This website shows that Carborundum universal is supplieng to Bloom energy. I’m not sure if this is a reliable source or not. Correct me if I’m wrong.
5 Likes
Sai Life PE 62, Market cap/Networth 10, Market cap to cash flow 67, EV/EBIDTA 32, ROCE 14%, Promotor stake 35% and CDMO is a lumpy business, stock is priced to perfection
12 Likes
I feel valuation is not always just the TTM P/E. Doing so ends up in a lot of value traps. The differential insight can only come from looking at capabilities that aren’t captured in the balance sheet as intangibles. That’s what tells us about the longevity and durability of one’s moat and a bulk of the valuation actually comes from these two (how much can the earnings grow AND for how long) than what the company did in the past alone (while it still matters). Its the growth and longevity that are primary drivers of valuation (see how a growing annuity is valued vs a lumpy one-time cash flow for eg.)
Most of the winners for me in the last couple of years have come from this approach and I have bought 50 P/E+ consistently in this period (Wocky, Shaily, Axiscades etc) and have had 2x-4x in few of these even when market did nothing since Sept ‘24. A business like Sai isn’t as lumpy as the other CDMOs. It has no commodity business which can drag while another molecule or two is growing. It is not reliant on one or two molecules. It is 100% innovator business. Doesn’t mean it wont suffer from lumpiness - but at least it will not be as lumpy as the others.
Take this for eg. The biggest molecules for Sai are Qulipta and Blujepa (in the last 2-3 qtrs that is). In Feb exports there is no Blujepa and still IDRx-42 made up for it. In Mar (20 days data only) there is neither Qulipta or Blujepa and yet Tyvaso and Quviviq have made up for it and for this quarter as a whole, none of the molecules are > 20% (Still this is the best quarter almost for Sai already if you remove Tradipitant from Q2 since this was from earlier quarters but shipped later)
Even this doesn’t cover the work they are doing for numerous innovators every month and how complex such of those chemistries are and how capable some of those molecules are in clinical trials. Do check out Sai thread and I have covered in depth some of these.
My pf is filled with high P/E stocks - Aeroflex, TD, Mtar are all 60 p/e or higher. I have seen how markets react in times of crises - capital seeks comfort in businesses that have good earnings visibility and relatively undisrupted cashflows and these bounce back durably after a market-wide correction. I have made mistakes of buying 12 P/E stocks only to see them derate to 8 P/E in such times even though they have strong counter-trend rallies that end up being bull traps.
This is of course not the only way to invest - its a way of protecting one’s capital and one of the important things in the market to know are when to protect capital and when to attempt growing it aggressively. Sometimes even protecting capital yields decent alpha by simply not losing and sometimes gaining. This strategy will not work as well in a roaring bull market and I am aware of it quite well. We are nowhere close to a roaring bull market at present.
111 Likes
Appreciate your detailed reply phreak, its all about your strategy you follow in different market conditions. Problem with high growth and high PE stocks is these may go in price or time correction once growth moderates and stock price comes to mean valuations. Timely exit is also important as any bad news is known at last to retail investor. There are very few business with consistent 20% plus growth.
Congratulations for new ATH, here fight is to limit losses. Conviction is getting tested on daily basis. 
21 Likes
This small paragraph talks hell lot of experiene and not sure how many understood the power of this. In the recent past, i have taken time out, read more, introspected my past way of investing and used lot of AI to understand better. This paragraph is a good summary of my learnings. But knowing and practicing this is another level. Thanks for re-iterating this for me.
18 Likes
Have not seen an update in a while from you Phreak.. Really looking forward to one.
10 Likes
I don’t have much to add so haven’t written anything. I continue to stay invested in the same businesses. I do have some things to say on the topic of AI from a technical, business and philosophical perspective as I have done here few times before. So continuing on those lines - feel free to skip the sections you don’t like. However, I think to understand my overall perspective, its important to know where am coming from. I don’t use AI to write so you won’t find the familiar “that’s not a trend, its a phenomenon” template of hype.
I have been immersed in AI, using at various levels and forms (claude, claude code, within phreakonomics using local LLMs, building mcps, organising knowledge bases in obsidian and so on), fine-tuning models (a 9b param model - nothing much to speak of), building agents (similar to claude code, but for 3d printing along with son), tweaking ways to run local models (using mlx-vlm, llama.cpp playing with turboquant and other kv quants, dflash and speculative decoding, obsessing on throughput token/sec without foregoing quality too much).
As a big fan of running local LLMs for a few years now, I can finally say we are really getting somewhere. My API costs have come down drastically from its peak because of the same and I can say the quality of local models is now so good that I am finally able to parse earnings PDFs locally within few mins at zero cost and its 95% accurate across pnl, bs, cf and segment data.
Few of the other things I use on phreakonomics - summarising announcements, scoring them and surfacing the important ones on a daily weekly basis (across categories like order wins, acquisitions, risks and disruptions, business updates etc), recording and transcribing (using parakeet and diarization pipeline), scoring and summarising conference calls and extracting guidance, synthesising sector level knowledge, generating business synopsis using ARs, transcripts and ppts, extracting business specific kpis from ppts, quick ways to track and compare earnings and so on. All these can be done today using a local LLM.
Even generating an exhaustive investment report using all these pre-digested inputs can be done locally in < 5 mins. The general idea is to organise things into information, insight, knowledge and wisdom hierarchy so that its easier for you to get access to anything you want quickly and also allow these as inputs to a LLM like Claude using a mcp which would give a way, way richer experience at way less token usage because the information hierarchy is pre-digested and allows for what’s important to rise up naturally.
I wouldn’t like to think am alone in this as is clear by the growth of anthropic and openai revenues. People are building all sorts of useful and useless things with AI. There are people building things of very high value with limited shelf life (because invariably margins will get razed to the ground and moats destroyed) spending a bomb on API costs. Anthropic doesn’t have compute to cater to its growth while OpenAI’s brash risk-taking promises to pay off big next year. Data centers are becoming the most crucial in all this because pretty much everything is running on supply shortage one after the other, from gpus, memory, photonics, cooling and power infra. It appears to be similar in size and scale to the railroads capex.
As an investor, its a once in a lifetime opportunity to benefit from but also one to be wary of since railroads capex in the US was one with a lot of bankruptcies but the public did benefit from all the railroads built. But it did play out over decades and the bankruptcies came in the 3rd or 4th decade. So we might still be very, very early in the cycle.
AI is a 5 layer cake - with application, models at the top and the chips, infra and energy below it. We would like to think the guys at the top would accumulate most of the value and so far it appears to be so - with OpenAI and Anthropic garnering values in 100s of billions with Anthropic reported valued at $1 trillion in pvt. markets. Following this, there is clamour for soverign models and soverign AI apps and India AI mission trying to promote these ventures. I am however somewhat skeptical of these because the open weights models coming out of China are so good and have no strings attached. A lot of the AI compute doesn’t need SOTA models - you dont need Opus to summarize things or generate reports - a Deepseek v4 flash can do it at a fraction of the cost. Even what sonnet does, a Deepseek v4 pro is able to do at a fraction of the cost ($25 vs $3.87/million tokens - and discounted at $0.85 currently). I do however see the point of sovereign data centers though, if not sovereign models as these deals may not last forever. But its undeniable that 60-70% of the work can be done by open weights benchmark equivalents to SOTA models of last year.
This also makes older GPUs like the H100 or even the A100 lot more valuable and I can see from what neoclouds are charging that annual rental for a year now covers what the card cost back in 2020. Even otherwise, they are able to cover the currently elevated cost as well within 2 years and these cards can be relevant for 7-10 years though accounting depreciation might be for 6 years (Because models keep getting efficient and your mundane information crunching tasks can be done even better than they can be done today).
India I think has a substantial edge in all this - while we don’t have an edge on models, applications and memory/logic chips, we do have decent trained labour in electricians and plumbers (both in great shortage and paid upwards of $200k/yr) and a lot of the DCs are becoming modular where they can build offsite and transported in containers, we have engineering capability in energy and transmission and electrical equipment. We have precision engineering capability to build liquid cooling components and build assemblies. We don’t particularly have issues with the US govt. and can get access to cutting edge Nvidia chips and neoclouds can be built here with companies serving the models that can cater to the 60-70% of the AI compute with open weights models. This is a substantial part of the pie. With middle-east facing security threats to their infra, our defence capability and connectivity and energy costs also makes us attractive to hyperscalers. I find it a bit irritating when I see people write us off in AI. I feel we have a big role to play.
Anthropic is the one which is high-flying with the highest risk at this point. They absurdly provoke everybody from their own customers to the govt. and while they have a great product, they have got crucial decisions on investments wrong and unfortunately, though they are correcting it now and it might be too late as its not easy to put up DC capacity. They are bidding up existing capacity which means we will continue to see deterioration from their models as they try to protect margins which are already non-existent. A quick comparison between what you get for a $20/mo pro vs what $20 can buy your in anthropic sonnet/opus API will tell you they are losing 5-10x on this sub. They have however built good apps and idiotically also leaked source code for it (for claude code at least). It is anyway unnecessary as its trivial to trick claude code into using a deepseek or a local LLM by setting base url and key and passing model name. So there is zero moat in claude code harness unlike what people think. Only people for which these are indispensable are the ones working in the cutting edge of drug discovery, cybersecurity or quant trading where the money lost/gained can be substantial and worth it paying hefty sums for enterprise access to Opus 4.7 or Mythos.
While I personally like to be on the cutting edge of tech for the sheer joy of it, it is also unsettling how good the models are getting. I find it interesting that communism vs capitalism is playing out the way it is. Chinese labs sharing knowledge and giving the world open weights models and huawei not looking for 70% margins like nvidia and cheap solar power is giving a run for the money of capitalists. I think China can really make a mockery of pricing like they have done in ev/solar/mining/chemicals/plastics/electronics etc already and that’s what we need to watch out for.
Disc: I am invested in the space through multiple businesses
90 Likes
@phreakv6 It’s fascinating how you manage such a concentrated portfolio and churn them, mostly at the right moment. Perhaps, the deep research you do gives you that degree of conviction.
I still remember you getting out of Wockhardt last year when you felt the valuation gap had closed and probably the differentiated insight was becoming common knowledge - and it was the right call in hindsight. I’m again surprised to see no mention of Axiscades in your recent portfolio disclosure. Ofcourse it’s not trading cheap, but if the management projections hold true (which they seem to be in track for), I feel there is a lot of upside. Is it something similar to what you did in Wocky that made you exit the position? Don’t know of anyone else that has done deeper work on Axiscades - hence would love to hear your take on it.
18 Likes
Hi @phreakv6 , I am invested in Aeroflex too. Have you got a chance to read about Thermax data centre liquid cooling solutions?
Posting a few snippets:
This is the concall transcript of last concall. They have won an international and one domestic order for their liquid cooling solutions and has indicated that they will ramp it up and do capex.
Other snippet is a brief description of their liquid cooling solutions which claims that it saves 80% water.
Thermax is a giant when it comes to industrial cooling solutions across the globe, also at much better valuations and better looking chart i.e. not very extended one.
Only issue is DC cooling will be just one part of many things they do, so rerating may not be as good as it happened in case of Aeroflex.
6 Likes
The above simple buildup show that Aeroflex baking in too much too soon? I booked out Aeroflex on the above basis..
4 Likes
@phreakv6 I done a rough calculation on mtar and want to know your perspective on this.
If mtar’s profits become 10x from here and market gives it a pe of 100 then still the market cap will be around 30000 crores. there is very much optimisim around it when all the future earnings have already priced in then what is the meaning of holding it for long term.
2 Likes
All the results are out for pf stocks. Everything has been great, some have been outstanding. Overall, no big surprises on the upside or downside. However, a lot has changed on the outlook for all the stocks I own. So I felt I will put down my thoughts on what’s new only.
MTAR - Guidance was upgraded to 80% revenue growth coupled with EBITDA margin of 24%. Earlier guidance was I think 50%. This is of course a night and day change in earnings power. Sometimes a stock at 300 P/E can be a contra value bet. In fact this is what I was trying to convey in my previous post, that while I was pessimistic on overall market valuations, I myself had loaded up on very high P/E stocks - precisely because P/E is not useful for valuation. Bloom has upped guidance and going by a recent report Goldman Sachs report, 25-50% of BTM power will be provided by fuel cells (though relatively expensive, its nothing compared to overall costs). So my guess is Bloom will have to continue doubling its capacity every year from here almost to get to something like 20 GW by 2030. The nuclear and defence segment as well is doing quite well and in addition to this, they have also received another data center order outside of Bloom. As they say, when it rains, it pours. I think FY27 PAT could be 250-270 Cr here and in all likelihood, 500 Cr by close of FY28. I know it sounds outlandish, but I feel its likely.
Sai - Market sold off despite a 109% PAT growth YoY for full year. I am not sure what market was expecting. Thankfully it did recover most of it though by close. I suspect people still havent understood how to read CDMO results. I saw the same thing happening in Neuland with market getting excited over the bumper quarterly number when it was sedate compared to its FY24 base. Anyway, the key takeaways are that tariff has no impact - its not new but reassuring to know. I had done extensive research on this - since CRO is 35% and exempt and development work is also exempt (Sai does a LOT of phase 1/2/3 supplies), among commercial innovator molecules bulk of it is to large pharma who have made deals with the govt. Only United Therapeutics and Tyvaso I felt could be impacted (if at all) so good to know tariff impact is a non-issue. Other than this, the new detail is that 1100-1300 Cr is planned capex next year after 650 Cr spent this year (220 KL coming in H2 and 220 KL next year). Since Sai has never invested without visibility, its clear they will continue on their growth path over next 2-3 years as well. With USD to INR nearing 96 and Sai deriving 100% of revenues nearly from exports, this investment gives access to a naturally hedged and growing future cashflows in USD. Some info from the innovators. In this Qulipta/Ubrelvy are end molecules for the intermediate Sai supplies and they are growing at healthy rates of ~50% (International doubling YoY). Rinvoq is the one that Anthem/Divis supplies (ignore the highlight on that)
Quviviq is growing 74% and Vanda has launched Tradipitant as Nereus. The other nice detail comparing Q3 and Q4 PPTs is that Active commercial molecules has gone up from 30 to 34 (+4) and Phase-3/Pre-reg has gone up from 6 to 11 (+5).
TD Power - Execution is on track with new capex coming online helping numbers as expected. They are well on track for 2400+ Cr (conservative) FY27 and 3200+ Cr FY28 (very conservative imo). Gas engines and gas turbines to DC is what will continue to drive the numbers and Innio has filed papers for IPO and is very bullish on its forecasts. TDPS management mentions they have a long-term commitment from Innio that extends to 2030! Just like MTAR will grow with bloom, I think TDPS will grow with Innio (and CAT’s subsidiary Solar turbines) since Innio is tripling capacity as per their IPO papers
I think between SOFC and gas engines/turbines, bulk of the BTM buildout will be captured - since it may not require grid backup or BESS. Just a 20% redundancy should guarantee a N+1 which will be cost-effective as well. Irrespective, both Mtar and TD are set till 2030.
Aeroflex - I feel this is the dark horse among the pack because there is a lot that can go right here if they can leverage capabilities quickly enough - and so far, they have shown that they can, since the liquid cooling skids wasn’t even part of business same time last year and is now a 5% contributor for FY26 and will be 20% in FY27 and most likely 40-50% in FY28. Sticking to the theme of only talking about what’s new, one of the most important things is this
The management has guided for a 3.25 lakh as avg realisation per skid. This is a considerable jump from 15000 skids with peak revenue of 350 Cr, yielding 2.33 lakhs per skid. This I feel is still very conservative because Aeroflex can contribute a lot more value per skid. I suspect they are even capturing 10% of what’s possible. Not just in the skids, but in the overall liquid cooling - especially if they can break into the rack level components which will be much higher in value.
There are initial signs of that. I am not sure if I put out my thesis on Hyd-Air here in the previous posts. If you haven’t yet, do have a look at their products. There are a lot of high value stuff in these, like the QDC and other fittings and as per specs they are interchangeable with some of the best in the business.
This was one of the things that got me exciting initially on Aeroflex in my initial research because this company is also setup at Chakan close to Vertiv’s own manufacturing plant. In recent call, there were some things on Hyd-Air I found interesting validating this thesis
My guess is Hyd-Air will continue to grow at a great rate and contribute to Aeroflex internal consumption for value add in skids and also in the last-mile rack level components. In terms of exit run rate, looks like post 15000 skid expansion, Mar 2026, they can exit at 60% util (750 skids) which means a 300 Cr contribution from LC in FY28 (750x12x3.25 lakhs) even if we assume no growth in that run-rate. From 0 last year to 600 odd skids per quarter now to 750 per month is breakneck scaleup for what is essentially a completely new line of business. I think 15000 skids is perhaps ~200-250 MW of capacity based on some assumptions. If India has to add 5-10 GW of DC capacity and Vertiv has 70-75% of this market share (as per Aeroflex Arihant call), and Vertiv has exclusivity with Aeroflex, you can imagine how much this capacity has to go up by in the next 4-5 years. It has to pretty much double every year, just like Mtar if things play out. Its still very early days though. On top of this, they can increase the % of value add by adding new products as well which is why I feel this can be the dark horse as there are several things that can keep surprising on the upside.
Aimtron - Among other stocks I have studied recently, liked Aimtron. In short, I like how they have scaled business from doing PCB design and assembly (65%) to doing complex box builds (35%) for global OEMs. I feel the AI box builds is a fairly large market size and I think Aimtron can capitalise on it. They have a decent order book (600 Cr vs FY26 rev of 300 Cr). Among EMS players, their growth has been very good and valuation relatively cheap. The main thing that excited me was their foray into optical transceivers.
Please note that the PLI is not yet approved, so its not worth getting excited about yet. If it does get underway, its a good import substitution play of a high value component. SFP Optical transceivers are used in several places in telecom and DCs and Aimtron seems to be targeting 1G to 25G speeds. At the top end, it can be used in network level use in DCs. At speeds as high as 800G, it is used between GPUs in Vera Rubin but that is extremely specialised and companies that have that capability can be counted in one hand. What Aimtron is attempting to is the most basic stuff - but even that requires capabilities beyond what they possess. They seem to have got an international player who might transfer technology.
If this optionality plays out, Aimtron can do very well but I think better to treat it as an optionality. I have a tiny position here to track and see how it goes and might scale up if they get PLI and ToT approved and have a clear path to how they will finance it considering they dont have the balance sheet or cash flow for it.
Disc: Invested in all names mentioned and very likely biased. I am very much a novice and write to gather my thoughts and this is not investment advice
106 Likes
Hi Phreak, DId you get a chance to study Sterlite Tech ? Any thesis on same
1 Like
In this context, this is the comment that caught my attention:
Rather than do some kind of a fabrication job and supply individual components, they are now doing the design itself. This moves them up the value chain to become a ‘solution provider’ rather than a parts supplier and is probably one reason why the realizations can be expected to go up.
(Disc.: Invested)
27 Likes
This is a good news for DC proxy players in India. Players like Aeroflex, STLTECH, Cemindia, Adani Green, Adani Energy Solutions etc. should benefit. Not sure about MTAR though.
20 Likes
Bad news for India. World is outsourcing their energy and water problems to India.
6 Likes
Some thoughts as usual across the two spaces am interested in CDMO and AI/DC
WuXi AppTec has been included in Pentagon list as a Chinese military company.
This is a strong tailwind for Sai since there is good chance of further tech transfers as it happened in 2023 (the drugs which WuXi Apptec make cant be sold under medicare plans and they can’t get any projects funded by USG etc.). Not that Sai requires it, considering the number of molecules and engagements they already have but it can be a good problem to have. I hope they can capitalise on it.
The other great thing that happened was Daraxonrasib came out with great results from phase-3 trial for pancreatic cancer that doubles median lifetime. A result so spectacular that the medical community gave it a standing ovation. Sai happens to be the CDMO supplier for this drug - and this class of drugs which are K-RAS inhibitors are the new thing with Eli Lilly having their own Olomorasib in phase-3 with good results (Again, Sai is a supplier for this drug as well). They also have a BMS K-RAS inhibitor in earlier stage - so Sai likely has a good grasp of this tech. The other thing is the modality of action of Daraxonrasib is peculiar - it doesn’t inhibit the way Lilly’s drug inhibits but actually gums up the protein so that it can’t function. This class of drugs are called Molecular glues and is a hot area of research. In small molecules, I am getting more and more confidence that Sai is clearly in the cutting edge of tech, even if all this doesn’t yet all show up in numbers.
Mineralys’ Lorundrostat’s (Sai supplies and drug has PDUFA date in Dec ‘26) competitor drug from AstraZeneca got approved last month (they paid up for a priority review to be ahead of Lorundrostat). The peak sales projections for Baxfendy (Baxdrostat brand name) are in the order of $5b since there are 20m untreated patients in US alone for uncontrolled hypertension. Mineralys being so confident of approval has gone and acquired the licence outright for $200m from the company they would have had to pay royalty to. Having seen the trial progress, I feel this drug is a 95% chance of approval in Dec opening up a massive market for Sai.
AI/DC
Periodically market seems to throw a fit on AI/DC stocks - but the underlying tailwind remains strong with so many new avenues opening up from CAD design, hardware/electronics design, gaming etc. I dont think SpaceX, Anthropic and OpenAI investors at these trillion+ valuations are going to make money though. I think they will keep serving 40% of the AI demand which itself can be so massive that we can’t imagine it right now though - but they are priced as if they will serve 100% of the demand.
I keep thinkign we must invest seriously in DC capacity or we will be left behind. Thankfully, few things have gone right here with AirTrunk alone committing 5GW by 2030. This means that the projection for India goes from 1.5GW today to something like 10-15GW by 2030. Most DC plays in India are focused outward into exports (like Mtar, TD) but Aeroflex might be one of the purest DC plays for India. With current shipments roughly equal to 10MW at ~700 skids and installed capacity of ~100 MW and going up to 200-250 MW in a few months, it is nowhere close to catering to this demand being an exclusive partner for Vertiv that holds 60-70% market share. There is a lot of growth available for them to capitalise on if they execute right.
In the US there is ruckus about regulating DCs due to water usage etc. which is the luddite equivalent of destroying the automated textile machines in the industrial revolution. With closed loop cooling and non-water based air cooling outer loop, the water usage is not going to be much but there is a lot of misinformation on the subject which should be put right in due course hopefully. Until then Mtar/TD may face bit of negative sentiment (coupled with lofty valuation, esp for Mtar) but I don’t see it persisting for long.
Another thing I noticed is Apple is finally headed in the right direction with AI. Watch from 35min mark. They have signed a $1b deal with Google for a gemini based 1.2T param model. I feel this opens up the AI market some more because it can increase consumer AI usage multifold. Apple should soon join the DC capacity buildout race along with the others since they want to self-host the models due to privacy of its users.
xAI renting out compute to Anthropic and Gemini has depressed GPU rental prices somewhat which people are taking to mean that AI demand is sagging which isn’t the case. I think it will be temporary as Anthropic has used the opportunity to launch a nerfed version of Mythos (Fable 5) which promises to be a strong token sink. There’s another usage pattern starting to gain traction as well called loops - essentially instead you prompting the AI, this involves giving AI a long-term goal and have it figure out how frequently and what to prompt and spin sub-agents to achieve that goal. I can foresee this being an absolute token hog as well.
There is one more theme I have mentioned here multiple times as it is dear to me, which is local AI. It is gaining a lot of traction - even if not local, self-hosted open weights models can really be a thing (openrouter at $1.3b valn seems like a steal compared to SpaceX $1.8t). Of course these need DC capacity too so its not a risk but can be a risk if hosted out of China instead as they are paying less margin on GPUs, memory, power. The local AI i dont foresee it materialising because this local consumer hardware is competing for the same fab, memory, gpu compute - and DCs can always afford to pay up higher than a consumer because they can sweat that asset 24/7 unlike a consumer GPU which will mostly lie idle - you simply can’t compete. 5090 at 4-5 lakhs seems like the long-term equilibrium compared to its launch price of 2 lakhs. Its the same reason why you can’t find a new PS5 Pro anywhere.
Disc: Invested in all names mentioned
69 Likes
Thanks @phreakv6. Didn’t know about water situation in US. Will read about it. Btw, by any chance, have you had a look at STLTECH? Optical fibre theme is one of the strongest sub-theme in AI DC space. Further the prices of preform is on the rise and STLTECH is the only Indian player who has command on this end to end. Besides there are strong domestic theme as well which may boost sales and margins.
4 Likes
Thank you @phreakv6 . When you get a chance and if you feel like it of course could you also pen down some thoughts on mistakes you’ve made in the past either where your initial thesis broke down or you sold too soon or waited for the right price that didn’t arrive or even instances where the allocation was wrong. Would be keen to see how you approach these situations. Thank you once again for the time and effort you’ve already put into this thread.
5 Likes