• America’s export controls won’t work…
  • The Perceptron’s latest win…
  • A surprising advancement in an “old school” industry…

Dear Reader,

Welcome to our weekly mailbag edition of The Bleeding Edge. All week, you submitted your questions about the biggest trends in technology. Today, I’ll do my best to answer them.

If we have a question we’d like answered in a future edition, we can submit it right here.

Now, let’s turn to the mailbag.

Can tech companies sidestep export controls?

Hi, Jeff.

Why doesn’t ASML sell their EUV machines to China and simply replace all US components with those made by other countries?

– John L.

Hi, John. Thanks for your question. On Monday, we discussed new export controls around semiconductor manufacturing equipment.

The rule from the Bureau of Industry and Security is titled “Implementation of Additional Export Controls.” And it outlines aggressive new export controls regarding the export of semiconductor technology.

And the target is clear. As I wrote on Monday, this is effectively an economic war with China. The new rule will stimy China’s ability to produce bleeding edge semiconductors.

And as I shared, the rule is misguided. And it may even backfire on the U.S. in the long term. For my full analysis, I’d encourage readers to catch up right here.

But as for your question, John. Could companies like ASML get around this control by simply “swapping out” U.S. components in their machines?

That’s a logical thought.  But the U.S. has literally banned ASML from selling any kind of EUV technology to China, irrespective of whether or not ASML’s EUV manufacturing machines include U.S.-made components or not. 

And ASML is the only company on the planet with EUV technology.  And yes, the U.S. does have that power despite ASML being a Dutch company.

But let’s say just for argument that U.S. components could be swapped out.

As most of my readers know, prior to my role as an analyst, I spent my entire career as a high technology executive. One of my roles was working as an executive at NXP Semiconductors Japan. NXP’s semiconductors were widely used in the automotive market.

The supply chains around building something like a car or a truck are incredibly complex. It takes literally years of design, planning, and securing the necessary components in order to manufacture the end product. With very few exceptions, it would have been unthinkable to “swap out” a few components at the last minute.

It’s a similar story with ASML’s lithography machines. These are the most advanced semiconductor manufacturing machines on the planet. Have a look at what one of them looks like below. A machine this complex simply can’t have all of its components “swapped out.”  Many are custom designed for ASML.

ASML’s EUV Machine

Source: ASML

It would take years to make the adjustments to rid ASML equipment from American components.  And back to my original point, ASML still couldn’t sell EUV technology anyway to China.

China would ideally like to purchase these machines directly from ASML. That would give it the ability to produce semiconductors on par with what U.S. companies like Qualcomm, NVIDIA, and AMD are producing.  But the reality is that China still has access to ASML’s machines through other channels.

The reality is that most fabless semiconductor companies use either TSMC in Taiwan or Samsung to manufacture their bleeding edge semiconductors. Both TSMC and Samsung manufacture using ASML’s EUV semiconductor manufacturing equipment.

And China-based companies hire TSMC and Samsung to produce their semiconductors. China’s fabless semiconductor company Biren is a perfect example of that.

This is one of the reasons why I think the policy is misguided. There will always be secondary markets and intermediaries. And there are even technical ways to procure advanced semiconductor components which can then be used to produce advanced semiconductors in mainland China. 

All these policies will do is damage U.S.-based technology companies and antagonize the Chinese government at a time of intense global crisis. And they will also increase the likelihood that China will assert control over Taiwan, and ultimately the Taiwanese semiconductor industry, which would be devastating for both the U.S. and Europe.

Another successful trade from the Perceptron…

Hi Jeff,

While a 40% gain over a weekend with Enzyme is not too shabby, it did go up more than 100% over the weekend. Since this is a trading service, are there any plans for the Perceptron to monitor and give signals over the weekend? With crypto markets available 24/7, I can see this happening again.

George B.

Hi George, thanks for writing in. To catch readers up, George is referring to a recent trade for our AI-powered trading service, Neural Net Profits. This service is powered by a proprietary neural network that we refer to as “the Perceptron.”

Last Friday afternoon, the Perceptron delivered us a very clear buy signal on a digital asset called Enzyme Finance (MLN). The Perceptron was predicting a big move with MLN. And we didn’t have to wait long.

As the chart shows, MLN traded mostly flat into Friday evening. But that night and into Saturday morning, the asset took off. We exited the trade on Monday for a 40% return in four days.

[As a sidenote, I know there were several readers who wrote in saying they could not take a position in MLN before this move came. And to those readers, please don’t get discouraged. There will be plenty of new trading opportunities with the Perceptron in the near future.]

But as you said, George, this trade is a little bittersweet. Had we exited sooner, our closed return would have been more than 100%.

We were working/monitoring over the weekend, and this was an important topic of discussion amongst me and my team on Monday. The Perceptron’s signal was still strong for Enzyme on Saturday (which is great); but on Sunday we saw the signal start to decline (our alert to monitor closely). By Monday we had a clear sell signal and sent out the alert.

An argument could be made that in a market like this it would be better to give up any possible upside and just take the gains quickly on Sunday morning.  Many subscribers would have probably missed a Sunday morning alert though. Not knowing however if a larger run was in store, there is also the risk of missing out on further upside. Hindsight is of course 20/20.

But with an active trading service like Neural Net Profits, it is important for us to adopt something of a “trader mindset” when we approach this product. 

I traded equity options strategies and futures options for years when I had the time to actively monitor my positions. One of the things that I learned quickly is that it is impossible to perfectly time the market on every trade. Sometimes it happens; but most of the time it doesn’t. So I found it wasn’t a good use of my time to think about the “if only I had” situations.

“If only” I had timed the top perfectly. “If only” I had waited a few more days for that perfect exit. “If only” I had been a little earlier or a little later.

The reality is that it is very rare to perfectly time the tops and bottoms of any trade. What’s most important is that we’re consistently securing winning trades and finding the next high probability trade. And with that mindset, I’d say a 40% return in four days is one heck of a trade.

There will be more.

Thanks again for writing in. And just to be clear, it is entirely possible for us to send out a buy or sell alert over the weekend.  We monitor the Perceptron and our positions daily and we have the team on standby to publish alerts when needed over the weekends.

An arms race for the grocery store of tomorrow…

Those Veeve carts sound like a great way to go. I’ll look forward to seeing any updates. From what I’ve heard Amazon is having trouble with its camera-based system.

A comment regarding the Kroger/Albertsons/Safeway issue. I believe that Kroger will not only cut costs by reducing employees but also by closing some of the Albertsons and Safeway stores in communities where they have previously competed with other Kroger owned stores such as Fred Meyers in the Pacific Northwest (primarily). It seems Kroger is aiming to become the Amazon of grocery outlets!

John H.

Hi, John. Thanks for writing in. The grocery store business isn’t what we’d normally expect in the world of The Bleeding Edge. But as John alluded, there are some interesting technological developments coming to our neighborhood grocery store.

The big news on Monday was that early stage company Veeve raised a nice venture capital round to fund its product development and expand its business. 

But the more interesting angle is that Veeve already had pilot programs for its smart shopping cart product with both Kroger and Albertsons. This puts Veeve in an incredible competitive position given the deal between Kroger and Albertsons. Veeve is in the right place at the right time…

Veeve’s carts use computer vision to scan every product we put into them. And for items priced by weight, the cart also has a built-in scale at the bottom of the cart. So, it knows exactly how heavy each added item is.

It’s a great way to remove the friction from the checkout process we’re accustomed to at the grocery store.

Veeve’s approach is different however than the approach being taken by Amazon, or by another competitor Grabango, which don’t require special carts. Cameras are used on the ceiling to track and identify which items are being taken off the shelf and placed into a bag or cart.

Back in 2020, I visited the Amazon Go store in Manhattan just after it opened. These are small convenience stores where you scan a QR code on your phone’s Amazon Go app as you walk in. That ties you to your Amazon account.

Then you take whatever you want off the shelf and walk out. AI-powered cameras track your purchases and charge your Amazon account as soon as you leave.

The technology worked perfectly. I even tried to trick the artificial intelligence by picking up items and then putting them back. I also tried obscuring the angles of my body with respect to the cameras. The AI spotted it each time and only charged me for the products I left the store with.

Now, if this technology is adopted in a larger grocery store setting, it’s likely that the company will have to fine tune the technology. The technology is still being refined, so there will always be areas for improvement; but there are some clear advantages to not having to employ expensive shopping carts to enable a system like this.

And you’re correct John, stores will have to be closed which will reduce operational costs. This will actually be required to get the deal approved. In markets where the combined company would have near monopoly control, they’ll have to either close or sell off stores to competitors in order to get past an antitrust review.

Either way, there will need to be improvements in operational efficiencies in order to compete with Walmart, and of course Amazon as it expands further into grocery. Technology is going to be a large part of that effort.

That’s all the time we have this week. Remember, if you’d like me to answer one of your questions, please send me a note by writing to [email protected]. I’ll do my best to get to it in a future mailbag.


Jeff Brown
Editor, The Bleeding Edge