Remember when everyone was freaking out about “circularity” in AI deals last month? Well, Wall Street has found a new word to lose sleep over: depreciation. And honestly, it’s kind of hilarious watching grown investors panic over basic accounting.
Here’s the deal: All those shiny, expensive GPUs and chips that tech companies are buying like they’re Pokemon cards? They might lose value way faster than anyone expected. Think of it like buying the latest iPhone – except these “phones” cost millions and might be obsolete in two years instead of six.
The drama started when some famous short-sellers (you know, the people who make money when stocks go down) started doing math that would make your high school algebra teacher proud. Michael Burry – yes, the guy from “The Big Short” – estimates that Big Tech is basically lying to themselves about how long their AI toys will stay valuable.
“By my estimates they will understate depreciation by $176 billion 2026-2028,” Burry tweeted, probably while cackling maniacally. He thinks these chips have a 2-3 year lifespan, not the 6 years companies are banking on. That’s like planning a 6-year car loan for a vehicle that turns into a pumpkin after 3 years.
But wait, it gets better. Peter Berezin from BCA Research did some napkin math and found that by 2030, tech companies could be sitting on $2.5 trillion worth of AI assets. If those depreciate at 20% annually, that’s $500 billion in yearly depreciation costs – more than their combined 2025 profits. Ouch.
The Nasdaq 100 is down 6.3% recently, and tech stocks have fallen over 9%. Apparently, the market doesn’t like the idea that all this AI spending might be the equivalent of buying a really expensive pet rock.
Kai Wu from Sparkline Capital put it in perspective: current AI spending already beats the internet boom when you adjust for GDP. And if you factor in faster depreciation? This AI buildout makes the railroad boom look like a lemonade stand.
Of course, not everyone’s buying the doom and gloom. Bernstein analyst Stacy Rasgon basically said “chill out, GPUs can work for 6 years.” But try telling that to investors who are suddenly realizing that Moore’s Law might make their billion-dollar investments feel very, very old very, very quickly.
The irony? We’re watching the market freak out over the possibility that technology might advance too fast. It’s like complaining that your sports car is too fast while you’re driving it off a cliff.
Whether this depreciation fear is justified or just another Wall Street anxiety attack remains to be seen. But one thing’s for sure: in a world where AI chips might have shorter lifespans than a TikTok trend, maybe it’s time to rethink what “long-term investment” actually means.