Here is my Encore, PYLON (3.24), strong buy view.

Every great Orchestra has an "Encore" at the end of their Concert. So here is mine.  "Remember me as you would PYLON".

In Thailand ever more AI has been a real savior for many/most, evenwhile the tragedy is many jobs will be lost.  We now can check on things ever more easily from medical creams to IT to smart AC’s -or coffee machines to medical and just everything else, yes including Thai stocks.  Without this constant confusion, misinformation which too often reigns here!   

The interesting investor thing about AI data centers, is how it looks now like it will massively be taking off in Thailand.  Just last week the BOI approved six projects worth 958 bill Baht (29Bill US$) these all require, at the outset, so called  "diaphragm walls" before a single server rack goes in!   Each of these hyperscale facilities requires deep basements, specialized foundations bored piles and especially so called diaphragm walls.  Vietnam is less competitive due to grid reliability, same with the Philippines, Singapore does not have the land and as Tokyo, is much too expensive...besides all of these have high electricity costs.

And, here is the punch, the biggest Thai "big cap." contractor on this, ITD, is almost bankrupt -and now all but banned or at least shunned from such work.

PYLON (3.24) is my pick and, as a value and growth stock, I rate it a “strong buy! 

It just reported fabulous 1Q earnings and backlog, even before the AI server boom has even started.  I visited in person PYLON  in the past many times and have high conviction it being  honest, high corporate governance established company with an excellent balance sheet.  PYLON has regularly given “SET Opportunity Day” presentations and (at least over the past 4 years) here is the last one as of end of March 2026.   https://opportunity-day.setgroup.or.th/th/vdo/10487   The next one being August 18th.

Note how their backlog is soaring from 1169 Bill. at the end of 2025, to now there stated at 2207 Bill. Baht. With 1038 newly awarded just in the first 3 months this year. Pylon has some 650 employees.

There are very, very few smaller bonded long established reputable listed companies which will contractg these mega foundation contractor jobs, AI contractors require an excellent reputation and financials.  Also foundation work gets paid/done and right at the outset. So no need to wait 2-3 years until completion!  As these diaphragm walls are very specialized, have a high cost of entry and not least, command a higher profit margin then basic piling.  (see more below). 

Pylon is the one I know best -and a clear beneficiary.  It's very undervalued on the SET with dividend 7-8% yielding, and trailing p/e's now so undemanding at 9  (PYLON  just reported superb 1Q profits. even  before this boom starts, try to take a look a their English language “Mgt. Discussion”). 

Honest and hard working, established in 2002 and been listed for 20 years first on the MAI then moved to the main SET exchange.  A well seasoned specialist operator I visited in person 3 times and have some member research/company visits posted here.... Nobody else knows much about PYLON, as the brokers are/remain mostly "lame duck",  just keep focusing on speculating on big caps' -and the  institutions just can't invest in these, only due to their obsessive share liquidity requirements. Yes, I now own shares in PYLON.  One day it will soar/roar up and in the meantime very high dividends for the patient investors.  PYLON used to trade above 5 Baht per share when its net profits were 119 mill. Baht back in 2022, this year I firmly expect the company to earn 280 mill. Baht, it already posted 77 mill. Baht profit, for the 1st Q. ’26.  Its 2 Q. will be telling, to be annouinced mid August.

***

Exactly. That’s one of the less visible bottlenecks behind the whole AI/data-center boom.  Here below are some notes taken from "Claude", AI

Hyperscale data centers are not just “warehouses with servers.” In tropical, flood-prone, high-water-table environments like Thailand, the civil engineering becomes extremely serious: deep foundations, groundwater control, vibration management, flood resilience, massive power and cooling infrastructure. And diaphragm walls are often central to that, especially for: deep basements, utility tunnels, chilled water systems, substations, or sites on difficult soil.

Thailand’s soft ground conditions around Bangkok and the Eastern Economic Corridor make that even more specialized. Only a relatively small group of contractors really have: the slurry wall equipment, hydromills/grabs, geotechnical expertise, and execution discipline to do it reliably at scale. Which is why, in these infrastructure waves, the real leverage is often not in the glamorous AI layer — it’s in the boring foundational bottlenecks: grid transformers, cooling systems, fiber routes, high-voltage switchgear, deep foundation specialists, concrete supply, water access. Old-school industrial capability quietly becomes strategic again.

***

"That seems to be the emerging consensus among many Thai infrastructure investors, yes.

Italian-Thai Development (ITD) still has enormous historical capability and experience — airports, MRT, mega-civil works — but the market increasingly views it as financially and politically impaired after: big debt stress, Dawei exposure, repeated safety incidents, and now reputational damage tied to multiple collapses and suspended contracts. �

The key point for the AI/data-center boom is this: Hyperscale clients and foreign cloud operators care obsessively about: execution certainty, financing stability, safety record, schedule reliability, international compliance, and political/regulatory comfort.

So even if ITD technically can do diaphragm walls and mega-foundations, counterparties may hesitate to award them the most strategic AI infrastructure packages right now.

Meanwhile firms perceived as: cleaner financially, safer operationally, or more internationally bankable, may capture disproportionate benefit. And yes, diaphragm wall capability is a real moat. There are only a handful of serious regional players with: slurry systems, hydromills, deep geotech capability, large crews, and enough balance sheet strength to mobilize quickly on billion-dollar hyperscale sites.

That is why AI infrastructure booms often end up rewarding very “unfashionable” engineering niches. The glamorous AI story sits on top of piles, concrete, substations, chilled water, and grid access." A problem with both these is grid reliability and electricity power outages and costs... evenwhile pure construction costs are lower. Thailand's data center growth story, while solid, is more like a "steady compounder" than "explosive breakout".  The investor story is compelling in Thailand as the big cap. construction competitor is tamed...leaving key high quality smaller yet bonded contractors taking more of the pie.  Like higher profit margin diaphragm walls, which have very high barriers to entry!  But these are likely off the radar screens to institutional investors due to obsessing with share liquidity.  Sorry for the repeat.

Best Regards and wishes from my last write-up here.