interesting that in this current round of market support 50% of national team money has been spent on Star-Chinext vs just 7% in the 2024-25 round "we are going to need a bigger boat" tho

* no AI model * faulty new folding hardware * 20% revenue exposure to China deflationary consumer market * CEO ignores interview requests from big brain tech podcasters Apple reclaims #1# world market cap.

The grass is always greener on the other side? We live in absolute clown car world. Americans in angst "we must win the AI race against a state directed economy that can build datacenters wherever they want". Chinese in angst "we have 1/10th of US compute power and we use it to make AI romance revenge slop microdramas while americans use it for research coding and automation." Same story with Tiktok when it was controlled by Bytedance. Americans thinking it was being used to feed slop to US kids while somehow the China version was some sort of educational content fountain for Chinese youth, and not the actual combo of short skirts face filters and farmers doing weird stuff. No one has any idea whats going on the other side just keep projecting own fantasies. Actually, probably no one even has any idea whats going on in their own side 😂

Dark days ahead for west coast VC professionals coming into contact with real life, but I am confident they will figure out a way to put aside their vegan + animal shelter donor traditions in order to give gazillions of money to hundreds of AI drug discovery start ups whose business model is to generate 1000 AI slop molecule candidates a day to be injected into monkeys and see what happens and "feedback learning to train the model". Show me the incentive I will show you the outcome.

China bank results for the smaller regional banks (city commercial etc) more mixed, but ma-boi Ningbo continues to deliver double digit growth YoY, up 2.5% QoQ, mostly relying on Jiangsu and Zhejiang export machine. NIM margin flattish, growth came from assets up 14% (corp loans up, retail loans flat) and fee income (wealth/asset management) up 25% (in this economy..). The QoQ growth rate in loans, deposits, and fee income has slowed down in Q3 though. Watch. NPL ratio flat, expense ratio flattish. They lost some money on FX derivatives, which probably tells you about the customer base. Yes the bank will not have fun if the export machine slows down. So with dd NI growth, ROE 14%, yours for 0.95 P/B 7.2x PE 3.9% LTM divi yield. For reference, big SOE bank chad CCB is 10% ROE 0.6x P/B 6.4x PE 4.8% divi yield. I own both, 15% of the stack. Maybe should be 30% these, 30% healthcare, and the rest cash.

Imagine the buyers of MBS at 3% back in 2021 thinking they will get refinanced in 5-10 years, and instead today they are holding 25 years paper at 3% vs rates at 7.5%. If they need liquidity I have no idea what pricing but maybe like 65 cents? Now imagine if the other half of their portfolio is private equity. Man it is tough.

uuuh Mindray 🧐 whatsss going on here? In this market?? Choose your fighter The technical guy: another head fake like in summer 2025, tho back then the index was massive tailwind while today this is a classic salmon upriver pattern The factor flows master: or just moving up in sympathy with TMO and Danaher (which btw are completely different business, but we live in an "everything retarded computer" factor flows market) The fundamentalist: Results remain garbage The second-derivate enjoyorrr: Yes, results are garbage, but the rate of worsening has decreased The asia hand policy connoisseur: well you know they are getting on with the government program to reward minority investors, this year started quarterly divis (rare in a-shares) and bbs

For all alibaba baggies (me too) growing cloud at 50% is very meh when google is doubling. Being "capacity constrained" is not a good excuse when at the same time you seem to be "price capped".

Some suggestions for whoever in Spain is in charge of this high speed train mess. The Wenzhou high speed train accident un China back in 2011 was similar in form (rear ending) cause (fault of equipment and then humans botching up) and outcome (40 dead). No big accidents in China since the Wenzhou one, 15 years ago, and with 100x the traffic (fingers crossed, at some point it is a matter of probability tho, but still). Report and responsibilities out 5 months after the accident (prior accident in Spain was like 9 months ago and still nothing?). I havent read the report, it is not my job, but it seems 54 people blamed and penalized, top to bottom, from the boss who sets the rules, to those that executed poorly, to the junior employee that failed to file a small fault report. Drivers cleared. Whoever read the report can correct this. It is probably not too off tho. (Maybe some names of uncomfortable people were dropped into the report for good measure, who knows but hey perfect is the enemy of the good no?) I defer to a sociologist to answer the question of whether "Your career is over if something bad happens even if you are not the direct cause but you are in the chain" can be a motivator for people taking their job seriously, instead of expecting a political figure above to resign and clean all, or put the blame on a machine or lack of budget.

Pharmaron next (a mini Wuxi Apptec, slightly rougher). 3Q26 fine too, though: --no increase in FY guide (still 15-20%) --no increase in purchase order growth vs 1H26 (+30%) Q3: --revenue up 20% --NI up 5% with FX eating a lot of the profit growth (ex FX up 24%) Yours for 18x EBIT 23x PE. Similar multiples to Wuxi, but much lower margins (15% vs 45% for ebit). You only own pharmaron vs Wuxi if you think in an upcycle it has more torque on earnings (NIM and EBIT% has gotten close to 18-20% in prior upcycle). Interesting that revenue from domestic pharmas is growing at 42%, faster than the rest. Still 60% of customer based is US pharmas (served from mainland). Chart below. No mention of AIDD in the release, maybe they said something in the call I haven't checked. This is what they had to say in summer: --Pharmaron sees rising orders from AIDD biotech companies, which now contribute 7-8% of its lab services revenue. --These platforms can generate molecules at scale (most against known targets), but limited high-quality experimental data still constrain their ability to predict drug-like properties. --Management sees little risk of AIDD disrupting the traditional CRDMO business model. Advances in computational drug discovery should expand the pipeline funnel and drive more candidates into development.

Wuxi Apptec just reported 3Q26, at a first glance all seems fine: --FY revenue growth guide moved up from 37% to 41% (expected I think) --Backlog up from rmb66bn in June to 77bn in September (up 36% YTD, from up 25% 1H26) --CF up 70%, capex guide up, EBIT margin flattish 45%, blah blah.. --BUT...MOST important...and after keeping silent quarter after quarter...we finally get an 🥁 "AI drug discovery" mention in the report. Wow. They really can keep up with the times. Imagine also Wuxi management this summer watching Genscript share price moon thinking "seriously? these guys?" Actually we get two: AIDD and AI4S. Check out the prose, it's exquisite: "We efficiently fulfill the rapidly growing demand from global AIDD/AI4S customers for dry & wet lab closed-loop R&D, further solidifying our industry leading position. We continue to enhance AI and automation capabilities, enabling multiple use cases in pharmacology, pathology, and beyond. In response to the needs of AIDD/AI4S customers, we provide high-throughput, diverse, customized, comprehensive, and integrated wet lab services, fostering novel partnerships with high efficiency and synergy."🥲 What does it mean? No one knows what it means, but it's provocative... it gets the people going... let's see how this mixes with a gigabear market out there yours for 17x LTM EV EBIT 22x LTM PE