ZenFlow Finance

Savita Oil Technologies. ₹419. The street sees a legacy lube company. We see India’s data center cooling play hiding in plain sight. A thread. 🧵

Most investors ask: “Which stock will 3x?” Munger asked: “What’s the fastest way I go broke?” That one inversion compounded Berkshire at 20%+ for 50 years. Run it on any thesis — Excited about a CDMO play? Ask what kills it. USFDA alert. Single-molecule revenue. Promoter pledging. Bullish on a capital goods name? Ask what kills it. Order inflow slowdown + high debt + stretched receivables. That’s a value trap, not a value buy. Most HNI portfolios are built on forward optimism. Almost none are stress-tested against their own failure. Munger called it Inversion. Borrowed from mathematician Carl Jacobi: “Invert, always invert.” Don’t chase the win. Eliminate the loss. — At Zenflow Finance, that’s the first question we ask before we build a position.

@deepontweet Yes

@ZenFlowFinance I have read about Castrol parent capability of data center cooling.

Savita Oil Technologies. ₹419. The street sees a legacy lube company. We see India’s data center cooling play hiding in plain sight. A thread. 🧵

@deepontweet Castro did mention but their capacity is mineral oil and not ester. SOTL can be a beneficiary due to the ester capacities

@ZenFlowFinance Why not Castrol?

@Gautamg717 We could take a probable guess since the GPUs ( Rubin ) will come around 2029-30. If it gets approved SOTL might become a big player . Also ester margins are 4-5x of mineral oil

@ZenFlowFinance The fluid used for cooling is still hasn’t been approved and we don’t know how much more time will it take.

@sjain_13 The DC cooling and EV thermal management thesis is solid. But a thesis without revenue is just a story. First sale changes everything — it de-risks the narrative and re-rates the multiple. Patience is the edge here.

@ZenFlowFinance Have been holding for 2 years now for this. Till they make their first sale for DC or EV cooling, can’t really be sure

🚨 INDIA’S AI INFRA RACE JUST LEVELLED UP 🚨 E2E Networks has gone live with next-gen NVIDIA B200 clusters using NVIDIA Certified Reference Architecture. This is BIG. Why it matters: • B200 = NVIDIA’s latest Blackwell architecture built for large-scale AI training & inference • E2E is positioning itself as India’s sovereign AI CloudGPU layer • Target use cases: Agentic AI, healthcare, finance, manufacturing, agriculture • Built with high-speed InfiniBand networking for hyperscale AI workloads • Backed by NVIDIA Enterprise software + Nemotron open models India’s AI opportunity is no longer just about apps. The real moat is shifting toward: ⚡ GPU infrastructure ⚡ Compute capacity ⚡ Sovereign AI stack ⚡ Local hosting + compliance ⚡ AI cloud platforms While everyone is chasing AI wrappers, companies building the “picks & shovels” of AI infra may quietly become the biggest winners of this cycle. E2E is aggressively expanding GPU clusters across Hopper + Blackwell systems and positioning itself as a foundational layer for IndiaAI Mission-scale demand. The AI infrastructure capex cycle in India is only getting started. 🇮🇳🔥 #AI #NVIDIA #E2ENetworks #IndiaAI #CloudGPU #DataCenter #ArtificialIntelligence #Blackwell #GPU #IndianStocks

@LearningEleven This research was inspired by your tweet from May 6th, sirji 🙏🏻. That post sparked the thinking — we just took it further.

Bhai - This is your work, my contribution is zero. 🙏 x.com/zenflowfinance…

The triangle is a trap. Efficiency, security, and cost don’t trade off. They compound — in both directions. Cheap and unsecured doesn’t save money. It just delays the bill, with interest. The organisations that win on AI over the next decade won’t be the ones who optimised the triangle. They’ll be the ones who stopped drawing it.

@ZenFlowFinance One of the most overlooked aspects of artificial intelligence is the trade-off between efficiency, security, and cost over the next decade.

The street is paying 15.8x for what looks like a legacy industrial company. What they’re actually getting: a zero-debt, consistently dividend-paying synthetic ester manufacturer at the intersection of India’s power supercycle and the global AI cooling boom. Mahad Phase 2 is commissioned. Q4 numbers show the margin inflection has started. Valuation hasn’t caught up yet. This is the window. $SOTL. ₹419. Watch this space. Not investment advice. Do your own research. Built on research inputs by: @LearningEleven — At Zenflow Finance, we track what actually moves markets.

Risks you can’t ignore- What keeps me honest on this one: — Immersion cooling adoption is slow. Data centers don’t retrofit overnight. QuantiCool revenues could lag by 12-18 months. — Base oil prices are crude derivatives. Any geopolitical spike compresses margins before SOTL can pass costs through. — Transformer oil is sticky but not immune if global players dump inventory into India. Every thesis has a bear case. Know yours before you size the position.