Meta业绩超预期，营收646亿美元、同比增长26%，并指引第四季度营收730亿至760亿美元，但股价本周下跌逾8%至642美元：1350亿至1450亿美元的资本开支意味着第四季度支出巨大，规模未知的C类股将稀释股东并增加供给，通常意味着股价走低；Muse看起来被过度炒作。

It's best to start things as we always do with Meta. After rising to $700 at the start of the week, Meta's stock is no better off. It would be best to take a technical look at this stock in a little while, but the stock has fallen by more than 8% this week . The stock ended the week at $642.

This comes as the company announced its third-quarter earnings. I mean, it was very good. Revenues reached $64.6 billion. That was a 26% year-on- year growth. They exceeded expectations.

They also provided guidance for the next quarter, the very important fourth quarter in which the company sells huge amounts of advertising. Revenues are expected to range between $73 billion and $76 billion.

Wall Street's average estimate was just under $74 billion. So, all of this sounds really good, doesn't it?

I thought this was one of the most interesting lines in the entire Meta report. The fact that the company is directing its capital expenditures, including loan principal payments and finance leases, to range between $135 billion and $145 billion .

Previously, the company had directed expectations to be between 130 and 145 billion, which essentially means raising the lower end of the estimates to 135.

Well, when you looked at Meta's cash flow statement for the nine months ended, meaning they only had one quarter left, the company spent—and I dare say "only"—$83 billion on capital expenditures, or the property, plant and equipment item was only $83 billion.

Well, folks, they just told you they were going to spend 135 to 140 billion all year, and they've only spent 83 billion dollars so far. This means they will be spending huge amounts of money in the fourth quarter.

Part of it may be related to data centers. They also have rental payments on this side. I mean, I don't know for sure. All I know is that the company will spend a fortune in the fourth quarter.

Now, the other factor that is affecting, and this is certainly what has been putting pressure on Meta shares over the past week, since they announced their earnings. These Class C shares are non-voting shares.

We don't need to go into the finer details of this matter. Google has done something similar in the past. This is not a big deal from a voting perspective. You will always be passed over from a voting perspective when it comes to Mark Zuckerberg and any of the companies run by their founders.

But what's interesting is that we have no idea how many Class C shares will be issued, which means that as a regular Meta shareholder, whether you own voting shares or not, your stake will be diluted.

This often means that more and more easing is coming. What does that mean? For every share you own in Meta now, more shares will be issued next year. We have no idea of the quantity,

So, how will Meta finance it? Well, part of that comes from direct cash flow. The company has $96 billion in free cash flow over the past 9 months. This does not include the next best quarter for advertising, in which they only spent $83 billion.

So, they will spend all the money that the advertising business brings in, but the other thing they will do is issue more Meta shares. This reduces your share. It also creates a new seller of shares in the market.

And it creates more supply . More supply usually means lower prices.

Now, what was also interesting on the earnings call was how much they talked about "Muse," and that reminded me of two or three years ago when they were talking on Meta calls about Metaverse, how everyone was going to use it , and that it was just too early, and we simply didn't understand.

Well, in the month and a half since the launch of the Muse AI agent, more than 10 million people in the United States and Canada have set up their personal Muse agents , but only about 25% of those who set it up use it on any given day.

I actually think , once again, that this has not been introduced to the wider world in part because of the cost of computing. What we need to understand about Muse's AI proxies is that they are not cheap to operate.

This is not like your Instagram feed . This is not like, you know, just your Facebook feed or even your WhatsApp conversations. This is extremely expensive computing required to provide you with one of these "Muse" agents.

Not only that, but you cannot release it globally. There is no profit mechanism in place yet , although they talked a lot on the earnings call about these things being great as shopping assistants and all those kinds of things.

However, this is a particularly light use in my opinion of something that has received so much hype.

We are also likely to see more competition with Anthropic's unveiling of their beta version. OpenAI has a version of this. Rumors suggest that Google also has a version of that.

Now, Meta is in the best position here, but it's not as if you won't face a lot of competition. If you can't put this up very quickly, it gives time for that competition to get in.

Meta's shares all fell to the bottom. Look at its closing price. Look where Meta closed. So, Meta was here at the top retesting the peaks. I was literally right here at the top, at the highest meta levels.

Would you have gotten over it? If you hadn't gone beyond it, you probably would have done exactly what happened, which is to come down here and retest the