tuta

Nasdaq-100 forward returns after one-day drops of 6.25% or more, like today’s (NDX, Jan 1985 to Nov 2026, n=33 days). Takeaways: ➤ 21 days later: average +5.2% vs +1.3% for any day. Index higher in 64% of cases vs 63%. ➤ 63 days later: average +4.9% vs +4.1% for any day. Index higher in 64% of cases vs 69%. By 3 months the edge is mostly gone. ➤ Further losses common: NDX fell another 10% or more within 63 days in 21 of 33 cases. ➤ In 1987 and 2020 a fast policy backstop ended the forced selling within two weeks. In 2000–01 and 2008 earnings and credit losses that policy couldn’t fix kept the index falling for months. Today’s trigger was the midterms. Texas elected Hinojosa as governor on a promise to freeze data center permits, and voters elsewhere passed restrictions in 40 of 41 local data center ballot measures. The obvious comparison is 2000. But the 2000 crash came from collapsing demand, while this comes from expected policy changes, and expectations can reverse. The closest precedent is April 2025: NDX fell 6.1% on tariffs, Trump hit pause three sessions later, and NDX was up 30.5% after 63 days. Then one person caused the shock and could reverse it. Here expectations depend on a new governor and Congress, so a V-reversal is harder. There is a flipside. The same Democratic wave that threatens the AI buildout also limits Trump’s room to escalate in Iran. A deal is barely priced. Lower oil would mean a less hawkish Fed, and lower rates help long-duration tech the most.

Does $PRL look insane or is it just me

Why Pearl will Win –––––––––––––––––– 01:14 BTC is dying 03:24 Breaking up with my highschool girlfriend 06:51 Why it can work 09:35 Pearl's power footprint 12:45 AI capex > asic capex 15:00 FP8 19:00 Negative overhead 22:04 Elite team 23:06 Hashrate = hinge profile height 28:01 AI lab adoption 30:48 I want my MTV 32:31 Mark Knopfler shreds on guitar 35:52 Useful work vs Duplex work 38:32 Cheap verification is required for money 41:20 Can Pearl close the loop? 45:17 Addressing fundraise FUD 51:21 Risks 1:06:27 Pearl wins by becoming multi-denominational religion

Become addicted to finishing things. Not the idea. Not the plan. The finish. Start small. Close the loop. Do it again tomorrow.

Imo the single most bearish chart for the AI trade here Inference in only ~37% of compute usage Source @SemiAnalysis_

Great write up on Pearl, thanks for the shoutout @Bankless This article does a good job highlighting how Pearl’s duplex work is an elegant solution to BTC’s security budget problem. Bitcoin has lost its monopoly on the megawatt to AI, a secular trend @prlnet benefits from.

The bull case for Bittensor was an appeal to Maximalism “Think Bitcoin but with arbitrary reward functions”. This was a good first pass at a useful implementation of Bitcoin, it might still work. But Pearl has narrowed in on a more authentic design “Think Bitcoin with MatMul”, The difference manifests in complexity Bittensor is far more complex than Pearl. If Bittensor got it right, that’s not necessarily a bad thing. But if they’ve got it wrong, then there’s revision after revision after revision to get to an end state that might never fully work (read: what version of Bittensor are we currently on?) By nature of the simplicity Pearl is kinda a more clear moonshot. It might also fail, but wherever it goes, you can get there quicker. And if Pearl is “correct” you’ll also get much bigger

Pearl is coming to Hyperliquid

here's the matrix + reasoning: as expected, it scored the lowest. cool tech, but ~306M new PRL gets mined over the next year vs ~336M circulating, ~91% dilution. no burn, no buyback, fees go to miners, and the company sells inference in USD. the business can grow and the token captures none of it, which is the antithesis of the value investor framework.