Your Centsational Market Update July

Hi love,

This is our last update before the holidays, so today I want to keep it focused: one short-term thing that's on everyone's screen right now, and one longer-term story that I think matters far more than any single week of price action because it touches the entire AI boom that's been driving so much of this market.

Let's unpack it together.

Where the Nasdaq sits right now

The Nasdaq — the US index most heavily weighted toward tech — has been going through what's called a correction: a pullback from recent highs, driven mainly by nerves around how much money the biggest tech companies are pouring into artificial intelligence. Corrections feel uncomfortable while you're inside them, but they are a completely normal part of how markets move. Nothing goes up in a straight line forever. Looking at where the index is sitting today, we're inside a zone that stood out to me on the chart even before this pullback began — the kind of area that has, historically, tended to be where selling pressure runs out of steam. That doesn't mean the correction is definitely finished, and it doesn't rule out prices drifting a little lower or simply sitting here for a while longer. Nobody — genuinely nobody — can call the exact day a correction ends. I'm sharing this as context for what we're watching, not as a signal to act on.

The real worry underneath the correction

The "Magnificent Seven" — the handful of giant US tech companies (Microsoft, Amazon, Apple, Alphabet, Meta, Nvidia, and Tesla) — are spending enormous amounts of money building data centers and buying computing power for AI. That spending is called capex, short for capital expenditure: money a company puts into long-term assets rather than day-to-day running costs.

The fear is simple: if these companies keep spending at this pace, it eats into their free cash flow — the cash left over after a business pays for everything it needs to run and grow, which is really the cleanest measure of how profitable a company actually is. And yes, free cash flow across the big seven has come down. But it's worth sitting with this: it is still positive. These companies are spending aggressively, but they are still generating real cash while doing it.

There's a second layer: all of that AI capex doesn't just vanish a large part of it flows straight into semiconductor sales, since data centers need chips to run. When you lay the two side by side, the current boom in chip sales looks a lot like the internet boom of the late 1990s at a similar stage meaning, historically speaking, there may still be real room left to run before this becomes the kind of overbuilt bubble people are worried about.

China's AI models

Now for the part I actually want you to remember from this update, because I think it will matter for years, not weeks. Chinese AI models have gotten very close in performance to the leading American ones. One model in particular, called Kimi K3, now performs nearly as well as top systems like Claude or ChatGPT but at a fraction of the price. AI systems are priced by the token, which is simply the unit used to measure how much text or data a model processes. On a price-per-token basis, Kimi K3 undercuts its American rivals dramatically, for barely any drop in quality. This matters because of a very old, very simple rule of competition: when two products perform almost identically, people gravitate to the cheaper one. Right now, the American AI market is structured as something close to an oligopoly — a market dominated by a small handful of large players (think of names like Nvidia, OpenAI, or Meta, and executives like Elon Musk or Mark Zuckerberg) who currently have enough pricing power to charge what they want. A wave of cheap, high-quality Chinese alternatives threatens that pricing power directly.

We've actually seen this exact pattern before… in cars, and in motorcycles. Chinese manufacturers didn't need to beat Western brands on prestige. They matched them closely enough on quality while undercutting them sharply on price, and that reshaped both industries in the West. The data on how much traffic is already flowing from American companies toward Chinese AI models is rising fast, which raises a real possibility: that China could effectively dump cheap AI capacity onto the global market over the medium term — "dumping" meaning selling below what would normally be a sustainable price, specifically to win market share. If that happens, it's a genuine risk factor for the valuations of AI-focused companies, Anthropic included.

Why China is playing this game at all

This part explains the "why," and it comes down to economics inside China itself. Unlike the US model, where productivity gains have tended to flow disproportionately to company profits and a small number of very wealthy owners, Chinese wage growth has been tracking productivity growth quite closely: workers, not just shareholders, have been capturing the gains. Company profit margins there have stayed roughly flat as a result.

That's a deliberate policy choice, not an accident. China has a structural problem: its households save an unusually high share of their income around 45% of GDP according to World Bank data, compared to lower savings rates in the Eurozone and the US. High savings sound prudent, but for an economy, it also means weak domestic spending. Chinese retail sales growth has been sitting close to flat year-over-year, which tells you how cautious the Chinese consumer currently is.

So Beijing has two goals that reinforce each other: keep wages rising to coax more spending out of its own consumers, and build AI models that are genuinely competitive on the world stage. Cheap, strong AI serves both aims at once.

The bigger picture remains this:

In the short term, this sideways, corrective phase in tech could easily continue a while longer — and again, nobody can time it precisely, so that's simply not a game worth playing. In the medium term, though, keep an eye on this Chinese AI story. It has the shape of something that reshaped entire industries before, and it's worth understanding now rather than being surprised by it later.

My role here isn't to tell you what to do with this information. It's to help you understand why it matters, so that when the headlines get louder, you're not reacting — you're already oriented.

We'll be back with the next update once the holidays are over.

If something in this felt unclear, or made you think, just hit reply and tell me. I read every message.

I've got you.

Francesca 🤍