Razor's Edge: Hire A Kid Groundhog's Day

The AI market definitely gives you groundhog's day vibes w Great Winfield "Money Game" overlap. AI Capex stocks are very much like a treasure hunt for most people trading today as they largely have not been thematically interesting for 25 years. Every analyst is basically no different then a retail trader cycling through what hasn't moved yet, and what is the next thing other people will buy in the rotational thematic game. Every new AI shovel discovery is effectively an IPO. They make the laser. They make the coolant. They make the copper cable.

They make the chemical gas you need to make the thing the other guy makes that makes the thing nvidia buys for the gpu that Openai needs for the datacenter Oracle is gonna build them.

We are definitely not at the stage of talking about earnings or cashflow for these things. Like every AAOI momo now will quote an aspirational topline number from the ceo as if that means something in industry where some market leaders doing that year in and year out been throwing off Gtlb levels of FCF.

AAOI had an adjusted EV the other day that was about what HPE paid for Juniper. This for a company that generated little over $100ml in rev this quarter and simply can't make money assembling leading transcievers or manufacturing them on decade old tech.

The market leaders in optics are worth over $100bl combined today and won't throw Zoom FCF till 2029 assuming everyone isn't losing money by then. Yes, everyone in capex isn't Sandisk going from losing money to 2x earnings in 3 months.

And as market kids look for shiny new toys the old ones are forgotten. But eventually you have ramped just about every pick and shovel you can find and actually need to do some qualitative work on what buildout investing is all about. For example, teens fwd earnings multiples 2 years out is not a sign of cheap in sectors where supply almost has no limits. Not all infra is the same. And winning business from name hyperscalers isn't special as there is effectively no other market. Apple supply chain investors should remember how that works long term.

The whole thing is an interesting exercise for older folks because on one end you have this huge advantage having invested in all these names before and thus can find them faster, but on the other end you are cynical enough about the economics to not want to own too long and are way more price sensitive.

Which begs the question....

What Comes Next?

Like if you are long Lite today and have the same % return in the last 3 months as a SNDK long whose biz did a 1bl in fcf basically last quarter are you gonna have the same kpis for your position over the next 3 months?

"Demand is strong and CPO in a few years..."

Meanwhile what is the SaaS Ai bear saying next quarter with the same numbers at -70% ytd? "AI gonna wipe out all workers eventually..."

At some point you own a business again.