MX - The Unfulfilled Promise Of Magnachip: Assessing Its Latest Quarter's Performance

2023-12-04 13:26:02 ET

Summary

• Magnachip Semiconductor Corporation is a deeply undervalued semiconductor company with a market cap of $260 million.
• The company's potential acquisition by a Chinese firm was blocked by CFIUS, causing its share price to decline.
• Magnachip's revenue is declining, but there is hope for a turnaround in 2024 with a recovery in the power segment.

Magnachip Semiconductor Corporation ( MX ) is a deeply undervalued semiconductor company trading at a market cap of ~$260 million. I've been bullish (and wrong) on Magnachip for a long time, and a reader asked me for my opinion on the last quarter . To be frank; I think it was quite bad. The main saving grace is that the company was already going through a bad part of the cycle, re-organizing and cutting jobs; revenue declining further doesn't hurt so much because it is at very low absolute levels already. There isn't a clear line of sight towards things turning around, which is why the market is just really fed up with this company. It's one thing if things are bad, but without near-term glimmers of hope, people just look elsewhere.

Before I dive into the specifics of the third quarter . I want to highlight the big-picture idea here. Magnachip was on the verge of being acquired for $1.4 billion at $29 per share as recently as March 2021 (it was a competitive process with many bidders, including U.S.-based ones). The "winning" buyer was a Chinese private equity firm called Wise Road Capital. This was blocked by CFIUS aka the Committee on Foreign Investment in the United States. That's even though Magnachip doesn't have a lot of business in the U.S. The agency reviews the national security implications of foreign investments in U.S. companies or operations. The current share price is $6.71 with $4.2 in cash and equivalents per share on the books. There is no debt to speak of.

The CFIUS-blocked semiconductor transactions I identified have generally done quite well. The Qualcomm ( QCOM ) overture to Broadcom ( AVGO ) was a semi deal blocked before Magnachip:

Lattice Semiconductor is the last remaining independent maker of FPGA's. I talked extensively with semiconductor expert Philip Freidin, see interview here , who worked at Advanced Micro Devices ( AMD ) and Xilinx, about that deal but also specifically about this very interesting semiconductor type. He helped me understand what was special about it.

Aixtron's returns are more pedestrian, but then again, it's not listed in the U.S:

GSC Holdings trades in Taiwan and declined by roughly 50% since the acquisition attempt in August 2016. There is no price data on YCharts, unfortunately.

China Resources Microelectronics Ltd and Hua Capital Management Co Ltd tried to buy Fairchild Semiconductor in 2016 for $2.46 billion in cash. ON Semiconductor Successfully Completed the Acquisition of Fairchild Semiconductor for $2.4 Billion in Cash later that year. ON Semi went on to crush it. My superficial understanding is that the acquisition did cement the company as the nr.2 in the semiconductor "power" subsegment.

As the deal fell apart, Magnachip traded down, and its share price was only supported by its net cash per share. Unfortunately, its net cash per share has dwindled through the downturn at a faster clip than I anticipated, getting us to the current state.

The company has two segments (which it is separating) in power and display. From the earnings call, it is clear the segments aren't officially separated until early 2024. The idea behind this separation is to make it easier to sell or spin-out components of the business. The share price is at a point where I'm not really thrilled to see the company hiving off a segment at a 30% premium, but at least in the short-term that option isn't likely on the table, for better or worse:

Finally, a few comments on our previously announced plan to separate our Display and Power businesses into separate legal entities. As we announced previously, our internal separation of the Display and Power businesses will be effectuated by establishing a separate operating company under Magnachip Semiconductor, Ltd. MSK, the company’s primary operating subsidiary.

In September, we established a limited liability company registered in South Korea, named Magnachip Mixed-Signal, Ltd., MMS. The separation will include the contribution of assets and liabilities of the display business and the power IC business to MMS, transfer of directly associated resources such as sales, marketing and R&D, as well as allocation of shared expenses of certain corporate functions including HR, Finance, Legal and IT. This internal separation is expected to be completed and go effect on January 1st, 2024.

The worst thing about the earnings call was the $5.4 million buyback. Sure, the company instituted a $50 million buyback program and has generally done a decent job in years past. I thought it was very good they did $25 million in Q2. With the share price falling further, the company only managed another $5 million.

If you believe in the long-term turnaround that's in motion, hit that buyback hard. Once the turnaround is obvious, management isn't getting another chance at these prices. This will be the easiest EPS growth ever created.

Now, moving to our fourth quarter guidance. Amid heightened global geopolitical and macroeconomic uncertainty, we expect Power demand to soften in Q4, driven by normal weaker Q4 seasonality and inventory correction in industrial end markets.

Power is actually the part of the business that was holding up best. It now seems to be falling apart. It is especially weak on the industrial side. This does make some sense because these power management chips are used in the renewable energy industry. Although it is obviously a growth area in the long term, currently, that sector is getting hammered. With the backdrop of the energy transition, it seems unlikely to me we'll see sustained (think multi-year) weakness from that sector.

Another growth area for the company is in automotive chips. But with doubts around electric vehicle ("EV") sales recently, I don't think anyone was getting their hopes up on that front. But there was an interesting exchange on the call about these chips:

Analyst

we've heard mixed data points as far as demand evaporating and but then we just heard from Allego [ph] this morning that the demand's actually really strong. Are you seeing anything specific as far as near-term demand in China?

Suji Desilva

On the automotive EV, I believe China is strong. I think the -- what we are hearing, the EVs in the non-China market is slowing down. So, I think that's because of the EV price in China is much competitive. Rest of the word, the EV price are very high. I think that's what it is. And I just said, the IGBT is a key solution for more, affordable EVs.

IGBTs are insulated gate bipolar transistors. IGBTs are used as inverters for electric vehicles, train traction, variable-frequency drives for motors, and power supplies. This type of chip helps to get a lot more mileage out of batteries at a relatively low cost. The company seems to be especially competitive at the lower end of the EV spectrum (which are notoriously expensive)

There are only a couple of analysts who are following the company. Their work is broadly painting a picture of nothing happening until the end of 2023. In 2024, we're going to see a semblance of a recovery. In my opinion, this is likely to start in H1 2024 but will be very weak potentially it can become overshadowed by revenue loss in power (if that slowdown continues). Analysts see $0.60 in negative EPS in 2024. That seems overly pessimistic to me as we've just gone through a lot of cost-cutting and revenue should be going up.

Analysts are seeing an even stronger recovery in 2025. I think this is a reasonable picture, but management can really accomplish a tremendous amount in terms of eventual value creation if it hits the buyback hard, gets design wins going, and shows beautiful margin profiles (the nearly finalized headcount reduction should have helped with that) once the turnaround is complete. Management actions continue to have an enormous impact.

You could question why I remain invested here if I roughly agree with analyst estimates and am basically at the mercy of management's favorable or less favorable actions. The answer is that I don't think the market is pricing in these analyst expectations. Some of their expectations like deeply negative 2024 EPS seem to be too bearish to me.

Market participants looking at this think they can get back in once the recovery is visible in the numbers. If you consider the cash per share and clean balance sheet , it looks even less likely the analyst's estimates are priced in.

In conclusion, Magnachip almost sold itself at the peak of the cycle. Instead, it went through a very challenging period. It is undervalued with a market cap of approximately $260 million vs. $160 million in cash and virtually no debt. It went through a headcount reduction program. Revenue seems likely to be picking up in '24. The lack of immediate prospects for a turnaround are keeping investors wary. Management didn't help convince anyone with a weak buyback in Q3. Ultimately, value creation depends a lot on execution from here on out. If management is confident in its execution, the low Magnachip Semiconductor Corporation share price is a tremendous opportunity for long-term value creation.

For further details see: