Market Lab Report - The AI race boosts tech stocks; IPO glut? Tech earnings rocket

The AI race boosts tech stocks 

AI is no longer just a product race. It is a defense race. Open models already run on machines no one country can switch off. So governments will not slow the build. They will claim a slice of it.

Data centers are the factories. Chips, power, and servers become national-security gear, the way steel and oil were. In a year or two, expect rules that reserve a chunk of that compute for defense and cyber, not only for chat apps. Cyber budgets rise because the same tools that defend a network can attack one. Each side has to keep spending so it is not the one behind.

That is the classic arms race. Whoever gets to artificial superintelligence first sets the terms. Second place does not get a cozy share of the win. So the buildout does not pause for power bills, town fights, or a soft quarter. Defense demand keeps the concrete pouring. Markets read that as more years of AI-stock bid.

**Short:** If ASI is the weapon, data centers are the arsenal. Countries that believe that will keep building. Rising tech stocks are the side effect.

IPO glut causes oversupply?

Some investors think an AI bubble would pop from too many new shares, not from fading demand for the technology. The cleaner lesson from 2000 is simpler: bubbles pop when central banks stop flooding the market with money.

Over the next year, U.S. share supply could rise nearly 5%, instead of the usual 1% yearly decline. SpaceX lockups expire. Anthropic and OpenAI may go public. The big cloud companies keep issuing stock to fund data centers. GMO strategists say each 1% rise in supply has historically cost the market about 4% over the next year, which would leave returns roughly 20% below normal.

That supply math misses the other side of a big IPO. A listing the size of OpenAI, Anthropic, or SpaceX is bullish overall. It pulls in new buyers, fresh capital, and attention, and it usually arrives because demand is already strong. Rising share count is a cost. A market able to absorb deals of that size is a sign of strength, not the pin in the bubble.

History points somewhere else. In late 1999, the Fed and other central banks pumped liquidity into markets because they feared Y2K would freeze payments and banking systems. January 1, 2000 arrived and nothing broke. Once the date proved to be a non-event, that emergency money was pulled back. The sudden slowdown in global liquidity — not a loss of faith in the internet, not the IPO wave, and not rising interest rates where the Fed began hiking in mid-1999 — is what popped the dot-com bubble. New shares and stretched valuations made the fall worse only after the money stopped.

The same pattern shows up again and again. Asset prices rise while central banks are adding liquidity and fall when they stop. The first two rate hikes usually underscore a strong economy, not a top. The old market rule “three steps and a stumble” warned that the third hike is the one that can mark a major peak though this is usually accompanied by slowing liquidity. Further, in a world sitting on record debt, the Fed is unlikely to get that far but even it did, global liquidity is a far better predictor of overall market direction in the leading tech-centric indices. The Fed was already using “not-QE” tools to moderate long-term yields and may need to reinstate the pace.

Big IPOs can land in the middle of that and still be a net positive. They are not what ends the boom. AI can still be the most useful technology ever built. The stock bubble ends if liquidity is drained, not if a few famous companies go public — and with this much debt, the Fed has more reason to keep money loose than to hike a third time.

Earnings rising like a rocket

Corporate America and the Street are getting more optimistic on themselves, and the tape is confirming it: FactSet says a record 72 S&P 500 names issued positive third-quarter guidance while only 44 guided lower, the fewest since Q3 2021 and about 30% under the five-year average. More than six in ten of those positive guides came from information technology, tying the group’s own record, with software and semiconductors doing the heavy lifting. Analysts lifted tech’s Q3 earnings estimates 3.5% over the quarter, and the sector is now expected to grow earnings 65%, the second-fastest pace in the index. The bottom-up S&P price target has climbed 3.6% since June 30 to 9,275, roughly 18.5% above Tuesday’s close, and analysts see all 11 sectors up more than 10% over the next 12 months.

A rising bar is a tailwind for leading stocks, but it also raises the stakes. Management still has the cleanest view of its own order book, and the fact that fewer companies are sounding the alarm is as telling as the optimism itself. With earnings around the corner, the trade is to stay with the leaders that are actually delivering, keep stops tight, and let results decide whether this optimism was earned or already priced in.