Here's a fun market psychology lesson: a few months ago, anything with "AI" in the pitch deck was going to the moon. Now? Markets are systematically punishing companies that look vulnerable to AI disruption. And the casualties are piling up fast.

Veteran investor Ed Yardeni put it perfectly in a note Thursday: investors have moved from "AI-phoria to AI-phobia." Four major sectors are feeling the pain particularly hard—software, brokers, insurers and asset managers. The shift isn't subtle, either. We're talking double-digit declines across the board and valuation multiples that have been chopped in half.

The interesting question isn't whether these stocks are down. They obviously are. The interesting question is whether they're now cheap enough to buy, or whether the market is correctly pricing in a future where AI eats their lunch.

Software Gets Absolutely Demolished

Software stocks have taken the beating of a lifetime. The iShares Tech-Expanded Software Sector ETF (IGV) is down nearly 20% year-to-date, making it the worst-performing industry group in the market.

The panic started when investors began asking an uncomfortable question: if AI-native tools like Anthropic's Claude can handle specialized workflows in legal services, finance and sales, do companies really need expensive enterprise software subscriptions anymore? That's an existential threat to the entire business model.

Data providers got hammered especially hard. Thomson Reuters Corp. (TRI) shares have fallen 31.1% year-to-date and 57.6% from their high last summer. RELX plc (RELX), the parent company of LexisNexis, is down 30% this year and 47.4% from its May peak.

FactSet Research Systems Inc. (FDX) has slid 30% year-to-date and 57.3% from its December 2, 2024 high. S&P Global Inc. (SPGI) is off 25% in 2025 and 30% from last August's peak.

If you're keeping score at home, those are the kinds of declines that usually require a profit warning or a major scandal. But earnings haven't collapsed yet. This is pure fear about the future.

"For those who lived through the advent of the Internet, this feels like déjà vu all over again," Yardeni said. And he's right—this has that late-90s vibe where everyone suddenly realizes that disruption works both ways.

The valuation compression tells the story. Application Software now trades at 23.7 times forward earnings, down from 35.3 at its recent high. Systems Software trades at 23.3, down from 35.5.

So the question becomes: are these stocks cheap for a reason, or are they just cheap?

Brokers Face the "Why Do We Need You?" Problem

Investment banks and brokerage firms are dealing with their own version of AI anxiety. The trigger was fintech firm Altruist rolling out AI tools that can recommend personalized tax strategies. Which sounds innocuous until you think through the implications.

If AI can optimize taxes today, what stops it from handling broader financial advice tomorrow? And if that happens, what exactly is the value proposition of a traditional financial advisor?

The S&P 500 Investment Banking & Brokerage index—tracked by the iShares U.S. Broker-Dealers & Securities Exchanges ETF (IAI)—is only modestly positive year-to-date. But individual names have gotten crushed.

Raymond James Financial Inc. (RJF) shares sunk 9% on Tuesday, the company's worst session since March 2020. Charles Schwab Corp. (SCHW) also dropped 8% the same day.

The industry's forward P/E has dropped from 24.7x to 15.9x. That's not a small adjustment—that's the market repricing the entire earnings stream. And this isn't really about current profits. It's about long-term margin pressure.

If AI reduces the friction in financial advice, traditional advisory models face structural competition. The kind that doesn't go away when market sentiment improves.

Insurance Brokers Wonder If They're Next

Insurance brokers are having their own existential moment after reports that AI-driven tools are being integrated directly into conversational platforms to generate personalized insurance quotes.

"OpenAI approved an insurance application for ChatGPT developed by the Spanish digital insurer Tuio," Yardeni highlighted as the key news disrupting the sector.

Think about that for a second. If underwriting, comparison and quoting all become seamless within AI interfaces, where exactly does the broker fit into the equation?

The S&P Insurance Brokers industry index—closely tracked by the State Street SPDR S&P Insurance ETF (KIE)—has fallen 4% year-to-date. Major players have dropped more sharply from their peaks.

The counterargument is that insurance remains relationship-driven and compliance-heavy, which are areas where incumbents maintain real advantages. But again, the core issue is earnings durability. Will AI meaningfully erode commissions over time? Or will brokers successfully incorporate AI into their own distribution channels and maintain their economics?

Nobody knows yet. But the market is pricing in the pessimistic scenario.

Asset Managers: Collateral Damage from Software Collapse

Alternative asset managers might be the most indirect casualties of the AI selloff, according to Yardeni. The concern here is exposure—these firms have significant stakes in private software companies through both equity investments and private credit.

As public software valuations collapse, exit opportunities shrink. And that raises uncomfortable questions about the marks on private portfolios. Are those valuations still realistic if the public comps have been cut in half?

Several large alternative managers are down double digits year-to-date and significantly more from their late-2024 highs. KKR Inc. (KKR) and Apollo Global Management Inc. (APO) shares have fallen 16% and 11% year-to-date, respectively.

"Before worries about their exposure to software weighed on the stocks, they'd been facing investors' fears that credit losses in a wide variety of industries lurk in their private loan portfolios," Yardeni said.

Meanwhile, Blue Owl Capital Inc. (OWL) has already fallen more than 50% from its record highs. That's venture capital bust territory.

So Are These Industries Actually Cheap Now?

"These industries' forward P/Es have plummeted to ridiculously low levels relative to current earnings projections," Yardeni said.

And he's got the numbers to back that up. According to Wall Street consensus estimates for 2026, earnings growth projections remain surprisingly intact:

• Financial Exchanges & Data: 7.6% growth, 22.5x forward P/E
• Insurance Brokers: 12.2% growth, 15.9x
• Investment Banking & Brokerage: 14.4% growth, 16.9x
• Asset Management & Custody Banks: 15.4% growth, 15.5x
• Application Software: 17.3% growth, 23.7x
• Systems Software: 19.8% growth, 23.3x

Those multiples are dramatically lower than recent peaks. In several cases, they've fallen from the mid-30s to the low-20s, or from the mid-20s to the mid-teens. On paper, that looks like a legitimate valuation reset.

But here's the thing: the market isn't debating valuation in isolation. It's debating durability. Will these companies actually deliver that earnings growth if AI starts competing directly with their core products? Or are analysts still using old models that don't account for the new competitive landscape?

"Will AI competition trigger downward earnings revisions as contracts are renewed? That's the risk," Yardeni said.

And that's really the question, isn't it? These stocks look cheap if earnings hold up. They look appropriately priced—or maybe even expensive—if AI starts eroding margins and revenue growth over the next few years.

Which means this isn't a simple "buy the dip" situation. It's a bet on how quickly AI disruption actually happens, and whether these incumbents can adapt fast enough to protect their economics. Some will figure it out. Others won't. The market just doesn't know which is which yet.