AEM Holdings

[Boon]

This looks "different" to the patent drawing, doesn't it ?

[weii]

Looking backwards:
 
Fundamental of AEM has certainly improved over the last few quarters.
 
Share price of AEM has certainly increased significantly over the same period.
 
Historical PE based on FY2016 EPS of SGD 0.1088 and share price of SGD 1.885 = 17
 
For those who have managed to buy in earlier at lower price (say below 0.50 a share), AEM is already a few-bagger to them.
 
Looking forward into FY2017 and beyond:

Based on order book and management guidance, 1H2017 revenue and PBT is going to surpass that for the entire FY2016.
 
If FY2017 EPS = 1.5 x FY2016 EPS, => Forward PE = 11
If FY2017 EPS = 2.0 x FY2016 EPS, => Forward PE = 8
 
This is still considered cheap, if growth potential is going to be significant, going forward.
 
If AEM’s “handlers” really has what it takes to reduce test costs significantly for the market players, questions remain:
 
1) How big a market is it for AEM’s handlers, if Intel is its only customer?  Bear in mind that Intel is still the number 1 chipmaker in the world - lots of chips to be tested.
2) How big a market would it be if Intel’s foundry customers were to be included?
3) How big a market would it be if other customers were to be included, assuming AEM holds the IP rights of the “handlers?
4) Does AEM hold the IP rights of the “handlers”?
____________________________________________________________________________________________

According to CIMB analysis report dated 8 Mar 2017,
AEM’s current major competitor is Cohu. AEM and Cohu share the same major customer; should this customer decide to switch all its future needs to the
current test handler being produced by AEM, there could be a case for Cohu to consider an acquisition of AEM to keep its test handler business with the same customer.

CIMB analyst assumes that the current addressable market for AEM’s Test Handlers with its major customer is approximately 180 units. This is based only
on the customer’s server chip and Silicon on chip business. We note that its major customer has other business segments such as IoT, modems, memory
etc that could represent business opportunities for AEM. CIMB analyst assumes that the Average Selling Price (ASP) is S$3.0m per Test Handler.

My understanding is that AEM holds the IP rights of the "handlers" and it can be used with other customers' testers. However, AEM is too busy serving its major customer with a backlog to look into this at the moment. 

[Boon]

180 units each with ASP of SGD 3 m,
 
=> Revenue = SGD 540 m
 
Assuming a net profit margin of 7% (~ FY2016 NPM),
 
=> NPAT = SGD 37.8 m (~ SGD 0.88 per share), which is not a lot.................... 
_______________________________________________________________________________________________________________________

[holymage]

Personally, I believe AEM still has room to grow for at least a couple of years, unless there is a sudden industry downturn.

1) I do believe that AEM's technology has an edge over their competitors. And I think Boon explained it well. To reinforce the point, if Intel really have the technology for the entire HDMT system (ATE market), why does Intel still need to purchase from AEM?
The next question is, is Intel really AEM's customer? The answer lies in AEM corporate presentation dated 20 Jan 2017. We can see that AEM has field support in Oregon (US), Chandler (US), Austin (US), Costa Rica, Chengdu (China), Penang (Malaysia) and Ho Chi Minh (Vietnam). Next, we can check Intel's Manufacturing Factsheet that Intel has similar operations in Oregon (US), Chandler (US), Austin (US), Costa Rica, Chengdu (China), Penang (Malaysia) and Ho Chi Minh (Vietnam). We can also see that majority of AEM's revenue is from US, Malaysia, China, Vietnam and Singapore.
Thus, we can see strong clues that AEM is supporting Intel's operations. The fact that Intel still purchases from AEM shows that AEM's technology is indeed superior. Personally, I read through AEM's patent and logically, it does seems efficient, due to its modularity and "scaleability". Although I am not too sure how easy will it be for their competitors to go around their patent and achieve similar breakthroughs.

http://download.intel.com/newsroom/kits/...tSheet.pdf


2a) I suspect AEM's future revenue stream will sustain largely based on consumables. I have always wondered what % of their business based on recurring income and what % is based on equipment. My query was answered is evident in AEM corporate presentation dated 28 Feb 2017, and we can observe that 2016 recurring income doubled that of 2014. The very skeptical investor may ask, "Is this true?". One can observe from the youtube video below that testing are performed in a cleanroom. This is because due to contact, there will be wear and tear, during which will generate particles, which will damage the chips. In fact, the same lessons can be observed from UMS business as well, and this question was answered by MrEngineer in the UMS thread.

2b) In addition, Intel could also refresh its existing old test handlers.

2c) Intel is resuming plans for Fab42 (Chandler) in 2020, for 7nm gate size (thanks to Donald Trump?). I would expect Intel to purchase new machines for additional operations.

https://www.youtube.com/watch?v=3xQct03l6bI
http://www.oregonlive.com/silicon-forest...to_me.html


3) Currently, Intel foundry business is minuscule compared to their core businesses.


4) Next, Intel's HDMT promotional video is just an illustration. As Intel doesn't sell its testers in the market, I suspect Intel intends to keep its technology as a trade secret, instead of patent their technology. From the video, the entire testing process is inside a box and designed to be automated. Does it make sense to have little LCD screens when it is supposed to be automated? How many workers do they need to constantly monitor the numerous LCD screens? Logically, we can tell that it is an illustration.


5) Investing based on the thesis on a potential acquisition is alike to wishing that you have a wealthy long distance relative who will pass away soon and you will inherit his wealth. Just my opinion.


These being said, I am not an industry insider nor familiar with semiconductor technology.

Hope it helps.

[amperex]

Thanks to Boon and holymage. What both of you posted certainly help me to understand more about AEM. These things are outside my circle of competence. But think I understand ( just ) enough to know that I should remain invested for couple of years at least.

[white collar]

Slighted confused here. Is the bonus / capitalisation 1 for 2 shares issue entitled to the dividend declared of 1.3c? I don't think it is.. is it?

[amperex]

Slighted confused here. Is the bonus / capitalisation 1 for 2 shares issue entitled to the dividend declared of 1.3c? I don't think it is.. is it?

If you owned 2000 shares then, now you have 3000 shares XB.
You will receive 1.3 cents per share for all 3000 shares.

[laowangnextdoor]

Slighted confused here. Is the bonus / capitalisation 1 for 2 shares issue entitled to the dividend declared of 1.3c? I don't think it is.. is it?

If you owned 2000 shares then, now you have 3000 shares XB.
You will receive 1.3 cents per share for all 3000 shares.

Pre-bonus 1.95 cents, ex bonus 1.3 cents

[zz...]

After last night's Q1 results announcement, price jumped 22% today

http://infopub.sgx.com/FileOpen/Results-...eID=449769

[barneyonline]

Orders from Intel solely has shoot up earnings... more companies are approaching to buy the new equipment from AEM holdings!! >.<
A 5 years research will cover minimum 5 years growth in earnings... the public will see... ^^