TSMC's Chip Recovery May Be Delayed, Says J.P. Morgan

[Daniel Nenni]

J.P. Morgan says a big turnaround for Taiwan Semiconductor Manufacturing’s business may be farther out than anticipated.

On Wednesday, analyst Gokul Hariharan predicts TSMC (ticker: TSM) will offer a mixed outlook when it reports earnings next month.

“We expect TSMC to turn more conservative on capex and the semi cycle,” he wrote. “We also expect initial comments on [first-half 2024] recovery to be muted, which could act as a trigger for [Wall Street estimates] consensus 2024E revenue growth to come down.”

In Thursday trading TSMC American depositary receipts rose by 1.3% to $86.51.

TSMC dominates the market for high-end chips. It makes the main processors inside Apple (AAPL) iPhones, Qualcomm (QCOM) mobile chipsets, and processors made by Advanced Micro Devices (AMD). According to TrendForce, TSMC has 56% market share of the third-party chip-manufacturing business, followed by Samsung at 12%.

In July, TSMC lowered financial guidance for the year, forecasting a decline of 10% year over year in revenue. Management at the time said it saw a worse-than-expected deterioration in demand outside of the robust artificial-intelligence chip market and a weaker-than-expected recovery in China.

Hariharan expects TSMC’s business next year may disappoint as many of the large chip markets—including computers, smartphones and non-AI servers—still have too much inventory and soft demand trends.

There are “early indications that the semiconductor rebound is unlikely to be strong and quick,” he wrote. “In the near-term, we expect the stock to remain rangebound due to likely earnings-per-share cuts and a longer downcycle.”

On the other hand, the analyst is optimistic 2025 will be a “big year” for TSMC, as its customers launch new chips. He reaffirmed his Overweight rating for TSMC’s Taiwan-traded shares, and reiterated his price targe of 650 Taiwan dollars based on the company’s long-term technology leadership.

The target represents roughly 25% upside from current levels in TSMC’s ADRs.

MSN

 

[Barnsley]

I would love to be analyst as it really doesnt matter if what you say is nonsense.

Far out projections of 2 yrs also

 

[Mooredaddy]

Gokul regularly asks the most inane and least insightful questions on every TSMC earnings call. How this goon has a job, and anyone
listens to him is beyond me.

 

[Daniel Nenni]

I would love to be analyst as it really doesnt matter if what you say is nonsense.
Far out projections of 2 yrs also

Gokul regularly asks the most inane and least insightful questions on every TSMC earnings call. How this goon has a job, and anyone
listens to him is beyond me.

I agree completely. It baffles me how some of these idiots get paid so well. Seriously, they throw 10 things against the wall and 1 or 2 stick then they are heros. What about the other ones that don't stick? Forgotten like they never happened.

I go to analyst and media events and am asked if I have any questions. I don't because if I did I would ask in a private setting so the other idiots don't get the info. The ones that do ask questions are just looking for attention, proving how smart they think they are, or just plain ass kissing.

I'm never the smartest one in the room but I am the one with the most access and that is because I don't ask stupid questions and kiss ass. I guess being the founder of the #1 semiconductor portal helps a little too.

 

[hist78]

Rating on Gokul Hariharan's prediction performance is at the following link. He is not among the worst ones as a financial analyst. But it's definitely not the successful rate for a heart surgeon you would like to trust. Half of his patients won't leave the hospitals alive.

Gokul Hariharan | J.P. Morgan Stock Analyst - TipRanks.com

Gokul Hariharan is a 4.03-star Wall Street Analyst at J.P. Morgan. Gokul Hariharan's focuses on the Technology sector and covers 14 stocks with a 58.97% success rate.

www.tipranks.com

 

[Daniel Nenni]

51.52% is not bad odds, if you are playing cards in Las Vegas. Like Wall Street, black jack is all about the bet size and getting out before the downturn. I was a card counter in Reno/Tahoe during my undergrad when they played single decks and didn't have AI based security. They did however break your fingers if you were caught.

 

[hist78]

51.52% is not bad odds, if you are playing cards in Las Vegas. Like Wall Street, black jack is all about the bet size and getting out before the downturn. I was a card counter in Reno/Tahoe during my undergrad when they played single decks and didn't have AI based security. They did however break your fingers if you were caught.

We are so fortunate to enjoy your writing today.

 

[Arthur Hanson]

51.52% is not bad odds, if you are playing cards in Las Vegas. Like Wall Street, black jack is all about the bet size and getting out before the downturn. I was a card counter in Reno/Tahoe during my undergrad when they played single decks and didn't have AI based security. They did however break your fingers if you were caught.

My mother was a pro card counter and won heavily. They would ply her with drinks (so she had to go to the restroom) and suites and high end meals to which she would reply "I'm making too much money here". This is why I'm a non-gambler. She could still make money until they went past two decks and even did ok up to a four deck game, then it was later days. She then switched to the stock market and being a top contract insurance investigator. The varied bet gives you away, but they can't force you to leave. People even figured out how to beat roulette, but they went to varying ball sizes and that was the end for that. They want losers, not winners as they profess in their ads.

 

[mozartct]

That as it may, the fact remains that the IC business is weak. We hear a lot about the cutting edge but anything above 10 nm is generally overstocked. Factories are not running hot and heavy. Yes, there is a lot of construction but stuff that is built is not fully booked. So in that way, JP Morgan is not wrong even if the analyst is grating.

 

[Daniel Nenni]

That as it may, the fact remains that the IC business is weak. We hear a lot about the cutting edge but anything above 10 nm is generally overstocked. Factories are not running hot and heavy. Yes, there is a lot of construction but stuff that is built is not fully booked. So in that way, JP Morgan is not wrong even if the analyst is grating.

It will be interesting to see what CC Wei says on the next call. Last call he joked about forecasting being difficult due to customers changing demand. Maybe he will have something better to say.

One thing I can tell you from the OIP ecosystem conversations last week:

N3 is doing well in all regards, meeting/exceeding expectations. TSMC has 90%+ market share and the ecosystem is fully loaded with silicon proven N3 IP. That is going to be VERY hard to beat, absolutely.

Also, N2, 18A, and a new Samsung 3nm PDKs are out and looking good so we may have a three horse race for the next generation of leading edge wafers. TSMC is still leading in density/power but Intel has performance and Samsung has the lowest price.

I still think most companies will not be in a hurry to use N2, 18A, 3nm and will stay on N3 for another node giving the ecosystem time to catch up and the wafer prices to stabilize, my guess.

 

[hist78]

It will be interesting to see what CC Wei says on the next call. Last call he joked about forecasting being difficult due to customers changing demand. Maybe he will have something better to say.

One thing I can tell you from the OIP ecosystem conversations last week:

N3 is doing well in all regards, meeting/exceeding expectations. TSMC has 90%+ market share and the ecosystem is fully loaded with silicon proven N3 IP. That is going to be VERY hard to beat, absolutely.

Also, N2, 18A, and a new Samsung 3nm PDKs are out and looking good so we may have a three horse race for the next generation of leading edge wafers. TSMC is still leading in density/power but Intel has performance and Samsung has the lowest price.

I still think most companies will not be in a hurry to use N2, 18A, 3nm and will stay on N3 for another node giving the ecosystem time to catch up and the wafer prices to stabilize, my guess.

"TSMC is still leading in density/power but Intel has performance and Samsung has the lowest price."

This is great for fabless companies to choose what is best for them.