Phreak's Thoughts, Ideas and Opinions

[phreakv6]

@Faiz_ahmed - This is a question I have been asked several times. I will try to explain best as I can. I don’t have some magical mental model that works everywhere. Its usually just the output that shows up in threads like these or on twitter which makes it seem like there is some kind of method. The short answer is that there isn’t. I turn a lot of stones and because my capital is small and I want it to perform the best, I am extremely picky on what I choose, even if several ideas seem very interesting, I make it a point to rank and yank (look up stack-ranking if not familiar) - we work best when working under such constraints even if they are made up. I realise I have almost no fomo when I reject something consciously or when I never made the effort to study something. Its just the way I am wired and that helps a lot. I try not to chase breadth - I am an investor and not an analyst (the distinction is very important). I have serious regret though when I like something but I am too lazy to act (Aegis for eg. I had done all the work in Feb and I loved how cheap it was but never bought since I liked the DC plays better - but in May, I had a good opp. to switch but I didn’t)

Ideas come at me from all sides, as I am sure it does for everyone here. I try to see if tailwind+tam+moat+valuation is satisfied (at least two or three). Very rarely all 4 click, especially in today’s market where everyone is using similar filters and are all having access to powerful AI models. Sometimes tailwinds are clustered, so that helps in looking at other plays for same tailwind - classic example recently is AI/data centers. I was little early to Claude code (started using it last Aug/Sept I think) so by Nov, I had become shocked at how good things had become, so that exposure made buying into AI stocks easy. So when I came across two ideas Aeroflex and Mtar (both incidentally from friends), I could not see them as separate unrelated ideas. I dug deeper post that into memory, power, cooling etc. in Feb using some good resources from semianalysis which convinced me I had to bet big on these things.

So a good understanding of the tailwind and moat is very important and this can only come from a deep understanding of a sector - be it CDMO or AI/DC plays or Power transmission ancillaries. So that is roughly my process. I try not to buy into stray stories that do not fit into my framework. I do make exception from time to time when valuation is very juicy or growth is not ignorable though. It helps me break out of the mould and explore that way. Buying into Bluejet last year for eg. was a way for me to break out of the mould to understand cdmo as a sector. That bet did not work out for me but it led to Sai. There are numerous stories I have given a pass since I did not have money but they went berserk (HFCL and STL tech, E2E - all around March). Several I have studied in depth but rejected (Acutaas - because of single molecule risk after having been bitten by Bluejet. Turns out market can turn a blind-eye when numbers are coming - same thing is happening in Navin Fluorine and Anthem also and PI Inds in the past. Market simply doesn’t care about the concentration risk until things break. Or maybe market chooses to stay ignorant). When an idea I work a lot on only to reject does well, it still makes me feel good because it means I am fishing in the right areas.

Whatever I am trying to say has been said in the past - its the concept of circle of competence. Try to stay within it (CDMO/Data centers/Power for now) but keep expanding it (Oil & Gas). Don’t chase things you don’t understand. If you avoid drawdowns, your returns will be spectacular. To avoid drawdowns, stay within what you understand and never chase without having a sense of value. There are whole swathes of market I simply don’t bother to look at - Sugar, Cement, FMCG, Textiles, BESS, Solar, Media, Alcohol, Real Estate, Banks/NBFCs, Logistics, Bulk chemicals etc. Others can make money in it but I know I won’t. I will invariably sell out a week or two after buying because they don’t excite me (happened with Banks/NBFCs twice in the past for me). Avoiding such large swathes makes it easy for me to focus on the things that matter. As a retail investor with numerous interests outside of markets, I find this way of slicing and dicing the markets work well for me. I might find myself at sea when people with breadth discuss stocks but I am ok with that.

@harshaga - Its phreakonomics. Its my personal app and its not public.

@Heavy_luck - There are lot of overlaps between Venus and other companies. Only thing common with all of them is that almost all of them trade at 50-60 p/e or higher while Venus trades at 18x. Market seems to have been unexcited after the AGM where the management guided for 15% growth for FY27. I feel management is conservative and even a 15% growth I feel is great given the valuation because there were couple of one-offs in FY26 (PLI and forex) - so a 15% growth might be actually like 30% real growth and company converts cash really well unlike all the rest. There’s no point looking solely at growth when valuation is this cheap - so I am staying put. Lets see. I might regret but I know I won’t lose my capital here and that is generally enough for me to give a long rope to a business.

---

I have no fresh ideas and have mostly remained invested in same businesses. Most are at ATHs and market has done well. I have bought more Vivid Electromech (switching from TD) whenever it has dipped under 1500. I think among DC stocks, its the one which is still trading very cheap (25x 1 yr fwd) while a peer like Marine electricals is trading at 93x. Maybe its due to SME illiquidity or because people don’t know/understand the business. Marine on paper has a bigger order book but they do entire Power Delivery Systems which include a lot of bought-out items while Vivid does only the PDU and so will be much higher margin (should be close to 30% if my back-of-the-envelope calc is right) and the PDU contribution should go up in FY27 going by the orders they have won in the last 4 months from CtrlS, STT Global. Unlike most other DC plays, Vivid already makes 50% from data centers and so can grow much faster riding on the DC tailwind. With the new capex coming in, they should have enough capacity to grow 40-50% for next 2-3 yrs without incremental investments.

Found this on IGGL twitter. Along with the DNPL/IGGL pipeline which is expected to be functional by Dec '26, there is one more pipeline also underway (Duliajan feeder line) which will provide sufficient access for all future scaleups of North-east small gas fields. The construction for this is underway in Aug and is expected to complete next year.

This again I feel is very under-appreciated as this sector itself is not covered by any institutional analysts. The way I see it (stupidly maybe), is like a large tract of land that sees its value go up when a state/national highway passes by it. So far, all these gas fields (esp. Oilmax ones) are valued like an optionality because evacuation wasn’t easy. With evacuation, all these fields should go up a lot in value next year. Because unlike the DNPL/IGGL pipeline coming up end of year which has 2.5 MMSCMD (DNPL expanded from 1 MMSCMD recently) capacity, DFL will give capacity to evacuate 6.7 MMSCMD

It will improve the realisation these fields can get for their gas as they don’t have to sell within Assam and can sell in UP or WB. Its like a village that suddenly can export its produce to the nearby town or faraway cities after having access to a national highway and thus fetch far higher prices than what it will get within the village. That’s what excites me about this. Oilmax has several fields in Assam and can benefit disproportionately once this feeder line goes live. The merger NCLT hearing is scheduled for Sept 4th as per latest update on NCLT site (its procedural in nature and should happen sooner or later).

Disc: I am a novice writing about what I do. None of this is advice.

[yogendra_kumawat]

Asian Energy Services Ltd:

KUIPER: manpower-supply business very low EBITDA ~ 7% important as how ASEL transforms this business to a higher margin like cable link laying and offshore infra possibly ASEL will able to achieve 11-12 % margin as guided. they have now access to the west Asia geographies this can actually play a important part in providing “end to end field solutions” may yield the margins. However, they haven’t clearly specified how Kuiper will be managed in terms of its management structure.

Oilmax

Tiphuk: won under DSF-3 (contract AA/ONDSF/TIPHUK/2021), originally discovered by ONGC in 1985. ONGC drilled 4 wells — one produced 30,000 cm/d until 1999 before being shut, two are workover-ready, two need workovers. So this wasn’t a dry/low-potential field — it produced commercially for over a decade before being mothballed.

Duarmara: contract AA/ONDSF/DUARMARA/2016, discovered by OIL in 1970, put into production the same year. Multiple oil/gas/condensate zones confirmed across 3 legacy wells. This is a genuinely old, previously-producing field though the “tightness” issue flagged on the call may create problems.

Amguri: not a DSF block — it was picked up from Canoro Resources (a defunct Canadian operator) via a separate deal, dormant for 11 years before Oilmax revived it in 2021. previously-producing reservoir mothballed for non-geological (operator exit) reasons. It does have potential but Amguri and is already producing but is near to the Assam ONGC fields , Gas migration/ water coning may cause issues or reservoir heterogeneity may cause problems for them.

Small DSF fields/blocks that had been given up by OIL/ONGC due to logistical issues or may be due to low potential latter could be the main reason of ceding these fields. You may not want to give away your gold, but what you consider copper :) could be someone else’s gold.

These fields may have been tested with oil/gas but have issues like tightness and reservoir heterogeneity which is common in oil reservoirs. This uncertainties poses a big risk in order to improve production. normal Testing or workover operation may take max 2-3 months if there has been issue of fishing or string breakage it can take more and rig cost per day is very high if successful there should have been a announcement till now.

Duarmara — management said they’d have clarity “within a month” as of May 20, the fact that it’s still not been announced by early September.

Indrora/Mevad —

well NM-01 at Mevad: Drilled to 1,650m, hit three separate hydrocarbon-bearing sands (Mandhali, Sobhasan, Kalol) Currently producing ~100 bopd from just the Sobhasan interval alone Estimated peak potential of 125–130 bopd from Sobhasan alone, with the Kalol-III zone showing hydrocarbon indicators but not yet tested if the company is already stacking multiple pay zones in its best well and still needs 6 more wells to reach 1,000 BOPD (vs. current combined ~200), the required average productivity per new well is still demanding — few dry or low-potential wells among the six would meaningfully dent the FY’27 target.

Drilling a new well without complications causes a investment of around 30-50 Crs depends on the depth of reservoirs so any dry or low-potential well can cause issues.

One macro data point that is important: Q1 FY27 already came in very strong The order book is encouraging at 1750 crs. and the business is very well- diversified apart from oil and gas to upstream projects management and that could de-risk things and the oil and gas sectoral tailwinds can do wonders for this company.

They have given aggressive targets in terms of oil production that may be difficult to achieve due to these reasons.

Disc: These are just my personal views and observations, not investment advice.

[Garuna]

Hi, thanks for this splendid thread.

On single molecule example (BLUEJET) and concentration risk, we can perhaps extrapolate this and say some of the companies from other sectors also have this problem, eg, MTAR’s single molecule parallel is single client, Bloom. Yash only makes Bushings, and someone else can come in and start making them .

[phreakv6]

@garuna - I don’t think its an equivalent comparison. Single molecule becomes a risk specifically because

1. De-stocking risk is real even in blockbuster molecules with mind-boggling growth like bempedoic acid. This of course is a minor risk in my personal opinion if I were an owner of the business as long as molecule is doing well and I don’t lose the customer. But as a shareholder it means sitting through several quarters of de-growth with no clue if the molecule has reached peak sales (NDAs will prevent discussing the molecule and sometimes CMO simply isnt clued in) or if there is a secondary or tertiary supply taking incremental volumes which brings us to the second risk

2. No innovator wants to go with a single source and will always diversify but might give lower volumes to the secondary if the primary is very reliable, in a reliable geography with great supply chains and has strong balance sheet. There is also the cost angle where someone with better processes/supply chains can manufacture for cheap which shifts market share.

3. While there is decent moat for API suppliers, for intermediates the moat is less strong. Even for API/formulation players, for most molecules it is not hard to switch if the innovator so wants. Moat is strong but not impenetrable.

4. Bigger risk sometimes comes out of competing innovator molecules and also from generics when drug goes off patent. After having studied this sector in some depth over the last year, this is more common than it seems. BA for eg. has competition now from Obicetrapib and Enlicitide (while injectable PCSK9 like repatha and praluent were expensive, Enlicitide can be a real risk). Sai’s migraine drug atogepant took serious market share from rimegepant (at least as per UK data where atogepant and crossed rimegepant sales). Another one of Sai’s drug lorundrostat which should get approval by end of year faces competition from another Astrazeneca drug called baxdrostat with both going almost head to head at launch and so on. Acutaas’s darolutamide faces risk from competitor’s drug going generic (enzalutamide).

So you can see how risky a single molecule reliance can be for a CDMO because it faces cost competition from other CDMOs, other innovator molecules, other generics, other innovator molecules going off patent cliff, supply chain risk and de-stocking.

Now coming to the comparison with Mtar and Yash. Mtar derives bulk of its growth from single customer Bloom today making hotboxes of SOFCs. There is only one known competitor (Kaori heat) and Mtar’s association with Bloom goes back 15-20 yrs and this product is not easy to switch to, to other precision engineering players. Mtar is being in the right place at the right time kind of play which happens out of unforeseen luck (that’s how I see SOFCs being cost viable suddenly for data centers). There is no de-stocking on near-term peak risk here either for another 2-3 years. There is however risk of competition from gas engines and so on but tailwind is so strong that everyone can thrive. So while there is risk, it is not in the same league as a single-molecule cdmo.

Yash’s bushings is perhaps an even weaker comparison as a single-product dependency. That’s like saying transformer companies make only transformers and so its a risk. Of course others can make bushings but its not straight-forward because there are only 9 notable names in the entire world that make bushings and qualification requirements and small market size keeps others out. Transformer companies themselves can make bushings but they would still have to import the cores and the fixed costs are high in this business that a player doing volumes will always have serious operating leverage, so making bushings is a hard sell (Yash making 10-15k bushings a year vs transformer player making 500 bushings for eg.). Yash is also backward integrating into cores which would help them do 30-35% margins in 3 years. No one else is backward integrating into cores and the technology is not simple and even fewer companies have the skill to make cores. There is no technology risk, de-stocking risk etc. and qualification to supply that takes years is a moderate moat which increases switching costs - also the fact that bushings is a very critical component that makes up 2-3% of transformer cost - why would anyone risk reputation? (esp when its going to be more expensive to make it inhouse)

In fact this business is so unique that I have found mutual funds buying into an SME for the first time.

[Aloysius_De_Sa]

Even for Bluejet, they do have a stable contrast media business and they are directly correlated with the growth in Xray and MRI in India and the RoW. With rise in incomes, increased longevity and increase in old age diseases, such diagnostic tools will only increase in use case. This becomes the cash cow business that will not simply disappear.

Their massive capex plans reminds me of Aether Industries that was languishing at 700rs for a long while till orders came and they monetized. I feel Bluejet is also at a spot where the management sees the demand (they spoke of a very large pipeline) and they are scaling for the same.