Why We Invested in Concrete

JPEG Trading has invested in Concrete's strategic round, led by Polychain Capital.

Our thesis is narrow, and it has almost nothing to do with yield.

The constraint nobody trades around

Ask why institutional capital has been slow to move on-chain and you will usually get an answer about yield quality, smart contract risk, or regulatory uncertainty. Those are real. They are also solvable by people who want to solve them.

The harder problem is structural. Registered advisers, funds, and a growing set of corporate treasuries operate under rules and audit requirements that keep client assets with a qualified custodian. That is not a risk preference an allocator can talk themselves out of after a good diligence call. It is a condition of doing business, enforced by their auditor, their board, and their regulator.

Permissionless DeFi, as a design philosophy, is the opposite of that. Assets move to the protocol. Self-custody is the point. Every mechanism that makes on-chain markets efficient assumes the asset is somewhere a smart contract can reach it.

So you get a standoff. The largest pools of capital in the world are structurally barred from the venue, and the venue's core design assumption is the thing barring them. Most attempts to close that gap have worked by making the on-chain part less on-chain: a wrapper, a feeder fund, a permissioned fork with three participants. Those products exist. They also throw away most of what made the underlying market worth accessing.

What AssetCX changes

AssetCX is a custody-integrated yield framework Concrete built with BitGo. The mechanic that matters: assets stay with the custodian, and the strategy still executes on-chain.

That is a different shape of solution than a wrapper. It does not ask the allocator to accept a diluted version of the on-chain market, and it does not ask them to violate the one constraint they cannot violate. It routes around the standoff instead of splitting the difference.

Whether this specific implementation becomes the standard is an open question. We think the odds are meaningfully better than zero, which is a higher bar than it sounds like in a category where most institutional DeFi products have been distribution plays wearing infrastructure language.

Vaults as the allocation primitive

The second half of the thesis is less about regulation and more about what a vault actually is.

Concrete's broader bet is that vaults stop being yield products and become programmable capital allocators: the unit an institution uses to express a strategy, with execution, accounting, rebalancing, permissions, and risk controls handled inside the vault rather than reconstructed by the allocator across six protocols and a spreadsheet.

We find that credible because we live on the other side of it. JPEG curates vaults. We know precisely how much distance sits between "there is on-chain yield here" and "an allocator can underwrite this." Most of that distance is not strategy design. It is auditable accounting, defined operational permissions, transparent risk controls, and the ability to answer a due diligence questionnaire without a two week research project. Those are the unglamorous things that decide whether capital can actually move.

Concrete is building that layer as infrastructure. We would rather own a piece of the layer than rebuild it privately.

Why us

JPEG runs the full asset lifecycle: first investment, underwriting, on-chain curation, and deep liquidity. We are a proprietary firm with no outside capital, which means our investments are our own balance sheet and our own conviction.

That vantage point is the reason we wanted this position. We see the friction at every stage, from an issuer's first raise to the moment an allocator decides they cannot get comfortable. Concrete is attacking the part of that chain we think is most load bearing and least well served.