Tecogen is reminiscent of other data center buildout winners $TSSI and $PSIX. Its products are patented and differentiated, and have only a 2-year payback period.
We've followed this name for a few years--wondered if the short-sighted "green"/woke anti-NG policies of NYS would result in their demise. Encouraging to see growing realization that NG is probably the preferred option for booming energy demand, and may provide a fast onramp for DCs well before SMRs get off the drawing board. Great insights on this by Murray Stahl (pg 6.) re: co-location of DCs in the Permian. https://www.frmocorp.com/_content/letters/2024.pdf
I see they mentioned $LB. The stock’s been fluctuating between $60 and $80 and has recently shown a gradual downward trend. Their subsidiary, PowerBridge, appointed the former CEO of Talen Energy as its CEO. He was behind the sale of a data center—built next to a nuclear power plant—to AWS, showing real vision for investing in this emerging concept!
Ahha, they are Horizon Kinetics! No.1 3-year performance hedge fund according to HedgeFollow!
Thank you for the interesting idea and write up! Very compelling but I'll play devils advocate
a) Is there a risk that these guys are just trying to jump on the next fad? I see that they had previously positioned themselves as a cannabis play in 2021/2022 during the cannabis stock mania:
What gives you confidence that this isn't more of the same?
b) What is the "moat" here? The stuff they are building is not rocket science so is there any opportunity for sustained pricing power and/or recurring revenues or will this get quickly competed away?
I don't know how you get into these so early. Vertiv the number one datacenter cooling company has partnered with Tecogen. This stock is guaranteed to explode. Hats off my friend ;)
Hi thanks for the write-up but the thesis is flawed because mobile gas energy systems generate electricity at a cost of about 0.15 per kwh. But you can only charge out to the customer at about 0.05 per kwh
I stand corrected. They generate electricity at a price cheaper than Bloom Energy and cheaper than mobile-gas powered generation. The thesis is solid given that 30% of data centres will need power that's not sourced from grid going forward
1. Why hasn't the company been been overflowing with orders the last few years if the RoI is so obvious?
2. May be answering my own question here but isn't the benefit of a hybrid system significantly mitigated by signing a PPA (guess this is only relevant for the larger sites)?
If you look through the past several quarters of transcripts, you see an evolution in management's thinking. At first, they didn't seem to realize that there would be a need for their products (again, likely because they thought of data centers only as the larger sites, like most of us do). However there's a fascinating moment on last August's call where they realize the opportunity from talking with some of their customers. The transcript speaks for itself, and also answers your second and third questions. I will paste it below:
"We've been working on data center projects for the past 6 months. I didn't want to talk about this before since the projects were too early of stage. When everyone thinks of data centers, they think of the massive facility that Amazon and Google are building. Initially, I thought that our products were going to be too small for this market. I asked one of our engineering and project development partners why he was recommending our products for data centers.
Why are you specifying Tecogen's equipment? Why aren't you specifying a large engine from Caterpillar or GE? What alternatives does the customer have? He walked me through the economics and the timeline that these customers are looking for. Now that we have made progress on multiple projects, I'd like to tell you about the needs of this market, why our solution is better than the alternative and when we should expect to see orders from this segment.
There are 3 sizes of data centers. Hyperscale is what gets all the press. Most data centers though are smaller. Most are colocation data centers. Here, multiple companies rent a portion of a bigger data center. Then there are the enterprise data centers, where a bank or a law firm has their own dedicated data center. In the past, data centers were built in the lowest utility rate areas and were primarily used for data storage. Today, with AI and distributed computing, the power draw is immense and liquid cooling is essential.
If you're Google, Meta, or Amazon, you can afford to build your own power plant or negotiate directly with utilities for favorable power contracts. Others are less short of power. Utilities are no longer able to provide all the power a data center needs. Power plants aren't being built fast enough. Utilities are also behind on upgrading infrastructure. The biggest need for solutions is in the colocation and enterprise tiers. If you need 8 megawatts of power in many regions, the utility will tell you that you'll get 5 and you'll have to source the remaining power yourself.
This has become such a constraint that space is no longer being rented in some data centers by square foot. It is being rented by required power for each tenant. If you're building a colocation or enterprise data center, what are your options if you don't have enough power? You could buy a backup diesel generator or a natural gas generator. Most generators though are only designed for 500 hours or so of annual operation. You may or may not have the appropriate service support for the uptime you'll need.
Emissions compliance for continuous operation is also tricky. Solar and wind won't work at night, but the data center needs power 24/7. You could build your own 3-megawatt turbine power plant, but that might take you 36 months or more to come online. Also, for an engine or turbine greater than 1 megawatt, lead times for certain components like electrical panels and switchgear are more than 24 months out. As more data centers are located closer to populated areas, neighboring buildings, noise restrictions, meeting regulatory and permitting also adds additional time.
How does all this affect Tecogen? If you have a data center that needs 1 megawatt to 3 megawatts of power, we have two solutions that will solve your problem. The first solution is our DTx engine-driven chillers. These can take all of the cooling load off electricity and move it to natural gas. They can be running on-site in 6 months or less. 5 to 6 units can be packaged in a container with all the connections for quick install on site. They can even be run with dry coolers to reduce on-site water usage from cooling towers.
For an 8-megawatt data center, this will immediately free up 1 to 2 megawatts, which might allow a customer to open on time or add additional tenants. Chiller modules can also be added as a data center expense."
Finally, it's good to look to the future, but I don't think we know what the world will be like in 2030. What I can say is the opportunity for frankly a 10x is there even if there wasn't a single additional data center built. And, even in that world, existing data centers without this product can always switch to it or use it when expanding.
I can see that colocation data centers need cogenerators, but why does that matter in the case that there's no growth in those data centers?
I'm trying to understand why there's a 10x opportunity here if no new DCs are built. What's the payback period for their cogenerators? Is it the 2-year period like for their chillers?
One follow-up: you mention the company has achieved 50% EBITDA margins, but I don't see their gross margin ever reaching 45%. And their max capacity is around $40m, so am I correct in thinking they only have $10m in spare utilisation before they need to look at expanding?
The ABI research shows an increase to 3k from 2k from 2024 to 2030 in # of medium datacenters (are these de novo, or does it include upsizing of small datacenters?). Since their research is paywalled, I wasn't able to see the context. Still seems like plenty of opportunity for a nanocap.
We've followed this name for a few years--wondered if the short-sighted "green"/woke anti-NG policies of NYS would result in their demise. Encouraging to see growing realization that NG is probably the preferred option for booming energy demand, and may provide a fast onramp for DCs well before SMRs get off the drawing board. Great insights on this by Murray Stahl (pg 6.) re: co-location of DCs in the Permian. https://www.frmocorp.com/_content/letters/2024.pdf
Great comments and intel, thanks for sharing!
I see they mentioned $LB. The stock’s been fluctuating between $60 and $80 and has recently shown a gradual downward trend. Their subsidiary, PowerBridge, appointed the former CEO of Talen Energy as its CEO. He was behind the sale of a data center—built next to a nuclear power plant—to AWS, showing real vision for investing in this emerging concept!
Ahha, they are Horizon Kinetics! No.1 3-year performance hedge fund according to HedgeFollow!
Interesting: https://investors.ies-corporate.com/news-releases/news-release-details/ies-holdings-acquires-arrow-engine-company
Thank you for the interesting idea and write up! Very compelling but I'll play devils advocate
a) Is there a risk that these guys are just trying to jump on the next fad? I see that they had previously positioned themselves as a cannabis play in 2021/2022 during the cannabis stock mania:
https://www.tecogen.com/news-events/press-releases/detail/405/tecogen-to-establish-new-business-unit-focused-on-indoor
What gives you confidence that this isn't more of the same?
b) What is the "moat" here? The stuff they are building is not rocket science so is there any opportunity for sustained pricing power and/or recurring revenues or will this get quickly competed away?
Thoughts on earnings? small loss but seems early days and they sound pretty bullish on ramping up revenue
I don't know how you get into these so early. Vertiv the number one datacenter cooling company has partnered with Tecogen. This stock is guaranteed to explode. Hats off my friend ;)
https://www.vertiv.com/en-us/about/news-and-insights/corporate-news/vertiv-and-tecogen-forge-global-partnership-to-enhance-cooling-solutions-for-power-constrained-data-centers/?utm_source=twitter&utm_medium=social-media&utm_campaign=ai-&utm_term=reputation&utm_content=en-link-post
Hi thanks for the write-up but the thesis is flawed because mobile gas energy systems generate electricity at a cost of about 0.15 per kwh. But you can only charge out to the customer at about 0.05 per kwh
I stand corrected. They generate electricity at a price cheaper than Bloom Energy and cheaper than mobile-gas powered generation. The thesis is solid given that 30% of data centres will need power that's not sourced from grid going forward
Hi, can I ask you where you got your insider trading information? The data you shared really supports your thesis on the stock, thanks!
Here's a link: https://www.nasdaq.com/market-activity/stocks/tgen/insider-activity
Couple questions:
1. Why hasn't the company been been overflowing with orders the last few years if the RoI is so obvious?
2. May be answering my own question here but isn't the benefit of a hybrid system significantly mitigated by signing a PPA (guess this is only relevant for the larger sites)?
3. How big is the modular/small opportunity? The following research actually has the number of small DCs falling in absolute terms by 2030 and medium essentially static relatively. https://www.abiresearch.com/blogs/2024/07/16/data-centers-by-region-size-company/
Great questions, thanks for asking.
If you look through the past several quarters of transcripts, you see an evolution in management's thinking. At first, they didn't seem to realize that there would be a need for their products (again, likely because they thought of data centers only as the larger sites, like most of us do). However there's a fascinating moment on last August's call where they realize the opportunity from talking with some of their customers. The transcript speaks for itself, and also answers your second and third questions. I will paste it below:
"We've been working on data center projects for the past 6 months. I didn't want to talk about this before since the projects were too early of stage. When everyone thinks of data centers, they think of the massive facility that Amazon and Google are building. Initially, I thought that our products were going to be too small for this market. I asked one of our engineering and project development partners why he was recommending our products for data centers.
Why are you specifying Tecogen's equipment? Why aren't you specifying a large engine from Caterpillar or GE? What alternatives does the customer have? He walked me through the economics and the timeline that these customers are looking for. Now that we have made progress on multiple projects, I'd like to tell you about the needs of this market, why our solution is better than the alternative and when we should expect to see orders from this segment.
There are 3 sizes of data centers. Hyperscale is what gets all the press. Most data centers though are smaller. Most are colocation data centers. Here, multiple companies rent a portion of a bigger data center. Then there are the enterprise data centers, where a bank or a law firm has their own dedicated data center. In the past, data centers were built in the lowest utility rate areas and were primarily used for data storage. Today, with AI and distributed computing, the power draw is immense and liquid cooling is essential.
If you're Google, Meta, or Amazon, you can afford to build your own power plant or negotiate directly with utilities for favorable power contracts. Others are less short of power. Utilities are no longer able to provide all the power a data center needs. Power plants aren't being built fast enough. Utilities are also behind on upgrading infrastructure. The biggest need for solutions is in the colocation and enterprise tiers. If you need 8 megawatts of power in many regions, the utility will tell you that you'll get 5 and you'll have to source the remaining power yourself.
This has become such a constraint that space is no longer being rented in some data centers by square foot. It is being rented by required power for each tenant. If you're building a colocation or enterprise data center, what are your options if you don't have enough power? You could buy a backup diesel generator or a natural gas generator. Most generators though are only designed for 500 hours or so of annual operation. You may or may not have the appropriate service support for the uptime you'll need.
Emissions compliance for continuous operation is also tricky. Solar and wind won't work at night, but the data center needs power 24/7. You could build your own 3-megawatt turbine power plant, but that might take you 36 months or more to come online. Also, for an engine or turbine greater than 1 megawatt, lead times for certain components like electrical panels and switchgear are more than 24 months out. As more data centers are located closer to populated areas, neighboring buildings, noise restrictions, meeting regulatory and permitting also adds additional time.
How does all this affect Tecogen? If you have a data center that needs 1 megawatt to 3 megawatts of power, we have two solutions that will solve your problem. The first solution is our DTx engine-driven chillers. These can take all of the cooling load off electricity and move it to natural gas. They can be running on-site in 6 months or less. 5 to 6 units can be packaged in a container with all the connections for quick install on site. They can even be run with dry coolers to reduce on-site water usage from cooling towers.
For an 8-megawatt data center, this will immediately free up 1 to 2 megawatts, which might allow a customer to open on time or add additional tenants. Chiller modules can also be added as a data center expense."
Finally, it's good to look to the future, but I don't think we know what the world will be like in 2030. What I can say is the opportunity for frankly a 10x is there even if there wasn't a single additional data center built. And, even in that world, existing data centers without this product can always switch to it or use it when expanding.
I can see that colocation data centers need cogenerators, but why does that matter in the case that there's no growth in those data centers?
I'm trying to understand why there's a 10x opportunity here if no new DCs are built. What's the payback period for their cogenerators? Is it the 2-year period like for their chillers?
Thanks for the answers!
One follow-up: you mention the company has achieved 50% EBITDA margins, but I don't see their gross margin ever reaching 45%. And their max capacity is around $40m, so am I correct in thinking they only have $10m in spare utilisation before they need to look at expanding?
The ABI research shows an increase to 3k from 2k from 2024 to 2030 in # of medium datacenters (are these de novo, or does it include upsizing of small datacenters?). Since their research is paywalled, I wasn't able to see the context. Still seems like plenty of opportunity for a nanocap.
Yeah, this doesn't keep me up at night at all.
This has 4x in 6 months already. Has it reached fair value?