Kulicke & Soffa KLIC Introduction

[wpr101]

Fast growing semi $KLIC Kulicke & Soffa could be undervalued

I wanted to kick off a discussion thread for this company Kulicke & Soffa KLIC, which is a US based semiconductor that had an impressive earnings report recently.

They make components and systems that connect silicon chips to packaging. There are a lot of new products being released working with thermo compression and “vertical wire”. Kulicke & Soffa is also entering the memory space where they saw memory shipments rise 93% qoq to 31.3M

For overall revenue in the latest quarter the company had 243M which was +50% yoy and +22% qoq. Even better they gave strong guidance of 310M next quarter which is +109% yoy and +28% qoq. Gross margins are 49% in the latest quarter with 43M EBITDA, and 35M net income.

Surprisingly even with guidance of 310M of revenue and a 100%+ yoy growth rate the market cap is 5.1B. I see this company as being undervalued and practically unknown by the market. I would be interested to hear more thoughts about how folks view this company and their prospects.

[ryshab]

Extremely impressive find! As always, you are extremely generous in sharing your great ideas with this group. Thank you for the meticulous research.

I will be honest - I haven’t seen the video review yet, have done no research but I ran the numbers. And just by that alone, this business looks very interesting.

Here are the highlights:

• Analyst trend beat already at 45% (above 20% is pretty good)
• Analyst next year’s estimates are too low at 18%. That’s a sign of undiscovered stock. Because the current revenue trend is upward and ntm growth rate is above 50%. If you plug in a 40% year two growth rate, which I believe this stocks will get to, the quality score is at 50 for my methodology. That places this already in top 20 stocks
• NTM P/S is 4. That’s ridiculously low for a business that is going to grow at 50% in the next 12 months.
• No dilution
• IV around 49 which screams undiscovered - so that should head up over the next 6-12 months.

Already in my active watchlist. Time to do some research on the business now.

[wpr101]

@ryshab Nice, will be interested to hear your opinion after you check out some more details!

@ryshab

Analyst next year’s estimates are too low at 18%. That’s a sign of undiscovered stock. Because the current revenue trend is upward and ntm growth rate is above 50%. If you plug in a 40% year two growth rate, which I believe this stocks will get to, the quality score is at 50 for my methodology. That places this already in top 20 stocks

Good point there on the analyst estimates. I noticed a day after the company reported and guided for 310M of revenue, none of the future quarters were projected at getting to over 300M of revenue.

Even looking forward now that analysts have updated their revenue models after the earnings, the highest projected revenue number over the next two years is just 339M. Just based of what I have found about K&S, with all the new products and innovations, they could realistically potentially guide for over 340M revenue next earnings report.

[innerpeace123]

Hi WPR101,

First off, thank you for putting in the time to write up this deep dive and for bringing a new name to the board! It’s always great to see fresh ideas and well-researched pitches here.

That being said, I have a different perspective on KLIC, and I want to push back a bit on the idea that this is an “undiscovered gem.”

To illustrate why, I put together a quick blind test comparing KLIC’s most recent quarterly numbers against two other semiconductor/AI-related companies (labeled COMP1 and COMP2). To strip out the distortion of trailing cyclical losses, I added an annualized run-rate P/E ratio by simply taking the most recent quarter’s earnings and multiplying by four (Recent Qtr EPS × 4).

Here is how the numbers stack up:

Metric	KLIC	COMP 1	COMP 2
Most Recent Qtr Revenue	~$242.6	~$23.86	~$68.10
Year-Ago Qtr Revenue	~$162.0	~$8.06	~$39.30
Revenue YoY Growth	+49.8%	+196.0%	+73.2%
Most Recent Qtr EPS (Non-GAAP Diluted)	$0.79	$12.20	$1.62
Year-Ago Qtr EPS (Non-GAAP Diluted)	-$0.52 (Loss)	$1.41	$0.89
EPS YoY Growth	Turnaround	+765%	+82.0%
Annualized EPS (Recent Qtr × 4)	$3.16	$48.80	$6.48
Run-Rate P/E Ratio	~32.3x	~14.8x	~34.7x

If we are looking for the kind of hyper-growth companies that this board typically focuses on, which of these companies would you choose? Most of us would immediately gravitate towards COMP1 for its sheer velocity of growth and incredibly low P/E ratio, or COMP2 for its growth and profitability.

The reveal: COMP1 is Micron (MU) and COMP2 is Nvidia (NVDA).

When you look at KLIC side-by-side with the actual engines of the AI boom, a few things become clear. KLIC’s current stock price and valuation are not a result of Wall Street “missing” the stock or failing to discover it. Instead, the pricing is simply an accurate reflection of the semiconductor memory and equipment cycle.

KLIC operates in the capital equipment space (specifically advanced packaging and bonding). This is a notoriously cyclical, boom-and-bust industry. Wall Street is fully aware of KLIC; however, the market refuses to assign a secular, SaaS-like growth multiple to a hardware equipment maker whose revenues fluctuate wildly from year to year. Even COMP1 (Micron), which is posting nearly 200% revenue growth, is only assigned a ~14.8x run-rate P/E because the market knows that peak cyclical earnings don’t last forever.

Just for full disclosure: I am currently long NVDA, and I do not own any shares of KLIC or MU.

I sincerely hope that anyone who buys KLIC or MU here makes a massive profit. There is definitely money to be made in cyclical upswings! However, if one is investing in this space right now, I firmly believe the thesis needs to be: “The AI data center boom is going to fundamentally break the historical boom-and-bust memory cycle, resulting in an extended, multi-year supercycle.”

If you believe AI changes the hardware cycle permanently, KLIC is an interesting cyclical play. But we shouldn’t confuse a predictable cyclical rebound with an undiscovered secular growth stock.

Thanks again for the write-up, WPR101! Really appreciate the discussion.

Best,

[hilltop4208]

Thanks for bringing this interesting company to our attention. I’ve really enjoyed your videos, and your thorough research is excellent.

One thing I would like to point out is that Haoseng Industrial Co., Ltd. is actually a Chinese company rather than a Korean company. It is also very likely that the products are ordered by US-based companies and shipped to China for packaging. After packaging, the products may then be shipped back to the US or distributed to other parts of the world.

[hilltop4208]

Comparing KLIC at run-rate P/E of 32.3x against Micron at 14.8x and NVDA at 34.7x proves the point only if you accept that run-rate P/E is the right valuation framework for all three companies simultaneously. It is not.

NVDA at 34.7x run-rate P/E is exactly what the peer comparison showed at the equipment level — MTSI trades at 150x run-rate P/E and LSCC at 195x. The market assigns very different multiples to different semiconductor categories based on perceived durability of earnings. The counter-argument correctly notes this when discussing Micron — “the market knows that peak cyclical earnings don’t last forever” — but then applies exactly that cyclical discount logic to reach a different conclusion about KLIC than the one @wpr101 reaches.

The peer comparison use by @wpr101 used (CAMT, MTSI, LSCC) is genuinely more relevant to KLIC than Micron or NVDA. Those are all capital equipment or specialty semiconductor companies. Comparing KLIC to NVDA is like comparing ELVA to Tesla — both are in the energy/transportation space but the businesses are structurally incomparable.

[wpr101]

Good discussion on the valuation. I’m usually not comparing valuation of a large/mega cap company to small/medium cap company because they are on a different part of their scaling journey. What I mean is we would expect a large cap to have a higher income margin as their market is more established and their product more refined.

If we took examples from software it stands out a bit more on the difference. Let’s take Microsoft vs Snowflake. MSFT has a 38% net income margin and SNOW has a -24% net income margin.

In the case given with KLIC vs MU and NVDA, I find in encouraging that KLIC already has a lower run-rate P/E than NVDA.

---

Someone over on the Reddit thread pointed out there is a direct competitor to K&S in the Netherlands called BE Semiconductor Industries with (symbol BESI.AS). Besi is in the same market with a higher focus on hybrid bonding while K&S speciality is wire bonding.

Besi had revenue of 214M in their latest quarter, +28% yoy, +10% qoq. They did give a really strong guide of sequential revenue growth of 30-40% next quarter and analyst see them getting 290M of revenue which would be +67% yoy and +36% qoq.

In this case K&S has more revenue in the current quarter and a higher guided revenue number. Yet the market cap for Besi is 24B, or nearly 5x that of KLIC!

It is also worth keeping in mind that usually US based semiconductors trade at a higher valuation to EU-based ones.

---

@hilltop4208

One thing I would like to point out is that Haoseng Industrial Co., Ltd. is actually a Chinese company rather than a Korean company. It is also very likely that the products are ordered by US-based companies and shipped to China for packaging.

Interesting, was trying to track down this thread and from what I found, AI through Claude was telling me that Haoseng is a private Taiwanese based company. The AI volunteered this company sometimes gets confused with Korean company Haesung Industrial Co., Ltd. (034810.KQ), which is why I mentioned I wasn’t sure if the name was being written slightly differently.

[Buffjan2]

The “All Others” business segment has wonky financials because of a discontinued line of business in 2025. From the latest 10Q:

On March 25, 2025, the Board of Directors of the Company approved a strategic plan related to the cessation of the Company’s Electronics Assembly (“EA”) equipment business. As part of the plan, the Company began the process of winding down the EA equipment business in an effort to prioritize core semiconductor assembly business opportunities and enhance overall through-cycle financial performance. The cessation of the EA equipment business is subject to a consultation process with the applicable works council and union representatives, which the Company initiated in the third fiscal quarter of 2025 and, as of April 4, 2026, has substantially completed. The wind down activities remain ongoing and are expected to be substantially completed by the end of fiscal 2026, after which there will be some service support activities to serve out the remaining customer obligations.

In connection with the cessation of the EA equipment business, the Company has reclassified a product line previously included in the Aftermarket Products and Services (“APS”) segment. This product line is now reported within “All Others,” consistent with how the Chief Operating Decision Maker (“CODM”) evaluates the financial information of the EA equipment aftermarket spares and services product line together with the EA capital equipment business as one operating segment for performance assessment and resource allocation purposes. Refer to Note 3: Goodwill and Intangible Assets and Note 15: Segment Information for additional details.

This is why the gross margins and SG&A expenses look so good in 2026 vs the same period in 2025. It will normalize once the 2025 comps roll off and it is 2026 vs 2027.

For the three and six months ended April 4, 2026, the increase in gross profit margin for the “All Others” category as compared to the prior year period was primarily due to sales of previously impaired inventory as a result of the cessation of the EA equipment business. In addition, the prior year period included inventory write-down charges incurred as a result of the cessation of the EA equipment business.

Selling, General and Administrative (“SG&A”)

For the three months ended April 4, 2026, the lower SG&A expenses as compared to the prior year period was primarily due to $8.3 million lower severance costs. This was partially offset by $2.6 million higher sales representative commissions and $0.6 million higher staff cost.

For the six months ended April 4, 2026, the lower SG&A expenses as compared to the prior year period was primarily due to $7.1 million lower severance costs. This was partially offset by $3.7 million higher sales representative commissions.

Gain relating to cessation of business

For the six months ended March 29, 2025, the gain relating to cessation of business was primarily due to the $71.1 million reimbursement for certain costs and expenses from the cancellation of the Project, a $3.2 million gain on the disposal of a subsidiary and a $1.7 million gain from the supplier settlement.

Impairment Charges

For the three and six months ended March 29, 2025, we incurred $39.8 million in impairment charges on long-lived assets, intangible assets and goodwill related to the cessation of the EA equipment business.

[innerpeace123]

Hi Hilltop,

Thanks for the thoughtful reply! I actually love the angle you brought up here, because it touches on what I think is the most fascinating question in semiconductor investing right now: What fundamentally separates a “cyclical stock” from a “secular growth stock”?

Taking a step back from the specific P/E numbers, I completely agree with you that the market assigns different multiples based on the perceived durability of earnings. Hardware is, by its very nature, bound to a physical cycle: Under-supply → Surging Profits → Massive CapEx/Expansion → Over-supply → Earnings Crash.

When we look at Nvidia, it has historically been a highly cyclical stock. The reason it currently acts and is valued like a pure secular growth stock is a unique combination of factors: the sheer scale of the AI paradigm shift, TSMC’s severe CoWoS packaging bottlenecks (which has prolonged the supply-constrained environment), and a massive software moat (CUDA). Instead of just a chipmaker, the market currently views it as a hardware company with unusually strong software ecosystem lock-in.

However, it is entirely possible that in a few years, once the initial massive wave of AI data center infrastructure is built out and hyperscaler CapEx normalizes, NVDA could very well revert to behaving and trading like a traditional cyclical stock.

When we look at KLIC, we are looking at the extreme upstream end of the hardware cycle: Capital Equipment. This part of the supply chain is subject to the classic bullwhip effect. The machines KLIC sells (wirebonders, advanced packaging tools) are highly durable. Once downstream foundries and OSATs (Outsourced Semiconductor Assembly and Test) fill their factory floors, their orders can decline dramatically for extended periods following major capex buildouts until the next major tech upgrade. This naturally leads to a far more cyclical revenue profile.

This dynamic is exactly why the market values capital equipment companies the way it does. If we look at KLIC’s actual peers in the packaging and inspection equipment space, including companies like BESI (BE Semiconductor), ASMPT, or Camtek (CAMT), we see this play out. BESI commands a massive premium right now because they are widely viewed as a leading early mover in Hybrid Bonding (the critical bottleneck for next-gen AI chips). KLIC, while making great strides in Thermocompression Bonding (TCB), still derives a large portion of its revenue from legacy wirebonding, which is tied to the highly cyclical consumer electronics and automotive markets.

To be clear, buying a deeply beaten-down equipment stock at the trough of a CapEx cycle is a brilliant, highly profitable investment strategy! The cyclical rebound can generate massive returns.

My only hesitation regarding KLIC for this specific board is whether it fits the mandate. I think it’s important to distinguish whether we are investing in a “predictable cyclical rebound” (where we need to time our exit before the CapEx cycle turns again) versus a “secular compounder” that can power through macro cycles. The distinction probably comes down to whether the company can sustain durable demand independent of capex timing.

Really appreciate you pushing back on the valuation framework. It’s a great reminder of how the market discounts different types of earnings, and this exact type of discussion is why I love this board!

Best,

[stocknovice]

Good discussion. I think we are fooling ourselves if we consider any of these companies with a this-time-is-different attitude as far as being cyclical at their core. Stealing a phrase from @GauchoRico, the question is whether there’s a “higher for longer” dynamic in play this cycle due to the sheer size and scale of the buildout.

Anyone in the AI-trade for more than a few months is likely already sitting on a nice profit cushion, and a lot of those profits have been found in established companies seeing a step up in revenue causing rapid and significant growth acceleration (as was shown with NVDA and MU upthread).

KLIC’s last couple quarters show that same pattern. The bull case is there’s room for it to run as its multiple and market cap catch up to its peers depending on which comps you use. The bear case is there’s nothing special here, and KLIC is not differentiated enough to deserve much more of a run than it’s already had. Regardless, I view KLIC as a short or medium term option only no matter how I eventually play it.

[hilltop4208]

Haoseng is a HK based distribution company working with KLIC for some years and its customers are mostly based in mainland China.