AMZN 股价被低估；受消费习惯和 AWS 业务多元化支撑，代理式购物带来的风险目前微不足道，因此维持多头仓位。

So now let's move on to Amazon because Amazon has surprisingly been getting hit in the market on the Muse announcements and people are getting more bearish on Amazon after Muse.

Now the first risk that I am able to identify is that if your everyday consumer is using an agent to shop then there will be less real eyeballs on e-commerce platforms like Amazon. The second key point is that this really impacts the high margin advertising business and turns e-commerce companies into commoditized distribution platforms to facilitate sales. Number three, in this world, e-commerce becomes a place where agents buy from and consumers never actually visit. And finally, this would also mean that Amazon loses a ton of data on its customers, how they browse, their habits, etc., which also impacts their advertising business, and how well they can actually advertise to real people.

The first point is that studies have shown that consumers do not fully trust agents to complete the entire purchase process. Instead, people are using agents to help find the best deals on a specific product that they know they already want to buy. Amazon is also building its own agents within its e-commerce platforms like Rufus to help consumers with their shopping needs. Rufus is also seeing strong demand and increasing the conversions on Amazon's website. Amazon is also beginning to guard its e-commerce platforms against agents to stop agentic shopping. In fact, it's already sued Perplexity for doing this.

As a very heavy Amazon shopper, I shop for basically everything on Amazon and it's the first place I go for every product now. I' I just have so much faith in this business. So let's talk about how I still shop now. So the first point here is that I still browse Amazon and look at options and reviews before buying pretty much anything. I want to know that what I am purchasing is what I believe is the best and I'm not sure if an agent would be able to do that for me or if I would even trust it to. However, once I have an item I want, then I would use an agent for future orders if I am a repeat buyer. For example, I have three cats and I run through a lot of cat litter. I have found what I believe is the best cat litter and now I order the same one routinely and I actually have an Amazon subscription that just orders the same cat litter and it just shows up and I don't even need to do anything. And I think that for me personally, that is where I could see myself using an agent is for those repeat orders after I have already identified the product that I want. But as I just said, Amazon also has subscriptions for these recurring items. So this wouldn't change my buying patterns all that much. and it's actually already largely automated for me. So, I do think that humans will still want to browse the marketplace and see options rather than just fully trust an agent to go through and pick something for them. This is especially true if advertising works its way to the agents and you can pay to be the product that an agents will choose because sometimes in my own experience I have noticed that the advertised item is not always the item that I want or the best item on the market. So again if an advertiser can pay to be the product that an agent chooses then that would actually cause me to trust my agent less.

So here is my conclusion as an Amazon shareholder. Amazon's advertising business is what is at most risk from agentic shopping apps like Meta's Muse. However, consumer habits are very tough to break and so far people still want to browse products and pick the best option. Aentic shopping will be great for repeat and tedious purchases, but I am not convinced it will completely remove people from Amazon's platforms. At least for me, I know that it will not. And then finally, agentic advertising will reduce my trust in shopping with an agent as I don't always want the highest bidding item to be what shows up at my doorstep and ultimately what I end up paying for. So overall, I think the real risk to Amazon's business today is probably going to be minute. Obviously, we're going to have to pay attention to where this goes in the future, but Amazon's advertising business is still accelerating. It's growing very quickly and Amazon also has its own shopping agents to assist consumers and it seems like Rufus is doing a good job and seeing a lot of demand.

So, I'm not personally worried about this yet for Amazon's business, especially considering Amazon is a very diverse business and it has the massive AWS and so many other businesses within it.

But I do believe that this could be a real future risk that Amazon is going to have to navigate and I am going to have to watch as a shareholder. And the main thing that I am going to be watching is if their advertising revenue does decelerate in any way. And so far that is not happening.

And in the meantime, I continue to believe that Amazon stock is very cheap and it remains the second largest position in my portfolio.

Now if we take a look at Amazon, it had 24.3% market share in 2021 and has now declined to 15.9% in 2026.

This next screenshot shows us that Morgan Stanley Research found that Marcato Libre is accelerating its incremental market share capture in Mexico now gaining over 2/3 more market share compared to Amazon and Walmart in the country.