Anaraj raj is at good valuation plus margins are increasing let’s how this quaterly result shows but have to invest but many stocks like voltamp are so costly that we can’t add much quantity and most important thing is we have to much more quantity like a pyramid for that fair price is imp
There are some players like Castrol India, Savita Oil in immersive cooling technology. They are talking about it, but it’s only an optionality.
But this space is crowded by other international players like NTT, Vertive, Castrol ( MNC ). APAR also does this.
Daikin ₹1 Lakh Cr. India R&D Hub Investment
This is the first R&D center outside Japan with a ₹1 lakh crore investment.
Sees AC market expansion with rising demand and data center cooling as a key growth segment despite weather disruptions.
1 Like
Optical Fiber Prices Triple in 18 Months Amid AI Boom & Helium Shortage
Prices up 3.3x (₹15 to ₹50/km in China). Bend-insensitive fiber for AI data centers at ₹65/km.
GPU racks need 16-36x more fiber than CPU setups. Hyperscalers alone are fueling 76% YoY demand growth.
Preform shortages extend lead times to 60+ weeks. Global capacity expansions are lagging, so prices stay elevated.
Major Beneficiaries:
2 Likes
3 companies control 75% of the large gas turbine market and all are sold out through 2030.
Wait times have exploded from 2-3 years to 7 years. Bloomberg estimates $400bn in planned power plants could be delayed or cancelled due to the turbine shortage.
TD Power manufactures generators for gas turbines and engines used in power plants for data centers globally.
As tech companies move to captive power to avoid high grid prices, TDPS is a direct beneficiary supplying generators in the 20-100 MW range.
Read this interesting Article
@karanshah137 and others - thanks for insights. Although I did not buy any stock for data center boom, instead had that much before all talks around data centers…because of momentum investors entering it has run up too much last few years. I have full confidence in the business & management of the company but not so much on momentum investors fleeing away the moment they find something else or growth normalizes a bit. (Trent example, which fell almost 60% from peak inspite of good results because it was not so good enough to keep the extended valuation & hence momentum alive).
So, in all these data center stocks, how to ensure not to end up like a Trent (not meaning to complete exit but at least be cognisant of the fact, and hence ready, that how much is the froth. I underestimated it by a huge margin in Trent, thinking it to be around 30% to max 40% when it was 50% greater at 60% froth), where management & business is excellent, results have been decent but stock fell 60% because of valuation derating? Even if someone is looking to add to their position, underestimating the froth may result in starting to add the dip too early during such massive resets and ending up in losses in new positions. Knowing the froth right may help not just in profit booking/complete exit but even someone who intends to remain a long term investor and build on to existing position during massive resets.
One criteria maybe order book growth or maybe execution timelines/delays, margin pressures etc. but how to read all that to map them to a right valuation these companies must have amid all this frenzy? In short, what must be a timeframe and/or markers when all this frenzy starts/begins to end and what must be the terminal valuation of these companies when it actually ends? OR is there any other way to look at it or judge it?
Disc: Invested in Trent (transactions recently) and Hitachi Energy. Not a buy/sell recommendation. Not eligible for any advice. Post only for learning purposes. I can be wrong in all my assessments.
3 Likes
This will always be an issue; there is no escape from it. Unless those data center stocks are in FnO, we cannot hedge the risk. But most data center stocks focused on the theme will always command such valuations.
Yes, I agree; I myself entered late in peak valuations of Anant Raj, and after the Deepseek moment, it couldn’t recover despite great numbers, but I’m still holding and averaging at dips and increasing my position, as I’m sure it can again fall 50% from highs, but in the long term, it will be fine.
I noticed in Trent and Anant Raj, people who found early and sat tight are still happy, as the underlying EPS is still doing fine; it’s just PE boom-bust cycles. I added Dixon like that around 3000 levels, and it went up to 19000 and now is around 10000, but I’m fine holding it, as my buy price is excellent, and I kept booking a few quantities, which made my remaining quantity free. So this approach can be done.
I’ll share what I’m currently doing. Track the order book to execution gap. As you could see in Schneider Electric Infra in Q2 FY26, it was not great, but in Q3 FY26 they came back strong.
Another thing I started doing is tracking hyperscalers like AWS/Google when they start talking about cost optimization or extending server life. So if they slow down their spending, the suppliers (Hitachi Energy, TDPS, etc.) will see a collapse in new orders 6-12 months later.
Those are early signs.
When companies from unrelated sectors, for example, let’s say a textile firm suddenly pivoting to data center EPC, start entering the space to capture the valuation premium, the froth is at its maximum.
If you read this article, I think once the data center capacity in India reaches a saturation point, for example, the 10 GW baseline mentioned in the Sify article, these companies will likely settle into a capital goods/utility multiple (20x - 30x P/E)
Instead of looking at only P/E, look at the enterprise value per MW. If it costs X Rs. to build a data center from scratch, and the stock market is valuing a company at 3X per MW of capacity, that may froth sign but it should more be treated as caution for Lumpsum or Partial Profit Booking Sign
(Disc: Invested in Stocks Mentioned. Not a BUY/SELL RECO. Sharing what I learned.)
10 Likes
@karanshah137, you should start a thread with your portfolio and thoughts. Will be helpful to the forum
I too believe southern India has become gold due to bombing of data centers in west asia. It is far away from our notorious neighbouring countries.
3 Likes
Nice article on what over investment in data center/ai can result into…indeed its a thin edge to run on….from current situation it looks like while overall capacity growth would be there over medium term but individual companies can face specific challenges….
I maybe wrong in my assessments
The snapshot mentioning varies sectors and underlying companies working towards data center is very informative. Not seing the name from Tatas though - TCS announced signficant entry & capacity into data centers and sister company Tata communication is already in tie up with STT GDC with DC assets in India & earlier in Singapore as well….interestingly they divested majority back in 2016 & 2019 to focus on the digital/software/fabric part….post covid, I am not sure with recent boom there strategy would be again to refocus on assets ( a mix strategy) or remain fabric focussed ….in any case they seem to hold significant experience and skin in game of data center business but still this well known name is not much talked about when data centers are discussed…. @karanshah137 and others woukd be good to know your thoughts on Tata communication specifically now with respect to huge investment focus by TCS on data centers…..Thanks!
Disc: Invested hence biased. Not a buy/sell recommendation. Not eligible for any advice.
@gaurav I appreciate it; that means a lot.
And yes, I do plan to share more once I’m a full-time market participant. Currently, being involved with a family office, I prefer keeping portfolio specifics private, as it wouldn’t really do justice to them due to conflicts of interest. But someday for sure will start a thread with my portfolio and thoughts
3 Likes
Yes, you are right, but due to TCS being more service-led, I’m away from it. And Tata Communications, I mentioned in a thread above.
Tata Communications has always had the infra + experience, but the real question is strategic direction: whether they go back to asset-heavy DC expansion or stay focused on the digital/fabric layer.
Also, TCS stepping in makes it more of a group-level bet now. How they divide roles and allocate capital within the Tata ecosystem will be the key thing to watch.
This is something we need to take care of, and we should only build green data centers; otherwise, our fate will be the same.
PRESTIGE GROUP, IRFAN RAZACK SAYS
Data centres present a huge opportunity, with multiple MoUs in place
Mtar is supplier of Bloom energy.
1 Like
Global data center power demand is set to accelerate:
Data center power demand is expected to grow +220% from 2023 levels, or +905 TWh, to a record 1,350 TWh by 2030, according to Goldman Sachs.
This is up from the previously expected +175%, driven by higher AI server shipment projections and increased deployment of more power-intensive servers for AI processing.
~60% of this growth is expected to come from the US, up from ~50% in prior estimates.
As a result, US data center power demand is set to reach ~750 TWh, followed by the rest of the world at ~600 TWh.
Furthermore, US data center capacity is projected to rise +197% between 2025 and 2030, to a record 95 gigawatts. The AI power boom is accelerating.
Talking about the Indian context:
Market to grow from ~$10 Bn (2025) to $22 Bn by 2030, Current capacity ~1.4–1.6 GW (164 facilities), Expected ~1.7–2 GW by 2026, and Long-term expansion to ~4–5 GW by 2030
The most important build cost for us is $6–7 million per MW, which is globally competitive.
Read This:
India’s AI boom is becoming an energy problem, not a tech one
This is what the biggest headwind is for India. While everything looks promising, we also need to take this in account as we know history of our Country and Focus more on Green DC only
