Saaspocalypse II - Electric Boogaloo
Volume 169 - The Week Ahead
Good morning,
I’m back after a week in Cannes, with a tan that won’t last and a phone screen full of alerts that made me question why I ever left. You take one week off and the Supreme Court tries to rewrite the tariff playbook (spoiler: it doesn’t matter), software loses another trillion in market cap, and gold decides it wants to be a five-handle commodity again. Magnifique.
I left the book as it was and only traded ES while I was away. Bought ES1 at 6797, six points off the lows, running 25% notional as a percentage of NAV with a stop at 6697 and a target of 7000. Sold into Friday's close at 6924.5 for a clean 127.5 handles. Not bad for a holiday trade…
Rebought yesterday at 6855 on the same sizing. Stop is 100 points below at 6755, target remains 7000. The thesis is simple: the range is holding, the lows are being defended, and I'd rather be tactically long into NVDA than sitting on my hands. If the stop gets hit, I'll reassess.
Anyway… let's get into it. There is a lot to unpack.
Software… We've Seen This Film Before. Or Have We?
I spent the first half of February writing about why software was not dead, why net exposure had collapsed to record lows, and why the biggest risk was being short into a squeeze. Two weeks later, we got a taste of that squeeze on the back of the Friday snapback... only for the sector to get absolutely obliterated again on Monday.
IBM down 13%, its worst day in 25 years. CRWD down 10% after Anthropic announced their Claude cybersecurity features. MSFT down another 3%. ServiceNow, Salesforce, Intuit, all getting dragged through the mud.
Here’s the thing, though, and I want to be very clear about this… the narrative has now expanded well beyond software. Trucking, logistics, commercial real estate, financial services, and even payments companies like DoorDash and Uber are getting caught in the crossfire. This is no longer a so-called SaaS repricing story. This is the market trying to price in the existential threat of AI to the entire services economy. And when the fear gets that broad and indiscriminate, history tells us we’re approaching the point of maximum pessimism rather than the beginning of a structural move lower.
The S&P North American Software index just posted its worst monthly decline since October 2008. The peak of the GFC. And the bifurcation in how funds are positioned tells the story better than any headline can.
The latest GS Prime data paints one of the more extreme positioning pictures I’ve seen in recent years, and it supports the thesis I laid out in Volume 168 almost perfectly.
TMT made up 70% of the total net selling for hedge funds last week. But within that number, the divergence is insane… funds are aggressively selling software and internet names while simultaneously buying semiconductors and memory. The result is hedge funds are now running their largest ever net long position in semis and their largest ever net short in software, at least since 2016 when the data begins. Record long semis. Record short software. If that isn't the definition of a crowded consensus trade, I don't know what is.
LOs finished the week $4B better for sale, bringing the February MTD total to negative $10B. That is the largest monthly sell skew for asset managers since March 2025 and one of the most significant selling episodes in four years. The other large sell months were these - August 2022 (Hiking Crusade bear market/re-rating $18B), March 2024 (profit taking? can’t even remember - $14B), and March 2025 (Tariff Tantrum - $22B). In other words, the asset management community is behaving as though we’re in a proper risk event, not a flat tape.
And here's the part that really caught my attention: globally, equities saw the largest net selling since liberation day... It seems people are taking down risk in what feels like preparation for index to finally reflect what single stocks have been saying for weeks. Which puts me at two thoughts. The first being that the whole move higher the whole time has been driven by people not having enough risk on, chasing, and buying what they don't own = more up. And the second being that maybe, just maybe, we are due a proper correction. More than 10%. The kind that actually resets things.
So where does that leave me on software?