We are entering the year 2026 and one thing is clear: anyone who still labels semiconductors as "cyclical" has missed the biggest structural change of our time.

After a massive growth spurt last year, the global chip industry is now on the verge of breaking the magic barrier of USD 1,000,000,000,000 (1 trillion) in annual sales. For comparison: that is almost a doubling within a few years.

Why 2026 is the "year of decision":

• From training to inference: while everything in 2024/25 was dominated by training large AI models, 2026 is the year of inference. AI moves to the end devices (edge AI). This means that we don't just need Nvidia H100/B200 monsters in data centers, but efficient chips in every smartphone, car and toaster. 📱🚗
• The 2nm milestone: TSMC and Samsung have fully ramped up mass production in the 2nm area (GAA technology). The efficiency benchmarks of the next five years will determine who is ahead here.
• Memory is the new gold: memory chips (DRAM/NAND) are no longer a commodity. Thanks to HBM4 (High Bandwidth Memory), memory has become the bottleneck for AI - and thus the cash cow for players such as SK Hynix and Micron.

My view of the market:

The industry is growing up. We are seeing a decoupling from the classic PC and smartphone cycle. Semiconductors are the new infrastructure asset class.

My watchlist for the start of the series:

The Dominators: $NVDA (-1,99%) (Nvidia), $AVGO (-3,16%) (Broadcom)

The enablers: $ASML (-2,27%) (ASML), $2330 (TSMC)

The comeback bet: $MU (-4,81%) (Micron - because of the HBM4 supercycle)

💡 What do you think? Have we already seen the top in valuations, or is the "$1 trillion supercycle" just beginning? Let me know your opinion in the comments! 👇

#investing
#semiconductors
#tech2026
#stocks
#künstlicheintelligenz
#chips

Goldman Sachs has looked into the issue and is convinced that it will be companies in particular that have extremely high personnel costs and, at the same time, great potential for automation.

The Goldman strategists use very specific figures. They take wage costs as a percentage of turnover and combine them with the proportion of jobs that can be automated using AI. The result is the so-called Phase 4 list of the AI revolution. AI hyperscalers, semiconductor values and data centers correspond to phase 1, phase 2 and phase 3 respectively.

Which sectors are phase 4 candidates? The answer is: software & services, professional services such as management consultancy, tax consultancy, insurance and IT services, because AI loves repetitive knowledge work.

The specific winners in the area of Banking & Finance include Bank of America $BAC (+1,13%)KeyCorp $KEY (+1,04%), PNC $PNC (+1,08%), US Bancorp $TBBK and Truist $TFC (+0,88%). Half of banking work is document processing - that's AI fodder.

For insurers Goldman Sachs calls Aon $AON (+0,73%)Marsh & McLennan $MMC (+2,69%), Willis Towers Watson $WTW (+0,71%). When AI checks insurance claims, it is a cost revolution.

In the area of IT service providers & consultants the list includes Accenture $ACN (+2,88%), Cognizant $CTSH (+2,19%)EPAM $EPAM (+0,84%)IBM $IBM (+0,44%), Twilio $TWLO (-3,54%), DXC $DXC (+1,17%) and SAIC $SAIC (+3,77%). These companies have consultants - and a large part of their activities can be automated with AI.

Among the software-specialists, Guidewire $GWRE (-4,71%), Manhattan Associates $MANH (+0,68%), Pegasystems $PEGA (-2,87%) and Tyler Technologies $TYL (+2,11%) are mentioned. These are companies that build software for the management of companies - and AI will allow them to deliver their products faster and more cheaply.

In the area of Healthcare Services & Life Science Tools Goldman names Labcorp $LH (-0,91%)IQVIA $IQV (-0,31%) and Certara $CERT (+1,47%)because AI is already involved in medical documentation, study evaluation and diagnostics.

What makes Phase 4 shares so exciting? Every working hour saved has a direct impact on the margin. High productivity stands for high profit leverage. The advantage: unlike hyperscalers or chip companies, they don't need huge investments, but can simply use AI. This means less risk, more leverage and a lower valuation.

Conclusion:

Currently everything revolves around Nvidia $NVDA (-1,99%) , AMD $AMD (-2,1%), TSMC $2330 - i.e. phase 2 stocks. But the next wave is not coming for companies that use AI.

Source text (excerpt) & graphic: Welt, 03.12.25

I realized that my portfolio is very tech-heavy and has been underperforming the market the last few days. I thought I would ask several AIs (for fun) how they would adjust my portfolio to remain growth-oriented but with more diversification and some stability.

Here is the list:

Tech:

6% $NV

5% $GOOGL (+1,2%)

5% $AMZN (+2,04%)

4% $MSFT (-0,88%)

5% $ASML (-2,27%)

5% $2330

4% $CSU (+2,06%)

5% $PANW (-0,99%) (new)

4% $CRWD (-2,9%)

6% $INTU (+0,32%) (new)

4% $CRM (-0,36%)

Industry:

3% $ETN (-0,09%) (new)

3% $APH (-2,14%)

3% $HEI (-0,46%)

Health:

4% $NOVO B (+1,77%)

4% $LLY (-1,53%) (new)

3% $ISRG (-1%) (new)

Finances:

4% $V (-0,51%)

3% $NU (+0%)

Energy/commodities:

6% $LIN (+2,02%) (new)

2% $NEE (+1,14%) (new)

Infrastructure:

3% $WM (+0,86%) (new)

Consumption:

9% $COST (+3,31%) (new)

The Ki:

BYD, Xiaomi, Iris Energy, Hims and Hers, Sofi, Applied Digital, Coreweave, Ondas, Rocket lab. Quasi China and all the hype stocks

However, AI has been arguing with itself. It thought Crowdstrike should be replaced by Palo Alto and, if Palo Alto was in, it should be replaced by Crowdstrike. The same with Novo and Eli lilly, visa and mastercard, salesforce and service now.

What do you think of the AI "optimized portfolio"?

I bought it in May from $2330 and things are going well so far. Currently up 60%... But since I couldn't invest as much as I would have liked at the time, I ended up taking a warrant with 8 leverage (yes, I was very convinced). Of course, it has now gone through the roof. Would you take profits here or let it run its course?

Thank you very much for your opinion!!!

If we look at the current PEG ratio, this is a very bullish sign for $GOOGL (+1,2%)
$IREN (+4,37%)
$SOFI (+2,77%)
$PYPL (-0,33%)
$2330

PEG Ratio: P/E/expected earnings growth

if this value is <1 the company is undervalued according to Lynch!

My thesis:

In a Taiwan/China conflict, Trump would never support Taiwan and therefore $2330 (TSMC) would be doomed, which is why the government recently invested $9 billion in $INTC (-1,56%) invested. In addition, Trump could have put pressure on Nvidia to have a chip manufacturer in the USA, which is why $NVDA (-1,99%) is now also investing $5 billion.

See Trump's contribution:

If it's a security play, Intel could be exciting if the investments are sustainable...

How do you see Intel? I'm staying on the sidelines as the company now has to develop with the money and show whether it can achieve the turnaround.

I would like to start a completely new portfolio that will primarily revolve around dividends.

As a core I was thinking of $TDIV (+1,31%)

Would you say this is a good core?

If not I would add $VHYL (+0,59%) add.

Additionally I would like to have a CC ETF as a kind of support, probably $JEGP (+0,34%) and or $SXYD (-0,36%)

I would like to represent the NASDAQ with $EQQQ (-0,26%) but I will represent it with $ASML (-2,27%) and $2330 will be added.

Allianz $ALV (+1,26%) and Munich Re $MUV2 (+1,31%) I definitely want to include, but they are too expensive for me financially, so I was thinking of the $EXH5 (+1,4%)

Oil shares are represented by $VAR (+0,93%) and one more.

Do you have any recommendations?

I am thinking about $CVX (+2,79%)
$EQNR (+2,29%) and $PETR4 (+1,43%)

I would also like renewable energies, but I'm not familiar with them.

Do you have any suggestions?

Becoming a defensive company $ULVR (+0,5%)
$D05 (-1,53%)
$O (+1,44%) and of course $NOVO B (+1,77%) Being.

$BATS (+2,14%) I already have in a portfolio, would it be too much of a lump to add $MO (+1,94%) to add to it?

I still have $KHC (+1,87%) on the watchlist but the split is not going so well, would it be wise to start with a savings plan?

Apart from that $RIO (-0,07%)
$NKE (+2,82%)
$1211 (-0,19%)
$SOFI (+2,77%) and $HAUTO (+0,56%) will be represented with smaller positions.

What is your opinion?

Would you improve anything?

What else would you add, especially in EE and defensive stocks?

Feedback is very important to me here, so far I have just been wandering aimlessly around the stock market without a fixed plan and strategy.

This is my first attempt to build something serious.

Greetings to all Getquins out there!

$ASML (-2,27%)

Quiet here, but ASML's share price has been rising for weeks.

Why?

J.P. Morgan declared that the worst quarter of ASML NV (AS:ASML) is behind the company and confirmed its status as a top pick.

It cited an improvement in order development, strong demand in the logistics and storage segments and a clearer growth path for 2027 as reasons.

After a disappointing second quarter report, in which ASML stated:

"We continue to prepare for growth in 2026, but cannot confirm it at this stage",

the news situation for the company improved in August and September, according to the brokerage house.

Several developments have reduced uncertainty over the summer. As no tariffs will be imposed on shipments of semiconductor equipment to the US,

"uncertainty around US-related orders is likely behind the company".

ASML is expected to receive orders for its planned US capacity, while Nvidia's AI-related spending,

$NVDA (-1,99%)

Broadcom

$AVGO (-3,16%)

und (-3,16%) Hyperscale

$GPUS

-Kunden continue to be strong. TSMC,

$2330

ASMLs largest customer, reported year-on-year sales growth of 37% through August 2025, and analysts expect this trend to continue for the rest of the year.

The expansion of the 2-nanometer process at TSMC and the expected acquisition of the A16 process by Nvidia are expected to increase the demand for EUV exposure and bring ASML higher orders in 2026 and especially 2027.

The memory market also looks promising, driven by demand for high-bandwidth memory solutions.

DRAM prices remained strong, supported by continued AI-related demand, and J.P. Morgan indicated that the tight supply and demand environment is expected to continue until new greenfield capacity comes online in 2027.