Sold AEM:SGX

 When I bought AEM in March I was betting there would be:

1. An economic recovery,
2. that would see the semiconductor industry in a up-cycle,
3. and would further benefit Intel Foundries, which has tailwinds behind it,
4. and would disproportionately flood down to AEM, as an Intel test equipment supplier.

Only 1 and 2 occured.

Following Intel's disastrous results 2 weeks ago, AEM released bad results this week.  Bearing in mind that 2023's results were already bad...

I can't tell if AEM's (or Intel's) problems are structural or cyclical.  Advantest and Terradyne's semicon sales and profits increased yoy in 1H2024, but they provide a wider range of Automated Test Equipment, not just SLT, which AEM focuses on.  They also have exposure to AI and lagging end chips (eg: automotive) that AEM/Intel doesn't have.  Advantest management stated (p5): "There was robust CapEx spending by customers for high-performance semiconductors in both SoC and memory, mainly related to generative AI. On the other hand, demand for mature process applications remained soft since the 3rd quarter of the previous fiscal year, resulting in a QoQ decline in sales."      ("Mature applications" --> PCs and servers --> Intel).

My mistakes were catching a falling knife...in a complex and unpredictable industry I can't follow.

There's a chance that this is the bottom.  Semiconductor stocks are also in a down cycle now, so its a bad time to sell.  AEM's lousy results may be the new CEO kitchen-sinking it.  It could just be a matter of timing, as Intel Foundry's capex catches up.  

But I don't know.  So I sold today, taking a nearly 50% loss.  Or 2.4% of my portfolio.  Take the loss and move on.

Getting the macro right does not compensate for getting the stock-picking wrong.

Bought AEM:SGX

AEM is a semicon supplier making test platforms for Intel, which should benefit from Intel's new Foundry business.  They made a loss in 2023 due to the biggest-ever covid-"stay-at-home"-semicon-bubble deflating, and the stock is down 35%.  I got this idea from the Value Investing Substack (part1) (part2) (paid links).

Its both a growth stock and a cyclical:

I think the 2H23 results were a cyclical drop, they should recover with the semiconductor cycle. And they have a structural tailwind behind them with the US trying to construct semiconductor fabs outside Taiwan and Intel re-establishing its manufacturing capability.

The key question for this company is: how much of its revenue is cyclical/non-cyclical, and recurring/non-recurring?  I am not sure yet, but it would affect wether its a trade or a buy-and-hold:

Source: 2023 Results Presentation (slide 11)

Its trading at around 10X peak earnings, not as cheap as I'd like but OK for a fast grower.  I bought a 5% position.

Risks:

• I don't have good knowledge of the semicon industry, hard for me to keep track of this niche (Suppliers of systems level testing equipment to Intel).

• In January an inventory shortfall was discovered, company said it was due to a manual mistake.  I believe theres no high-level fraud.

• The chart looks like shit and I may lose some fingers catching a falling knife.

• Or dead-SGX-stocks remain moribund, while Q's and crypto rocket in a new liquidity bubble.

I'm in a negative cash position now, so won't be buying anything else.  Will take a few months to pay off from my dividends/salary:

This blog will probably be quiet for a while.