Synopsys stock down by more than 1/3

[swka]

Earnings call clearly did not get received well by the street. But 1/3 in one hour is historical. I recall my cadence friend coming to office 30 years ago and thought their stock just split, turns out Fed just raided their office due to some law suite as I recall, so instead of split, it was sell side pressure.

While earnings number is disappointing, the 1/3 drop is probably has a lot to do with investors looking for signs to get out of AI bubble.

 

[siliconbruh999]

Earnings call clearly did not get received well by the street. But 1/3 in one hour is historical. I recall my cadence friend coming to office 30 years ago and thought their stock just split, turns out Fed just raided their office due to some law suite as I recall, so instead of split, it was sell side pressure.

While earnings number is disappointing, the 1/3 drop is probably has a lot to do with investors looking for signs to get out of AI bubble.

So should i Buy the DIP

 

[blueone]

While earnings number is disappointing, the 1/3 drop is probably has a lot to do with investors looking for signs to get out of AI bubble.

I disagree. I think this drop has everything to do with a PE of 70 and disappointing results and a weak forecast. At a PE of 70 perfection is expected, and a weak forecast is hemlock. What really annoys me is that their bad results is also dragging Cadence's price down, and for no good reason. The AI bubble can't succeed without these two companies.

 

[swka]

with such magnitude, cadence will be down, too many people w/o enough understanding will consider the two the same of the whole segment as a whole.

earnings result sure was the trigger. But does Synopsys business was 1/3 weaker than yesterday? unfortunately that's how market works.

at beginning of the 2025, Synopsys was over $600 per share. drops to mid 400 a few months later. Back to 600 a few months ago. So give it another quarter, it will be back to 600, perhaps

 

[Daniel Nenni]

Complete Wallstreet nonsense or has the AI bubble started to weaken? I will write about this on Friday.

 

[MKWVentures]

Complete Wallstreet nonsense or has the AI bubble started to weaken? I will write about this on Friday.

Other AI stocks are way up today (right or wrong).
Current spin THIS WEEK is that hypergrowth continues .... Bubble is growing??

 

[tooLongInEDA]

Complete Wallstreet nonsense or has the AI bubble started to weaken? I will write about this on Friday.

Don't think it's the former (for a welcome change). Here's what the Synopsys CEO actually said:

“Our results were primarily impacted by underperformance in the IP business as we had the expectation of deals that did not materialize, driven largely by the following three factors: one, new export restrictions disrupted design starts in China, compounding China weakness; two, challenges at a major foundry customer are also having a sizable impact on the year,” Ghazi said on a call with analysts. “And finally, we made certain road map and resource decisions that did not yield their intended results.”

So point #1 affects EDA as a whole. #2 (is this Intel ?) might only affect SNPS. #3 (poor investments/execution) is SNPS specific. Hard to see why Cadence wouldn't also be affected by #1 (equally in China, also in the IP business).

I also wouldn't be surprised if we haven't been through a one-off design (and associated EDA licence) surge with everyone piling into AI chips and a shake out coming when the long term winners and losers emerge. I distinctly recall every man and his dog trying to get into wireless processors a few decades ago.

Also full year earnings forecast cut from $15.20 to $12.80. So a 17% forecast earnings drop and a 30% share price drop. Not completely crazy.

Once again, I do find myself asking how companies get their forecasting quite this wrong (as Intel repeatedly did in the last few years). But then I need to remember the DJT volatility multiplier. Who can plan ahead with that sort of instability ?

 

[swka]

Don't think it's the former (for a welcome change). Here's what the Synopsys CEO actually said:

“Our results were primarily impacted by underperformance in the IP business as we had the expectation of deals that did not materialize, driven largely by the following three factors: one, new export restrictions disrupted design starts in China, compounding China weakness; two, challenges at a major foundry customer are also having a sizable impact on the year,” Ghazi said on a call with analysts. “And finally, we made certain road map and resource decisions that did not yield their intended results.”

So point #1 affects EDA as a whole. #2 (is this Intel ?) might only affect SNPS. #3 (poor investments/execution) is SNPS specific. Hard to see why Cadence wouldn't also be affected by #1 (equally in China, also in the IP business).

I also wouldn't be surprised if we haven't been through a one-off design (and associated EDA licence) surge with everyone piling into AI chips and a shake out coming when the long term winners and losers emerge. I distinctly recall every man and his dog trying to get into wireless processors a few decades ago.

Also full year earnings forecast cut from $15.20 to $12.80. So a 17% forecast earnings drop and a 30% share price drop. Not completely crazy.

Irrational Analysis (https://irrationalanalysis.substack.com/p/masas-rocket-powered-merry-go-round) agrees with you on Intel being the issue for IP. EDA side, since all players switched to time based licensing, it is a lot less "lumpy". Synopsys CFO did say booking is strong. Problem is indeed how much IP growth slow down and for how long.

 

[Daniel Nenni]

Intel and China I guess? Intel is a very large Synopsys customer. There must be a big IP buy if 14A is to go online in the coming years. Maybe that was pushed out? IP really is the canary in the foundry coal mine. No canary no foundry business. Or LBT may be moving Intel over to Cadence? Or maybe it was the blood moon?

 

[swka]

Intel and China I guess? Intel is a very large Synopsys customer. There must be a big IP buy if 14A is to go online in the coming years. Maybe that was pushed out? IP really is the canary in the foundry coal mine. No canary no foundry business. Or LBT may be moving Intel over to Cadence? Or maybe it was the blood moon?

Cadence is likely to gain position on EDA side from Intel, but their IP offering is less competitive.

The other foundry big for Synopsys IP is Samsung, and we know they have been struggling for a while.

 

[Daniel Nenni]

Cadence is likely to gain position on EDA side from Intel, but their IP offering is less competitive.

The other foundry big for Synopsys IP is Samsung, and we know they have been struggling for a while.

Agreed, although Samsung 2nm just got a big boost from Tesla but a big IP buy is probably a year away. Rapidus is also buying IP for their 2nm. In regards to foundation IP: Intel 18A and 14A are not meeting expectations, Samsung 2nm was delayed.

 

[blueone]

Once again, I do find myself asking how companies get their forecasting quite this wrong (as Intel repeatedly did in the last few years).

Investors really dislike senior managers who do not appear to be aware of key business factors, and don't communicate their concerns until the release of quarterly results. And especially not when their stock is priced with substantial growth baked in, like Synopsys.

 

[KevinK]

Intel and China I guess? Intel is a very large Synopsys customer. There must be a big IP buy if 14A is to go online in the coming years. Maybe that was pushed out? IP really is the canary in the foundry coal mine. No canary no foundry business. Or LBT may be moving Intel over to Cadence? Or maybe it was the blood moon?

Read or listen to the first couple of Q&A questions. It's mostly about IP.
* China - the 6 week licensing revocation window due to Trump negotiations killed off business, but also made Chinese customers question longer term "supply" - likely to affect both Synopsys and Cadence. We'll see if China tries to do domestic substitution and where the weak links will be.
* Foundry IP - pretty clear if was Intel - lots of market IP developed for 18A that will never get used by external customers
* Product roadmap - seems somewhat linked to the previous issue, being too tied to Intel strategies - from Sassine "for example, in Edge AI opportunities for IP that we put resources on delivering to these opportunities, and it came at some road map cost on which foundry to make that investment and for data center delay in some of our IP titles."

Sassine's insights into the changing nature of the IP business is also interesting:

"We have to serve the multiple foundries for those multiple markets in both interface IP and foundation IP. And there's more and more customization in particular, for interface IP. And these customizations are moving from an off-the-shelf to a more subsystem delivery which is it takes longer, it takes more resources and our ability to change the business model or the need to change the business model is an ongoing dialogue with our customers because as they're expecting us to do more work than just off-the-shelf IP, there's an opportunity for higher monetization.

 

[swka]

Read or listen to the first couple of Q&A questions. It's mostly about IP.
* China - the 6 week licensing revocation window due to Trump negotiations killed off business, but also made Chinese customers question longer term "supply" - likely to affect both Synopsys and Cadence. We'll see if China tries to do domestic substitution and where the weak links will be.
* Foundry IP - pretty clear if was Intel - lots of market IP developed for 18A that will never get used by external customers
* Product roadmap - seems somewhat linked to the previous issue, being too tied to Intel strategies - from Sassine "for example, in Edge AI opportunities for IP that we put resources on delivering to these opportunities, and it came at some road map cost on which foundry to make that investment and for data center delay in some of our IP titles."

Sassine's insights into the changing nature of the IP business is also interesting:

"We have to serve the multiple foundries for those multiple markets in both interface IP and foundation IP. And there's more and more customization in particular, for interface IP. And these customizations are moving from an off-the-shelf to a more subsystem delivery which is it takes longer, it takes more resources and our ability to change the business model or the need to change the business model is an ongoing dialogue with our customers because as they're expecting us to do more work than just off-the-shelf IP, there's an opportunity for higher monetization.

great analysis.

Cadence IP footprint is much smaller than Synopsys. Used to be a disadvantage as IP growth is always much higher than EDA (not so much margin though). But this time, China impact on Cadence should be mush smaller than on Synopsys.

On the other hand, China has been stockpiling IP purchase in anticipation of geo-political conflict. So perhaps this is more akin to inventory adjustment. IP usual business model does not normally lead to this behavioral, but the "localized" business norm is very different in China

Thanks for the summary on Sassine's comment. Wrong bet or priority on edge IP versus inside DC. Very much similar to overall AI market, where focus is all on DC and infrastructure, Everyone is saying real money will be from inference, no one knows when that will materially happen.

 

[Daniel Nenni]

Irrational Analysis (https://irrationalanalysis.substack.com/p/masas-rocket-powered-merry-go-round) agrees with you on Intel being the issue for IP. EDA side, since all players switched to time based licensing, it is a lot less "lumpy". Synopsys CFO did say booking is strong. Problem is indeed how much IP growth slow down and for how long.

Not too far off.... Foundation IP for 14A is probably not paid for yet so that may be an upside. Same with Samsung 2nm now that Tesla has made a major commitment. Definitely do not count Synopsys out since they own the IP market.

 

[KevinK]

Everyone is saying real money will be from inference, no one knows when that will materially happen.

Just to be clear, I think you are talking about edge inference, right ? Plenty of inference in the data center - that's the big build-out reach right now, with bigger, deeper models and more users.

 

[siliconbruh999]

Intel 18A and 14A are not meeting expectations,

It's kind of weird when IP is not meeting expectations but the process is doing fine for Intel Internal use just shows how difficult is too do IP for external.

 

[Daniel Nenni]

It's kind of weird when IP is not meeting expectations but the process is doing fine for Intel Internal use just shows how difficult is too do IP for external.

18A is working for Intel products which do not use Synopsys IP. There are two parts to commercial IP, the NRE and the royalties. IP companies generally break even on the NRE but the royalties can be really profitable. No customers, no royalties, it can kill a company if they are all in.

 

[Barnsley]

Earnings call clearly did not get received well by the street. But 1/3 in one hour is historical. I recall my cadence friend coming to office 30 years ago and thought their stock just split, turns out Fed just raided their office due to some law suite as I recall, so instead of split, it was sell side pressure.

While earnings number is disappointing, the 1/3 drop is probably has a lot to do with investors looking for signs to get out of AI bubble.

Just up the license fee by 50% is their usual policy.

 

[Fred Chen]

This didn't age well: https://investor.synopsys.com/news/...18A-and-Intel-18A-P-Technologies/default.aspx