TSMC Q1 2016 Earnings Call Transcript is up

[Daniel Nenni]

This is the official TSMC version, not the Seeking Alpha one which has led me astray in the past:

Co. reported 1Q16 EPS of TWD2.50. Expects 2016 revenue growth to be 5-10% and 2Q16 revenue (based on current business outlook and exchange rate assumptionsof $1 to TWD32.30) to be TWD215-218b

http://www.tsmc.com/uploadfile/ir/quarterly/2016/1yk4i/E/TSMC 1Q16 transcript.pdf

View attachment 17077
Lots of interesting things to ponder and misrepresent.... One thing that is perfectly clear, at least to me, is that TSMC has the iPhone 7 business:

Due to the world macroeconomic uncertainties, we reduced our estimate of 2016 smartphone growth from 8% to 7%, PC from minus 3% to minus 6%, tablet from minus 7% to minus 9%, while maintaining digital consumer electronics growth rate at minus 5%. In spite of the reductions of these growth rates, there are still growth areas in smartphones, broadband network, wireless infrastructures and gaming.

So contrary to the media perception smartphones are still growing.

For the whole year of 2016, we estimate the growth of world semiconductor to be about 1%. We maintain our estimate for the foundry market growth of about 5% and TSMC revenue growth of 5% to 10% this year.

I think TSMC will hit double digit growth. The semiconductor industry as a whole I'm not so optimistic. I think Intel is going to hit a serious bump and memory will be price challenged.

The volume production of N7 will start from first half 2018.

Usually when semiconductor companies say first half it means June 30th. From what I hear it will be Q1 2018. TSMC seems to be much more conservative now that Morris is not on the calls.

Currently development of a reliable EUV mask pellicle is another focus to enable a mass production operation.

This was also mentioned at SPIE, serious pellicle problems are afoot.

We expect 16-nanometer will contribute above 20% wafer revenue this year.

iPone 7...

Our expectation of InFO contributing more than $100m per quarter in 4Q this year remains unchanged. In addition to high-volume preparation and product qualification, we are working on yield improvement and cost reduction.

iPhone 7...

This is all from the prepared statement. The Q&A was a bit more fuzzy....... Questions? Comments? Insults?

 

[benb]

My thoughts: Strong ramping demand on 16nm, strong demand on 28nm; basically strength on strength. New lines of business opening up in InFO packages, server chips, and IoT. This is the language of a durably successful enterprise.

CC said the specialties (in smartphones, fingerprint and MEMS) are more than a quarter of total wafer revenue and contribute increasingly to revenue. Also, high voltage and automotive embedded flash. These are areas TSMC has to themselves, where Intel and Samsung don't compete.

I can spot signs of a future where x86 doesn't exist or is irrelevant. Since Haswell, Broadwell and Skylake performance increase has been single digits, it may even be time for Intel to consider an alternative.

 

[Daniel Nenni]

One thing I can tell you is that after my last Taiwan trip I have never been more optimistic. The FinFET competitive threats have really made TSMC a stronger company. Even at 28nm, the FD-SOI threat is taken seriously. There were TSMC people at the FD-SOI Symposium last week but no one from UMC or SMIC. Big mistake. TSMC will protect their 28nm business with their lives, absolutely.

 

[count]

The TSMC calls are always a nice read. No BS.

 

[hist78]

The TSMC calls are always a nice read. No BS.

On the other hand, Intel's conference calls are often loaded with elegant but meaningless English words. In a short and non-glamour way to describe it: It's full of BS.

 

[Speedfriend]

I can spot signs of a future where x86 doesn't exist or is irrelevant. Since Haswell, Broadwell and Skylake performance increase has been single digits, it may even be time for Intel to consider an alternative.

But what has the improvement in performance per watt been?

 

[Staf_Verhaegen]

I can spot signs of a future where x86 doesn't exist or is irrelevant. Since Haswell, Broadwell and Skylake performance increase has been single digits, it may even be time for Intel to consider an alternative.

Personally I don't see the server/cloud being majority of x86; I do see ARM and openpower giving leverage to the cloud service providers to put pressure on Intel's margin which may be problematic for Intel as it's not something they are used.

 

[Arthur Hanson]

Staf and hist, Thanks for the excellent observations and insights, they will help with my investment strategies.