Eh, boss! Come, sit down lah. Muthu boy here, your favourite prata seller at the corner stall. While flipping roti prata, I also flip some good stories. Today I tell you one damn interesting one. You got time ah? Order one kosong prata, I talk.
If I tell you got one company, last five years its share price drop more than 90%, like kena hammer until flat... but now got chance to become next 10x AI infrastructure dark horse, you believe or not?
More shiok is this: They don't make chips, don't build big AI models, and many investors never even hear their name before. But what they got? Maybe more important than chips. This company is FuelCell Energy, $FuelCell(FCEL)$
Today, like my prata, I break it down nice and clear for you — three big questions.
First: Why suddenly this old forgotten fuel cell company come back into Wall Street radar?
Everybody think AI short of chips. But go ask the real big bosses — OpenAI, Google, Microsoft, Meta, Amazon, Oracle, even Elon Musk xAI. Chips can buy, servers can buy, talent also can poach. What they cannot easily get? Electricity, lah! Crazy amount of electricity.
Example: OpenAI alone, by 2030, their data centres need about 30 gigawatts. One nuclear plant give roughly 1 gigawatt. Means OpenAI need 30 nuclear plants just for themselves! Add all the others, total AI power demand maybe cross 200 gigawatts. That is like building 200 new nuclear plants in five years. Wah, impossible right?
Nuclear power? Take 7 to 12 years to build, with all the politics and protest. AI cannot wait. Wind and solar? Depends on weather, and need huge land. Data centres want power 24/7, no fluctuation. Natural gas? Turbines all booked until 2028-2030, and emission permits also headache.
So every road blocked. This is the big problem.
Now, enter fuel cells. No need wait 10 years. No big land. Can put beside the data centre. Sign today, power tomorrow. This is what the big boys want.
Look at $Bloom Energy Corp(BE)$
What is a fuel cell?
A fuel cell is a device that makes electricity from fuel and air. Instead of burning the fuel to make heat to drive a mechanical generator, fuel cells react the fuel and air electrochemically, without combustion. The electrochemical approach avoids pollutants that are created by high flame temperatures, and it is a more direct and efficient way to make power from a fuel.
Fuel cells are configured in stacks of individual cells connected in a series. FuelCell Energy’s carbonate stacks have up to 400 cells per stack and produce between 250 kW and 400 kW of power. FuelCell Energy’s standard MW-scale module contains four stacks, nets around 1.25 MW of power, and can make electricity for sites like universities or hospitals. Fuel cell technology has powered everything from laptops to space shuttles. The modular design of fuel cell plants lets them scale up to a site’s energy needs. In South Korea, one fuel cell park produces 59 MW of power.
A fuel cell power plant is made up of one or more fuel cell modules. The plant also has electrical and mechanical systems that deliver gasses to and from the stack module and to convert the DC fuel cell power to AC power at the desired voltage. The small footprint, quiet operation, and clean emissions profile make them ideal for highly populated areas. FuelCell Energy’s plants are designed to be safe, clean, reliable, and highly efficient.
Hydrogen production is key to advancing intermittent renewables like wind and solar. FuelCell Energy’s platforms, both current and in development, can provide power during undersupply and store power during oversupply. A process called “electrolysis” uses power to produce hydrogen, an energy carrier. Stored energy can be transported for use elsewhere. So, the energy from sources like wind and solar can be available for use anytime, anywhere.
So why look at FCEL?
Because the market is so hungry, even the number two, three, four players also can eat full. Not enough supply. FCEL is one of the few that can deliver big MW-scale systems for data centres.
They already got standard 12.5 MW power modules — like Lego blocks. Need 50 MW? Stack four. Need 100 MW? Stack more. Data centre bosses love this. They don't want story, they want fast power.
Now, the real question — can FCEL make money or still keep losing?
In the past, they supply university, hospital, small industrial. Orders small, factory not full, lose money, then issue new shares, dilute old shareholders. Classic sad story.
But now? AI demand is different level. They don't care only about cheap price. They care got power or not. Faster you can install, faster they pay.
Market talking about a potential 60 MW data centre order. If this one confirm, wah... game changer. Means big tech already test and trust them. First one open the door, the rest will come running because everyone also desperate for power.
When orders come in, factory utilisation go up, cost per unit drop, margin improve, cash flow better. Slowly no need keep dilute. From "loss-making green company" become "AI power supplier". Different label, different valuation.
Now, how much can it go up? Let me tell you even with very conservative calculation.
They plan expand factory in Connecticut from 100 MW to 350 MW capacity. I don't trust management fully, so I cut half — only 175 MW sold in 2027.
Each MW sell about US$2.5 million. So revenue about $437 million. Current revenue only about $158 million. Means almost triple.
Assume gross margin improve to 15% only (not crazy). Give it only 18x sales multiple (half of what Bloom got at peak). After heavy dilution to 65 million shares, still can reach around $121 per share.
Current price is $17.33 after 19% discount on Friday . Even my super careful calculation give big upside. Of course, not guarantee, this is just if they execute.
Boss, this is classic turnaround story. Risk-reward very asymmetric. If wrong, limited downside. If right, can fly.
But Muthu boy remind you: Don't go all in. FCEL got history of losing money and issuing shares. Control your position. Watch these three things:
Confirm the 60 MW data centre order.
Bosses start buying their own shares.
Gross margin keep improving and turn profitable.
If these three happen, then the story become real.
So, if the 60 MW order confirmed, you want to just watch from side or give FCEL small position to "gamble" one small prata-size bet?
Come, your prata ready. Eat first, think later. See you next time, boss!
@Tiger_comments @Daily_Discussion @MillionaireTiger @TigerPM @TigerObserver
Comments