Killer Bees - When Software Eats Software, and One Drug Kingpin Eats Another

The Sting

Here's the strange thing about this week: the market is simultaneously convinced that AI will destroy every software business and that Alphabet should spend $180 billion building AI infrastructure. Novo Nordisk invented the weight-loss drug revolution and is now warning of its first sales decline since 2017, while Eli Lilly the company that showed up late to the party just guided for 27% revenue growth. Sometimes the story isn't who built it first. It's who built it better.

The SaaSpocalypse Isn't What You Think

The iShares Expanded Tech-Software ETF (IGV) is down 19% YTD after a brutal 15% January its worst month since October 2008. ServiceNow crashed from an all-time high of $239 to $109. Salesforce is sitting near its 52-week low. Intuit is down 34% YTD. The trigger? Anthropic released some new plugins for its Claude Cowork tool on Friday that automate legal contracts, sales workflows, and marketing tasks. Thomson Reuters immediately shed 20% in five days. The market decided, in about 48 hours, that AI agents will eat enterprise software alive.

But here's what the panic trade misses: these platforms aren't really selling software. They're selling decades of encoded business process knowledge. Every dropdown menu in Salesforce, every workflow trigger in ServiceNow, every edge case handler in Thomson Reuters represents institutional memory accumulated from millions of users doing actual work. Software built to support legacy processes will probably die. The real question isn't whether AI can write code because it obviously can. The question is whether AI can replicate the business process IQ that these platforms have spent years learning.

Consider: Adobe now trades at a forward P/E of 12, down from a five-year average of 30. Either enterprise software is genuinely dead, or the market is pricing in an existential threat that will take years to materialize. The selloff has wiped $830 billion in market cap from the S&P 500 software and services index since January 28. Hedge funds are piling into shorts. Alphabet just reported that its Google Cloud business grew 48% in Q4, and it announced plans to spend $180 billion on AI infrastructure. If AI destroys software companies, who exactly is going to buy all that cloud capacity?

The market reaction tells you something important: this isn't about fundamentals. It's about fear. Software stocks are oversold by every technical indicator, RSI in oversold territory, valuations at multi-year lows. As BTIG's Jonathan Krinsky put it, the sector is "probably oversold enough for a bounce," but "it is going to take a long time to repair and build a new base." Translation: the trade might be crowded, but that doesn't make it wrong.

The Weight-Loss Drug Reversal of Fortune

Novo Nordisk pioneered GLP-1 drugs with Ozempic and Wegovy. They were first. They defined the category. And now they're guiding for a 5-13% sales decline in 2026 their first drop since 2017. Shares dropped 17% on Wednesday alone, erasing another $50 billion in market cap. Meanwhile, Eli Lilly the company that showed up second with Mounjaro and Zepbound just posted a 43% revenue jump and guided for $80-83 billion in 2026 sales, beating the Street's $77.6 billion estimate. Lilly's stock popped 10%.

The divergence is staggering. Novo's market cap has collapsed from $600 billion at its 2024 peak to about $196 billion. Lilly is back near $1 trillion. Over the past year, Novo is down 45%; Lilly is up 30%. Same industry, same macro, completely opposite trajectories.

What happened? Two things. First, Lilly's tirzepatide (the active ingredient in Mounjaro/Zepbound) simply works better 20% weight loss in trials versus 14% for Novo's semaglutide. Second, Novo got caught in a pricing vise: Trump's "Most Favored Nations" deal is forcing prices down, compounding pharmacies are selling cheaper knockoffs, and patents are expiring in China, Brazil, and Canada. Novo CEO Mike Doustdar put it bluntly: "People should expect that it goes down before it comes back up." Lilly, meanwhile, is leaning into volume growth and preparing to launch an oral weight-loss pill that could extend its lead even further.

The valuation gap has become almost absurd: Lilly trades at 50x earnings, Novo at just 15x. Either Novo is a deep value play or Lilly's premium is justified by superior execution. The market has clearly made its bet.

Chart of the Week: Tale of Two Pharmas

12-month stock performance in the GLP-1 wars

This table captures the brutality of execution in a competitive market. Novo invented the category and is trading at a discount to its five-year average P/E (30x). Lilly arrived second and commands a premium. The 6-percentage-point difference in clinical weight loss efficacy has translated into a $792 billion market cap gap.

The Hive Mind

FinTwit has declared a full-on "SaaSpocalypse," with software bears getting increasingly loud and short interest reportedly climbing. Meanwhile, Novo message volume on Stocktwits spiked 2,900% over the past year retail is watching the carnage closely. The sentiment is actually bullish on both Novo and Lilly, which suggests dip-buyers are circling. Options flow on software names is more balanced than the headlines suggest. Retail is loud; positioning is hedged. The interesting thing: Bessemer's Byron Deeter one of the OG cloud software investors posted "Chaos creates opportunity!" on X this week, saying now is the time to make bets. That's either contrarian wisdom or catching a falling knife.

The Thread

Two very different industries, one shared lesson: being first doesn't mean staying first. Novo pioneered weight-loss drugs; Lilly built better ones. Enterprise software defined the cloud era; AI might redefine it entirely. The question isn't whether disruption is coming it's whether incumbents can encode enough institutional knowledge to stay relevant, or whether they'll become the next cautionary tale about moats that weren't as wide as everyone thought. In both cases, the market has picked its winners. History will tell us if it was right.