data centers scaled up, the infrastructure that powers them faces a fundamental re-architecting. The traditional paths of voltage conversion are being pushed to their limits, forcing innovation across the entire power chain:

1️⃣ Grid transformers stepping down from 13k–34kV to 800V
2️⃣ Server racks converting 800V to 48V
3️⃣ AI chips operating little under 1V

Each stage in this cascade represents both a technical challenge and an opportunity. The AI disruption is happening deep in the physical layers of the power grid, server racks, and the silicon

The winning stack is
34 kV AC → 800 V DC → 48 V SELV → 1 V (at the die).

This is what the industry is now converging on. Let me explain from the basics

INSIGHT 0: PHYSICS NEVER BLUFFS

Electrical Power (P) = Voltage (V) × Current (I)

Power is the amount of work that can be done. Voltage is the potential difference, and current is the rate of the amount of electrons flow.

Now the best analogy to think about Voltage is the water pressure, and current as the amount of water flowing through the pipes.

Now if you want to increase power, you can increase by increasing the I or V. But the problem is losses are defined by the following

Loss = I² × R.

So ideally, you want to have current as low as possible since current has quadratic relationship with Losses.

So Double voltage → halve current → quarter the losses

Lower current also shrinks:
– Cable cross-section (copper ≈ $8/kg)
– Bus-bar mass (try lifting 2 m of 4,000 A aluminum)
– Connector bulk & labor.

Higher Voltage is basically free efficiency?

HERON the new startup from Drew Bagliono (Ex-Tesla CTO or VP, they couldnt decide his designation as well, Basically X-battery boss), explains it well in their white paper. Basically they want to delete a bunch of steps in the middle and increase the efficiency by 2-4%. Meaning if you have a 1GW data centre, you will save ±40KW. This is huge.

The fact that Drew is starting a startup to supply Solid States Transistors (SST) tells you all you know about the market shortage. There are a few incumbents who might be able to produce it like Siemens, ABB etc. and it remains to be seen if they will be able to come through.

I have a few small cap stocks to reveal who will be the beneficiaries from this trend in the following posts.

INSIGHT 1: FROM GRID TO HALL
(34 kV AC → 800 V DC)

Status-quo :
• 13-34kV utility line → 480 V AC transformer → central UPS (AC->DC->AC) → PDUs.
• Five fridge-sized boxes, ~93 % round-trip, weeks of field wiring. Meh. 93-94% with the best equipment mind you.

Upgraded path:
• Solid-state Transformers spit 800 V DC straight into the hall.

NVIDIA’s been on this 800V train now since a couple of years.

Moving to 800V DC transmission is quite a big deal, as it allows for significant efficiency improvements, large cost savings in materials needed, and much higher upper limits on per-rack and per-data center power delivery.

While the 800 VDC notion isn’t new, having the world’s largest chip designer fully embrace the standard throws substantial weight behind it

Lets face it, we all want this on our chest too. Mark of confidence

SiC MOSFETs, film caps, breakers all sweet-spot around 1 kV.

Punchline: The same copper that once fed 15 MW now pumps 30 MW. Congrats, you just doubled your data-hall without touching real estate.

Porche was the first automaker to switch to 800V charging system. They got the fastest charging car in the Eueopean continent. 262kms in 10 mins. Naice.

INSIGHT 2: INSIDE THE RACK
(800 V → 48 V)

Okay now you have reached the rack with 800V. What happens then?

# STATUS QUO

• 400/480 V AC enters a rack level PDU/PSU.

• That PSU outputs **12 V (sometimes 24 V) DC**, typically ≤5 kW.

• Current at 12 V is huge (5 kW / 12 V = 420 A) so each rack stops at ~20 kW before copper, breakers and heat become unmanageable.

# NEW TREND

• A few high-frequency LLC or resonant converters take 800 V→48 V, yielding a rack level **DC bus-bar at 48 V SELV** (safety extra-low-voltage).

• 48 V lets the rack carry 100 kW with 2 cm-wide copper plates instead of water-cooled cables.

But Why 48 V, not 100 V?

• ≤60 V is exempt from most touch-safe regulations, avoiding arc-flash PPE inside the row.

• Commodity MOSFETs & controllers work efficiently up to 60 V.

• Server vendors (OCP “Open Rack v3”, OCP “ORV3”) already converged on 48 V blind-mate connectors.

Result: single GPU racks exceeding 150KW become practical; the rack PSU loses only ≈1 % instead of ≈3 % at 12 V, not to mention the copper and cooling it just avoided

INSIGHT 3: ON THE BOARD
(48 V → 1 V @ <1,000 A)

# STATUS QUO

• 12 V → 1 V multi-phase bucks; 600 A rivers across FR-4. Entire PCB looks like a copper heatsink. PCB’s would then look like this.

This image is very sarcastic btw. Not how you can produce PCB’s.

# NEW TREND
• Factorized 48 V → 1 V modules sit millimetres from the Ball grid array (connectors).
• Until the final inch, you’re only pushing 12 A; transient droop becomes a rounding error.
• GPUs still sip 0.9 V—physics of 5 nm CMOS didn’t change, and We just kept cramming more transistors.

Nvidia GPUs have evolved from about 400 watts on Ampere, to 700 watts on Hopper, to around 1400 watts on Blackwell, with next-generation chips expected to exceed 2000 watts.

Super-nerd stat: NVIDIA B200 (1.4 kW) needs ~1,400 A at 1 V. Try routing that at 12 V upstream. Your board traces will literally delaminate.

There is a player who could do such amperage. They got banished by NVIDIA 2 years ago. Today they got earnings, lets observe if they have any news.

Power semiconductors are usually very price elastic and quite a hard business.

Okay so let’s Summarize

• Efficiency: Nuke two conversion stages → +3-5 % site-wide. That’s 30-50 MW saved on a 1GW. Enough juice for another AI wing

• Power Density on racks going from: 20 kW/rack → 200 kW/rack. Meaning 10× compute per square foot.

• Copper savings: Current halves each hop → conductor mass ⨉ 0.25; 2025 LME copper is $9,000/tonne—do the math

• Cooling headroom: Less PSU heat → more delta-T for the H100 heat sinks

• Batteries: 800 V matches packs and skip the double conversion, slot in LFP cabinets, and you can go sip colada all day.

Here again, I have a misunderstood small cap that will benefit from this trend. Many thought leaders left it for dead but it might see a quick 2X in market cap.


TL;DR
Voltage is the cheat code. Crank it up early and keep it up for as long as possible. Then step it down ONLY inches from silicon, and you can unlock power dense, greener, cheaper AI factories.

Anything else is just setting money—and copper—on fire.

Finally got this post out. Will share the beneficiaries of this trends in the next post.

No financial advice and only used AI to proofread

The market has slowly started to realize how important utility-grade batteries are for data centers. The flavour of the month has been Grid Batteries in the stock market.

66x investment growth in utility-grade batteries over 10 years. It is also interesting to note that much of it is being colocated at the source of demand.

Batteries at data centers serve specific, valuable functions

1. Power quality management - handling rapid load swings from AI workloads.

2. Peak shaving and load management (price arbing)- reducing demand charges by up to 30%​ and making money by storing energy when it’s cheap.

3. Backup power replacement - replacing diesel generators (For eg. Microsoft is aiming for diesel-free data centers by 2030)​

4. Reducing interconnection delay - enabling faster interconnection. In some US regions backlogs are 5-9 years long. In EU it can be easily 10-13 years

5. Renewable integration - storing excess solar/wind for 24/7 use​

The first one is probably the most important one.

Data centre GPU training can destabilise the grid very fast. Look below an actual load from a AI data centres.

Dylan Patel explained how Meta engineers accidently open sourced this code named “Power_plant_no_blow” (unbelievable) to create fake wasteful demand in their training runs just to not have so much variability in their in the demand loads.

I spent an hour looking for the timestamp video for you all to hear it from the dragon’s mouth. Click on the video and it will take you to the timestamp.

Okay, so is this info still not priced in?

It’s the scale and extent of it that are not yet realized. All major hyperscale data centers will want batteries on premise.

I recently just started beeing active on substack. So go check out my original thesis on X (My original thesis) and come back. I had been posting non stop posts about FLNC in May/June. It has 4X since then so the ARR software valuation does not apply, however there have been a few more things that happened since then.

Quite a few things have happened since my post.This is according to the company slides:

This is what I am thinking:

• The One Big Beautiful Bill Act, or the OBBBB, came out with strong support for battery storage. It differentiates BESS from other sources of generation by recognizing FLNC’s technology as a dependable and dispatchable source of electricity, much like nuclear or gas plant

• MArket’s understood that AI data centre co location might be imperative to handle grid stability

• P.S. Stock’s about 30-35% SHORT according to Finviz and Fintel

  
  Did you just say 30% short interest after such a big move? Okay, lets give shorts a real scare stare.
  

Also btw, most of Fluence is mostly institutions owned; and their average price is around 18. They held through the whole drawdown. You think they will already sell it?

Fundamentals have a lot to like. Yes the margins could or should improve, but its a good trend. Also Stock based compensation on check.

Okay, let me also include some Excerpt from their last earnings call

Interestingly, he mentions there are some discussions with AI data centre operators. Hmmmm

I am definitely holding my breath (or heart) for the announcements. That might further rerate the stock.

Technology Advancements and Software Offerings
Beyond hardware, Fluence’s software platform (SaaS for grid optimization and battery management), is a rare in the sector.
I project this high margin ARR (145M in 2025) is probably going to 300-500M in 3-5 years. That alone could value the company at $3B in 1-2 years. Hardware business is also cherry on top.

Fluence’s software is very good in handling different types of assets from different companies. This is where it shines the most and could be unstoppable.

I might decide to deep dive into their software, as well as issues in general BESS systems need to handle at the software level in the coming days.

My Price targets for in the short to mid term is $36. Long term depends on how many AI data centre deals they are able to capture.

Risks:

• LFP batteries do have a risk of thermal runaway. South korean government lost 858TB of data (not backed up) due to data centre fire.

• In Jan 2025, there was a major fire at Moss Landing battery storage facility in California. Culprit was thermal runaway began with an internal defect known as abnormal lithium metal deposition and dendritic growth inside a single cell, which then propagated heat through neighboring cells.

  
  Might have contributed to downturn in Fluence’s stock starting in Jan 2025. Luckily, Nobody was harmed.

Since then numerous security measures are in place. Also, reminds me of media focusing on a few battery combustions in TESLA vehicles vs thousands of fires in gas cars. EV fire is largely a controlled problem. Utility grade fire will be a solved problem too.

Now there is a massive bottleneck to install batteries fast to the grid. It is not cell production. I will talk about it in the next post and reveal some plays on this problem.