Chips -- Trump's Tariffs -- February 12, 2025

Locator: 48546CHIPS.

Link here to Barrons. 

There is so much packed in these two (which I divided into three) paragraphs: 

On the surface, it might seem logical that a large tariff would incentivize domestic production of advanced chips.

But the reality is that there’s no way to divert production domestically over the next few years given a lack of capacity. Advanced chip factories can’t be created out of thin air. It takes roughly four years to build a new fab at the cost of $10 billion to $20 billion.

Money and time aren’t the only factors, though. Technical expertise is also needed. Only Taiwan Semiconductor Manufacturing (TSMC), has proven it can make chips that meet the requirements of top chip designers like Nvidia and Apple (AAPL).

Tell that to Jeff Bezos, AWS.

Tell that to whomever is running Intel (INTC). now. 

Once tariffs on chips are imposed, there are going to be winners and losers. 

[An aside: is there any difference between Biden choosing winners among EVs and non-EV carmakers; and, between Trump affecting winners and losers among chip manufacturers?]

Which brings us to Amazon and Rivian. 

At one time it looked like a no-brainer for Amazon, which had an exclusive arrangement with Rivian, to buy Rivian outright. 

That exclusivity arrangement ended in 2023 (?) and Rivian is now hoping to sells its products to all comers.

Some still suggest that Amazon might / should acquire Rivian. 

Comments and observations:

• buying Rivian won't move the needle for Amazon
• market cap
• AMZN: $2.5 trillion
• Rivian: $12 billion
• 12 with a lot of zeroes / 2.5 with even more zeroes = 0.0048 = 0.5%.

If acquiring Rivian does move the needle, the needle will move in the wrong direction for Amazon if Jeff Bezos buys Rivian.

What would Amazon gain from buying Rivian? Headaches.

The most recent App Economy AMZN graphic suggests that Amazon competes with:

• AAPL
• MSFT
• GOOG
• META.

Tesla, Rivian, Lucid do not show up on that list.

Drilling in on Amazon, Amazon's revenues, most recent quarter (some numbers rounded up):

• online stores (retail): $75 billion (+7% y/y)
• third-party services (retail): $48 billion (+9% y/y)
• subscriptions: $12 billion (10% y/y)
• advertising: $17 billion (+18% y/y)
• AWS: $30 billion (+20% y/y)

AWS is growing 20% year-on-year. Retail? About 7%. Last-mile delivery is not Amazon's Achille's heel. Amazon delivery is doing just fine with its current business model and is way ahead of whomever is in second place.

Buried in AWS: Amazon is making its own custom chips at its Annapurna Labs, an Israeli microelectronics company, a wholly owned subsidiary of Amazon, reportedly acquired by AWS for less than $500 million.

That line about taking four years and billions of dollars to build a fab factory:

• that may be true, but money seems not to be the issue, here;
• TSMC showed how fast they could be a greenfield fab factory in Arizona; and,
  
• DeepSeek raised questions about the need for high-cost-21st-century-state-of-the-art chips in the first place.
  

Amazon logistics:

But still confusing -- Amazon committed to buying 100,000 Rivian vans -- and Amazon was Rivian's biggest shareholder (needs to be fact-checked) -- raising questions about franchise and non-franchise "last-mile" delivery. 

And then there's the owner. A lot of folks / investors just don't see Jeff Bezos interested in owning a trucking company, even if it's Rivian. 

Bezos seems much more interested in AI and space. In fact, it's now becoming obvious that the fourth industrial revolution (which I call the sixth industrial revolution) is becoming more and more a "marriage" of "terrestrial AI" and "space AI." From a recent post update:

AI: catch-all for sixth industrial revolution. 

• Chips.

• Large data centers:
• copper
• natural gas
• mundane storage

• Space communications:
• SpaceX
• Blue Origin (Jeff Bezos)
  
• Loft Orbital
  

• E-commerce: 
• Amazon
• Walmart
  

• US military-industrial complex:
• Department of Defense
• unmanned vehicles, drones

 I don't see "trucking" in that mix. 

And Trump's tariffs on chips? This will be a huge win for a handful of companies.

Bottom line:

• Rivian might be a nice "bolt-on" for Amazon retail;
• if Jeff Bezos was interested in Rivian, he would have already bought it by now;
• has the EV environment changed under Trump?
• would Bezos rather see more emphasis on drones than on trucks going into the 2030s?
• Bezos already has a drone subsidiary, Prime Air;
  
• Rivian won't move the needle for Amazon; it probably won't even move the needle for Amazon retail;
• does the franchise model for last-mile delivery seem to be working just fine for Amazon?
• would Bezos buying Rivian be transformational;
• trucking seems analogue; the Magnificent 7 are digital, not analogue;
• Rivian doesn't seem to fit Amazon's current menagerie of eclectic companies. it already has Zoox
• buy Rivian and roll it into Amazon EV Logistics (Zoox + Rivian)?