Stormbit in the Cybereconomy

The thesis, taken literally

The argument runs in three moves. First: capitalism’s famous failures — leverage bubbles, housing, wealth concentration, climate, runaway technology risk — are not moral failures but capacity failures. Herbert Simon’s bounded rationality on one side, Mancur Olson’s collective-action problem on the other: we cannot discover, compute, or commit well enough for the invisible hand to do what the textbook says it does. Second: the bounds are now removable. AI dissolves the discovery and cognition limits; smart contracts dissolve the enforcement limit — for the first time an agreement can be executed by code rather than promised by a counterparty. Third: what follows is a cybernetic economy — one that self-regulates against the risks its own expansion creates — and the things worth building are its coordination mechanisms: customer protection embedded in code (their flagship example is Lido’s Dual Governance, a staker veto enforced at the contract level), machine-readable markets that AI agents can transact in — Coase’s transaction costs collapsing into tool calls — and market mechanisms generated on the fly for a stated intent.

One line of scrutiny before I use the rubric on myself. The thesis ends at a “zero-risk attractor,” and I do not believe in it: Frank Knight’s distinction — the one The Second Axis is built on — is precisely that some uncertainty is not measurable, and no amount of measurement infrastructure converges it to zero. But you do not need the limit state for the direction to be right. The direction is: stop leaving the second number implicit. On that, the fund and I are describing the same vacancy from two altitudes.

Where a collar already lives in this

The thesis’s first named failure — Challenge 1, the debt-leverage cascade — is my subject. Mispriced collateral relaxes credit, cheap credit inflates leverage, and the loop compounds until the collateral drops and the machine force-sells everyone at the bottom. Every lending protocol on-chain runs this loop today, executed with perfect fidelity and no discretion to be merciful. The Stormbit premise attacks the loop at its hinge: a borrower who owns a put owns a floor, and a position with a floor has no liquidation price. Protection stops being a promise about behavior and becomes a property of the instrument.

Read against the fund’s vocabulary, that is regulation through technology — consumer protection no regulator mandated, embedded at the mechanism level. The isomorphism with their own example is almost embarrassing to state: Lido gave stakers a contract-level veto over governance; a collar loan gives borrowers a contract-level veto over liquidation. Same pattern, one seat over — the principal-agent conflict between a lender who liquidates to survive and a borrower who must survive the liquidation, dissolved by code instead of arbitrated by it.

And the intent economy is not foreign territory either. A collar quote, in our engine, is already a machine-readable object: signed, dealer-executable legs — strike, expiry, price, signature — that software can fetch without a human in the loop. An instrumented market, in their words. If the coming economy is one where your agent transacts on your behalf, downside protection is precisely the kind of thing an agent should be buying for its principal: unglamorous, continuous, priced.

The honest audit

Then I did what the essay form owes the reader: I built the rubric and scored us, allowing the framework to say no. Five load-bearing pillars from the thesis; a score per pillar for where Stormbit stands today — not the roadmap, the today.

Stormbit today, scored against the five load-bearing pillars of the cybernetic-economy thesis. The rubric was built to be able to say no. Diagnosis and destination score high; the enforcement substrate and the dependency graph do not — yet.

The shape of the result matters more than the average. On the two pillars that are about credit — killing the cascade, expanding with the downside structurally bounded — the fit is deep, because they are the reason the project exists. On the two pillars that are about how the guarantee is held, the honest answer is worse: today the floor is a signed, dealer-executable quote, hedged by regulated desks on centralized venues. A cryptographic signature makes a promise non-repudiable. It does not make it self-enforcing. And the execution path runs through exactly the kind of concentrated infrastructure the thesis warns about.

One line summarizes the audit: a cybernetic thesis running on a pre-cybernetic enforcement stack. I find no way around that sentence, so I am publishing it. The gap between the two halves of the scorecard is not an embarrassment; it is the to-do list.

Nine ideas

What follows is the to-do list written as ideas — the ways a no-liquidation floor becomes native to the economy the thesis describes. None of this is shipped. Some of it is weeks of work on rails that already run; some of it is the harder, slower kind. All of it hands the user a power they do not have today.

1. The coded borrower right. Encode, in the loan contract itself, that while the put is live no liquidation path exists against the borrower — not as policy, as an absence in the state machine. This is the Lido Dual Governance pattern moved one seat over, and everything else on this list is downstream of it.
2. Floor-as-a-service. Publish the quoting engine as an agent-native endpoint, so any AI agent can ask “what does it cost to floor this position until October?” and receive signed, executable legs. Protection becomes a callable primitive of the agent economy rather than a page a human visits.
3. The intent auction. A borrower — or their agent — states constraints: floor at least here, financing at most this, this horizon. Dealers compete to fill against the signed reference quote. That is the thesis’s intent pipeline, verbatim, applied to credit: the borrower stops being a price-taker facing one number.
4. The second-axis oracle. For any covered position, publish the market’s live price of flooring it — the second number, quoted next to the first. The Second Axis argued the number is blank; this is the instrument that prints it.
5. The protection certificate. A signed, machine-verifiable receipt — floor, expiry, notional, counterparty — that any third party’s agent can check without trusting a dashboard. The right’s receipt.
6. The put as a line-item. Unbundle the floor from the loan and sell it separately to vaults and curators, so a vault’s disclosure carries both of its numbers: what it pays, and what its protection costs. Yield with a footnote becomes yield with a term.
7. The mechanism-aware structurer. Instead of a fixed menu of put-plus-call, an AI that searches the live surface for whatever structure satisfies a stated risk constraint — laddered floors, partial caps, spreads — and prices it. Market mechanisms generated per intent, which is the thesis’s sixth pillar stated as a product.
8. The floor-keeper. A delegated agent holding one standing intent — keep my floor at X percent of spot — that watches expiries, rolls the put, funds it within a stated cap tolerance, and simulates before it signs. What an institutional desk has and an individual does not.
9. The assurance pool. Book-level protection that executes only if enough depositors opt in to share the premium, refund-plus-bonus if the threshold fails — Tabarrok’s dominant assurance contract, applied to collective downside protection. The most speculative entry, and the one that would prove the coordination point best.

Order of operations

If the list has a spine, it is the first idea. The right comes first because it converts the project’s central sentence from marketing into mechanism: liquidation is a property of debt, not lending is either enforced by code or it is a slogan. The service comes second because it is the distribution — the cheapest honest demonstration that protection can be a primitive agents buy, on rails that already quote. The auction comes third because it is the market — the moment the floor stops being one desk’s price and becomes a competed number.

The reason to write this down in public is the same as it was for the second axis. The vacancy is visible now. The economy the fund describes — agents transacting, mechanisms generated, protection embedded — is going to include a floor under credit, because it is the single most obvious thing an agent should buy for the person it works for. Someone is going to build the coded right. I would rather publish the blueprint and be held to it.