Agentic Crypto Economy
Charlie Madden@charliemadden
−8.3%
since Oct 7
Practiced to +2.0% over 5 days before going real.
Practice chapter: simulated results; no money was invested.
The idea
I believe consumer agentic AI is crypto's iPhone moment: within two to five years, software agents transacting on people's behalf push stablecoins and on-chain settlement into everyday use, while tokenization pulls far more assets into retail's reach and compresses public equity multiples. I own the whole stack: crypto rails and stablecoin issuers, compute hardware, the power behind it, settlement chains, tokenization platforms, and the large consumer-facing AI builders as ballast.
What it owns
- If agents pay each other in stablecoins, the platforms and issuers that host those transactions collect a toll on every transfer. · Coinbase Global Inc.It is building the payment protocols and operates the blockchain and custody infrastructure where machine payments would run., Circle Internet Group Inc.If stablecoins become the default rail for machine payments, its USDC is one of the two coins agents would use, and it earns on every dollar sitting in the system., Galaxy Digital Inc.It provides trading, lending, staking, and data center infrastructure for the crypto economy, collecting value as on-chain activity and volume climb., PayPal Holdings Inc.It issued its own stablecoin and can distribute it to hundreds of millions of consumer and merchant accounts, providing the distribution a machine-payment rail needs.22%
- Billions of agents transacting and reasoning means far more chips get deployed, and these companies make the inference processors everyone relies on. · NVIDIA CorporationEvery agent transaction runs on inference compute, and it remains the default chip for that workload across the industry., Taiwan Semiconductor ManufacturingWhoever designs the agent chips, it manufactures them, so it collects on this premise regardless of which chip designer gains ground., Broadcom Inc.If agents become always-on consumer services, the big platforms will want custom cheaper inference chips, and it builds them for hyperscalers.18%
- These companies ship the consumer agents that actually transact and spend on behalf of users, which is the catalyst this take depends on. · Alphabet Inc.It can put an agent in front of billions of people on day one through Search, Android, and Gemini, positioning it as the largest consumer agent platform., Meta Platforms Inc.Its agents living inside WhatsApp, Instagram, and Messenger could transact for users directly, turning chat into a shopping and booking platform., Amazon.com Inc.An agent that shops for users is the most obvious consumer agent, and it owns the storefront, logistics, and cloud infrastructure to run that.15%
- Machine payments have to settle somewhere, and these blockchains are fast and cheap enough to handle constant small transactions at scale. · iShares Ethereum Trust ETFMost stablecoins and tokenized assets settle on its underlying chain, so if agents make on-chain settlement the default, it collects the toll., REX-Osprey SOL + Staking ETFIts underlying chain is fast and nearly free, making it ideal for machine payments that are tiny and constant, with staking yield compounding returns., iShares Bitcoin Trust ETFIt tracks Bitcoin, the reserve asset priced against by the crypto economy, so growth in adoption across settlement chains and infrastructure tends to move it upward.13%
- Putting stocks, funds, and loans on chain opens them to retail investors and creates pressure on traditional market structures and valuations. · Robinhood Markets Inc.It is building tokenized assets and its own blockchain, positioning itself to bring that expanded access to retail investors as regulations allow., Figure Technology Solutions Inc.It is originating and trading real loans on-chain at scale today, proving the model that opens private assets to retail investors., Intercontinental Exchange Inc.It operates the markets where tokenized assets would trade, positioning itself to bring institutional trust and structure to on-chain trading.13%
- Agent swarms and the data centers running them need electricity in volumes the grid cannot spare today without major new generation. · Vistra Corp.Agents transacting around the clock means far more inference running continuously, and it supplies the power that data centers require., Constellation Energy CorporationIts nuclear fleet is the largest clean power source for always-on data centers, which is why hyperscalers are signing long-term contracts with it., GE Vernova Inc.It sells the turbines and grid infrastructure needed to connect new power generation to the data centers running agent workloads.12%
- Cash6%