Investment Philosophy · 2026
Conviction, not consensus.
A working document on how we read the on-chain economy and where we choose to deploy. For the firm walkthrough — sectors, portfolio, team — see the VC Deck.
01 / The Macro
Post-hype, pre-utility.
The 2024–2025 cycle did what every prior cycle did: distributed speculation widely, then exhausted itself. What remains is the rare cohort of teams whose work doesn't depend on attention to survive — and a market quietly rewriting the rules of capital formation around them.
Two structural trends drive our deployment in 2026. First, AI and crypto are converging on the same infrastructure questions: who owns the compute, who verifies the model, who pays the agent. The answer cannot be a single corporation. The answer will be a protocol.
Second, the regulatory and accounting machinery for real-world assets to move on-chain has finally caught up with the demand. Treasury bills, private credit, trade receivables — the boring instruments that move trillions — are coming to programmable rails. The platforms that win this transition will be the next generation of payments and credit infrastructure.
We do not compete on cheque size. We compete on the quality of the screen — inbound from operators we have backed, outbound from sector maps we keep current, and the discipline to say no to nine deals out of ten.
02 / How We Pick
The conviction rubric.
Every allocation passes through the same three filters. Most projects fail at the first.
01
Verifiable utility
Does the product solve a problem that requires a blockchain to solve well? If the answer is "a database would also work," we pass. The token must earn its presence.
02
Operator-grade team
Demonstrable, attributable history of shipping. Public credibility — code, prior exits, peer endorsements — over founder mystique. Anonymous teams must offer on-chain provenance in lieu of a CV.
03
Defensible token design
Tokenomics that survive a year of unlocks. Team allocation under 25%. Cliff and vesting that align long-term holders, not exit liquidity. Anything else is dressed-up extraction.
03 / What We Pass On
Discipline is what we say no to.
A short list — explicit, because it's easier to write down than to recall in the moment.
01Anonymous teams without verifiable on-chain history.
02Pivots-to-AI without architectural reason to be on-chain.
03Team allocations above 25% with weak or absent vesting.
04Projects whose entire surface is Twitter and Telegram.
05Yield claims that do not pencil under stress.
06Mercenary KOL flips and pure speculation tokens.
07Synthetic exposure marketed as real-world ownership.
08L1s without a throughput-bound use case to fill them.
Sector deep-dives
Six lanes. Each with its own bar.
AI Infrastructure & Agents
~25%The crypto rails for the autonomous-agent economy — compute, identity, payment, and verification primitives that a single corporation cannot own.
Read sector thesis →Real World Assets
~20%The boring instruments that move trillions — T-bills, credit, commodities — coming to programmable rails. The next generation of payment and credit infrastructure.
Read sector thesis →L1/L2 Infrastructure
~20%Modular execution, privacy primitives, and cross-chain demand-led infrastructure — bets placed on technical edge, not narrative.
Read sector thesis →DePIN
~15%Decentralized physical infrastructure with measurable supply-side economics — hardware-token alignment that survives unit-economics scrutiny.
Read sector thesis →DeFi
~10%Novel primitives, not yield-aggregators. Capital-efficiency innovation and the RWA-DeFi seam where most of the next decade’s flow will move.
Read sector thesis →Consumer & Gaming
~10%Games with real retention, social products with on-chain identity value, niche communities with a reason to live on-chain.
Read sector thesis →
04 / Value Beyond Capital
The work begins after the wire.
The cheque is the easiest part. What follows is the work most firms outsource or skip: token design, listing strategy, market structure, KOL coordination, and the slow accumulation of counterparties across more than a hundred partner companies.
The mechanics are detailed in the firm walkthrough. The short version: we do not write passive cheques.