CryptoFund Radar

Only 12% of Americans even use X for news anymore (down from 15% in 2020).

If you're deep in Election Twitter, you're literally a minority of a minority—more obsessed with data than 99.5% of people.

Most folks have no idea what's actually happening. The echo chamber is real.

This applies to crypto too. When you're grinding CT daily, remember: normies aren't seeing what you're seeing. The alpha gap is massive.

Yan from Delphi on AI nationalization scenarios:

Two paths:
1. Full govt takeover = binary control
2. Govt investment = skin in the game

The second one is bullish. Think about it - if the state becomes a shareholder instead of a regulator, incentives flip. They want number go up.

This matters for $NVDA, hyperscalers, and any AI infra play. If Uncle Sam starts writing checks instead of writing laws, risk-on mode activates.

Watch how DC structures these deals. Equity stakes > executive orders.

PumpFun is back.

September fees hit $140M across all products — almost matching the January 2025 ATH of ~$150M.

The real alpha? Most of that is now coming from PumpSwap.

PumpSwap alone crossed $100M in fees for the first time. Volume up 50% YoY as $SOL memecoin degen szn heats up again.

If you're not watching PumpSwap flow, you're missing where the liquidity is rotating.

Focus on building what's next, not what's now. The real alpha is serving the builders grinding their way up—not chasing hype cycles. If your product isn't solving problems for people pushing boundaries, you're just another exit liquidity provider.

KOLs flipping their thesis every hour while quietly dumping bags from burner wallets.

You're watching them shill the narrative. They're watching their exit liquidity.

Trust the chart. Not the thread.

Ceteris dropping truth bombs: crypto doesn't need VC money like it used to.

Why? The infra's already built. We're done with that phase.

What we need now: scrappy app teams that can ship fast with minimal capital. Not another $50M raise for a team of 100 building vaporware.

The meta has shifted. Less dilution, more execution. If you're raising massive rounds in 2024, you're probably ngmi.

The real alpha on AI agents like $TOWN, Instinct, etc isn't the hype—it's the grind.

They'll brute force any task you give them until every path is dead. Relentless energy, zero sleep, infinite retries.

The catch? They lack context. They don't know when they've actually won. They just keep going.

This is both the edge and the risk. You get execution on steroids, but you need to define success upfront or they'll optimize for the wrong thing.

TL;DR: AI agents are insane workers but blind to outcomes. Use them right or watch them burn cycles on nothing.

Onchain options just hit different in September.

Notional volume DOUBLED to $4.83B. Derive still runs the show at 79.3% market share (down from 88.1%), but the game's changing fast.

Hypercall went absolutely nuclear: $9.4M → $538.8M in ONE month. Now sitting at #2.

Paradex opened their RFQ system to everyone and nearly 4x'd their volume.

August? Hypercall + Paradex combined were sub-4% of the market. Now they're forcing Derive to actually compete.

This is what real market structure looks like when protocols stop being monopolies. Onchain options finally heating up.

Infiltrated a Chinese OC syndicate that's been washing $1B+ for Lazarus Group.

Went undercover as a client. Collected enough intel to help freeze funds from the Feb 2025 Bybit hack and trace dirty money onchain.

This is how deep the laundering infrastructure goes. Not just mixers and bridges—actual crime syndicates moving billions for state-sponsored hackers.

More coming.

Kevin's take: AI might kill the debasement trade.

If AGI + agents + robotics actually help governments outgrow their debt spiral, that's bearish for $BTC long-term.

Right now, $BTC's main thesis = governments can't fix their debt problem, so money printer goes brr forever.

But if AI productivity boom lets them grow out of it? The macro narrative shifts hard.

Not saying it happens tomorrow, but it's a real tail risk nobody's pricing in.

$NEAR Intents just had its best fee month of 2024 in September.

Why? NEAR made Hyperliquid perps private by default. Cross-chain deposits now route through $NEAR Intents swaps to hit those positions.

Privacy = sticky liquidity. Users are keeping funds parked on NEAR Intents between trades instead of bridging out.

Confidential balances hit $70M on launch day.

Demand is converting to real usage. Not just hype.

RH launchpad tokens are absolute garbage.

Every single one bleeding out post-launch. Same pattern every time—hype the IDO, dump on retail, then radio silence.

If you're still aping into these thinking "this time is different," you're exit liquidity.

Stop feeding VCs.

Two paths for AI agents in commerce:

1. Pure alpha: Agent does deep research, finds you the actual best product
2. Pay-to-play: Agent just shills whatever advertiser paid most

Right now we're in the golden window where AI agents can actually deliver unbiased product recs. No ad networks corrupting the feed yet.

But it's coming. Same playbook as Google search, social media, every platform that started "neutral" then got monetized into oblivion.

The question: How long until $META and others turn agent recommendations into another ad slot? 6 months? 12?

Enjoy the pure signal while it lasts. This window won't stay open forever.

José's portfolio thesis is simple: AI plays, hedge assets, or nothing.

He's only holding:
• AI-exposed names
• Defense stocks
• Gold
• Crypto

Everything else? Out.

This is the macro setup—uncertainty is high, AI is the only growth narrative with legs, and hard assets are insurance against chaos.

If it's not riding the AI wave or protecting downside, it's dead weight.

Stress? Uncertainty? Just grind harder.

Two outcomes: stack more sats or unlock new alpha. There's no losing.

Bill Ackman calls it "compounding small wins daily." Brian Armstrong says it cleaner: action = information.

Stop overthinking. Ship. Learn. Repeat.

The market rewards builders and punishes waiters.

Feeling uncertain or stressed about your bags?

Stop doomscrolling. Get back to work.

You'll either stack more sats or learn what not to do next time. Both are wins.

The market doesn't care about your feelings. Your output does.

Running so many agents rn that my M5 Max (maxed specs) is plugged in AND the battery is still draining

Agent szn is here and it's eating compute for breakfast 🤖⚡

AI agents are about to flip every market on its head.

Imagine billions of people instantly comparing quality, pricing, and alternatives for ANY product or service. No more information asymmetry. No more getting ripped off.

This shifts massive value back to consumers and rewards high-quality businesses that actually deliver.

The inefficient middlemen and overpriced garbage? Dead.

At scale, this is one of the most disruptive forces we'll see. And it's bullish for transparency, competition, and real value creation.

The future is efficient. The future is agent-driven.

Barbell strategy in full effect this cycle 🎯

Ceteris breaking it down: going heavy on the extremes while dodging the dead middle.

One side: $ZEC and $HYPE (established narratives with actual conviction)
Other side: onchain experiments and micro-cap games (pure degen plays with asymmetric upside)

Everything in between? No man's land. No edge, no alpha.

Most mid-caps are stuck in limbo—not safe enough to park capital, not volatile enough for real gains. Why fight for 2x when you can aim for 10x or protect downside?

This is how you survive chop and catch the next leg up. Pick your battles.

Tokenized stocks hit ~$3B market cap but only 6% is actually being used in DeFi. The rest? Dead weight sitting in wallets or flipping fast.

The Problem: Liquidity Doesn't Exist When You Need It

Minting stock tokens is trivial now. Buy a share through a broker, wrap it, done. 84% of tokenized stock value works this way. Market 10x'd since early 2025.

But here's the issue: $TSLA alone trades under 5+ different tickers onchain. Kraken has xStocks. Binance, Coinbase, Robinhood all launched their own. Every version tracks the same price, so issuers compete on utility.

The real bottleneck? No standing bid when you need it most.

Lending protocols that accept tokenized $NVDA as collateral need to dump it instantly when price drops. If there's no buyer in that exact block, the protocol eats the loss. Most liquidity sits with issuers' trading desks, they quote for single orders then vanish. Nothing left onchain to sell into. Redeeming for underlying shares is too slow.

Some Solutions Emerging

A few market makers now post onchain liquidity 24/7. Backpack lets users swap brokerage shares for Solana tokens 1:1. Arbitrage keeps tokens pegged. Alpaca holds underlying shares and plans to open same swap.

DTC targeting October launch for tokenized stocks/ETFs/Treasuries. Most supply stays with institutions but could help market makers rebalance after hours.

The Endgame

Wallets and aggregators will route orders to whichever ticker has deepest liquidity. Fragmentation across 5+ versions of the same stock won't last. Winners will be the tokens you can actually borrow against and earn yield on, not the ones that just sit there.