Description

Opportunity to short a limited growth over-earner benefiting from a cyclical super-cycle that the company and sell-side are wrongly painting as on the cusp of a V-shaped recovery. During COVID, demand for wire bonders went parabolic as supply chains went sideways and panic buying resulted in KLICs backlog growing +700%.  KLIC has since stopped disclosing its backlog as it is declining (KLIC has historically disclosed backlog quarterly since 1995).  KLIC’s earnings will continue to normalize, the touted V-shaped recovery will not materialize, and the stock will drop.

KLIC manufactures and sells capital equipment and tools used to assemble semiconductor devices including IC chips, diodes, LEDs, and power modules. KLIC primarily focuses on back-end chip manufacturing equipment specializing in packaging equipment where it has an estimated ~60% market share in wire bonding (~75% of total semi chip packaging uses wire bonding, which has been around since the 1950s).  Wire bonders are used primarily for trailing-edge or low-value semis, and wire bonders have an average useful life of ~10 years.  Advanced chips don’t use wire bonding and instead use flip chips or thermo-compression bonding. TCB and other exotics are at dramatically higher ASPs so even though they are 25% of the volume their TAM is larger and growing. 

Wire bonding market is only a $1B TAM and will grow at best 2%-3% annually.  KLIC has also some advanced packaging efforts (the equivalent of wire bonding for more advanced nodes using flip chip) but has a much lower market share and is viewed as a laggard.  KLIC has three main product lines:

1. Wire Bonding: connecting the silicon to the chip via wires to I/Os.  Wire bonders cost as little as $70k and as much as $200k per machine. 28nm and above chips all use wire bonders.
2. Ball Bonding or Die Bonders: connecting the silicon to the chip via small balls. Die bonders cost $600k to $1.5M.
3. Thermocompression Bonding (TCB): Advanced packaging version of wire bonders.  TCB machines cost $1.8M to $2M per machine.  Chips 10nm and below have to use TCB.

KLIC's customers include chip makers, IDMs, and OSAT companies. Sales to KLIC’s 10 largest customers were 49.1% of sales in 2022. KLIC’s main competitors in wire bonding are ASMPT and Shinkawa.  On the advanced packaging side they compete with BESI and ASMPT.     From FY00 to FY19 KLIC had a revenue CAGR of -2.6% and from FY19-FY22 KLIC’s revenue CAGR’d at 40.7%.  2021 + 2022 KLIC revenue of ~$3B represent five years of average revenue for KLIC.  From FY00 until FY19 KLIC had an average EBIT margin of 7.1%.  In FY22 that EBIT margin shot up to 31% and sell-side expects it to drop to 11.3% this year before going up to 16.5% in FY24.   

You can track KLIC based on semi fab commentary as well as Chinese wire bonder imports.   As can be seen below imports of wire bonders has fallen off a cliff.  Given wire bonders are SemiCap they will be a trailing edge not leading edge spend on any recovery that materializes (aka not likely to be a V-shaped one).  

What Can We Make:

Assuming a 13x P/E (10-year average fwd multiple is 13.0x) with $2.18 in FY25 EPS + $1.72 in cumulative dividends implies -47% downside or a -25% IRR. KLIC has always run with a large net cash balance (and always will).  But if we want to view it on an ex-cash basis: $30 target price less $13 in cash => ~8x normalized ex-cash P/E.

How much can we lose?: KLIC proves resilient and is able to hit $1.3B in FY24 revenue (street is at $935M) and is able to achieve 20% EBIT margins (this is 1,100bps above KLIC’s long-term EBIT margin average)  resulting in $3.20 in EPS (well above street at $2.24). Assuming they trade at 20x P/E (800bps above the long-term fwd P/E) and $1.82 in cumulative dividends implying +17% upside or a +7% IRR.  

• Note: at KLIC’s 2021 investor day KLIC guided to $1.5B in revenue in FY25 with $6 in EPS. 

Background

Wire bonding is a slow to no growth category.  As advanced chips take share, wire bonding will drop from ~75% of chips made to around 50%-55%.  ~25% of semiconductors today are flip chips and this is up from 10%-15% a few years ago. If fab utilization drops below 85% demand for wire bonders turns negative, currently global fab utilization is <70% (TSMC is right around 70%).  Lead times have already dropped to 1-2 months from up to 10 months in summer 2022.  Wire bonders are more heavily leveraged to PCs and consumers/IoT but used in mainstream smartphones and on multi-die packages.  Auto (Power MOSFET) is also higher leverage to bonders.  Semi unit growth in the last 10 years was 6.5% with bonders a little slower cannibalized by flip chip.

OSATs and semis have started guiding down with the broader pain in semis doesn't appear to be turning; and everyone is lowering CapEx (revenue for KLIC) guidance: MU (DRAM uses wire bonders), TSM, Samsung, ASE (10% customer), ASM, etc. KLIC customers can cancel orders without a penalty. According to Gartner total semiconductor revenue grew 26% in 201 and 1% in 2022 and capital equipment spending grew 44% in 2021 and 7% in 2022 per Semi.org. KLIC customers will often re-sell aged equipment when utilization rates drop and can sell them at a 50% discount to a new machine (15% if in very good shape). These machines are easily transportable hence the robust secondary market.

Potential Threats

The LT bull case on KLIC is advanced packaging and Mini/Micro LED.  The mini/micro LED product is a mass transfer die platform being developed for the display market. Right now the market for micro LED displays is $250M and it will grow at a 10%-15% CAGR. KLIC's future in mini LEDs doesn't appear to be panning out and AAPL just announced that future iPad Pros will likely use OLEDs and not mini LEDs. Further TSMC is building a planar micro LED component and is fabricating a grid of LEDs in one substrate which will obviate the need to purchase micro LEDs from KLIC. BESI NA dominates advanced packaging and KLIC is struggling to be relevant here. Based on primary research discussions it is painfully obvious KLIC has very little hope of winning share within the advanced packaging market which is dominated by BESI and ASMPT (those two have something like 90% share ofthe flip chip bondng market). 

Path to Profitability:

• Core wire bonder business reverts to the $750M (above the ~$700M revenue level it hovered at for years) after dipping below in 2023/2024. In FY21 wire bonder revenue was $1.4B
• Advanced Packaging and Mini/Micro LEDs grow at 10% annually in 2025 and beyond after growing much faster in 2022 + 2023.
• Gross margins go back to historical levels of 47% after peaking in FY22 at 50% and EBIT margins go to ~15% down from 31% in FY22 but well above the LT pre-COVID average of ~11%
• KLIC utilizes 40% of FCF for share buybacks and increases the dividend 3% annually each year.

These assumptions result in $892M in FY25 revenue, $161M in FY25 EBITDA, and $2.18 in FY25 EPS.  Assuming 13x P/E it results in a $30 stock (on an ex-cash basis this works out to roughly ~8x normalized EPS).  

Catalyst

• Continued weakness in Chinese wire bonder import data disproving the “V-shaped” narrative
• Continued shrinking of the Mini/Micro LED business (has declined -70% or greater YoY three quarters in a row).