Dynamic, autonomous protocol

Chesterton’s Fence

In “The Thing” by G.K. Chesterton, the author describes the classic case of the reformer who notices a fence, and fails to see the reason for its existence. However, before they decide to remove it, they must figure out why it exists in the first place. If they do not do this, they are likely to do more harm than good. Similarly, with Nero, we must first understand the reasons why other protocols have not replaced their fences with our shiny new ones.

• Computing things on-chain is expensive

But no so much on StarkNet.

• Adaptive control theory is hard

Which is why we’re fortunate to have some of the most incredible control theorists on our team.

• We are still early

The space is still young and ripe for experimentation, and while most are busy porting building blocks from the traditional finance world, more ambitious creatives are running experiments to address credible neutrality, fairness, and governance attack vectors.

The father of Kwonzi schemes once said “A decentralised economy needs decentralised money.” He’s not wrong in this instance, but it’s hardly the full picture. At Nero, we say, “Unregulated systems need self-regulating mechanisms.”

Governance is an attack vector. Furthermore, it doesn’t align incentives efficiently. Humans are error-prone and often make suboptimal decisions whose consequences are not immediately apparent and take time to correct, especially in this new, decentralised decision-making paradigm. This is not to say governance will be absent in Nero — only minimised in such a way that no actor would be able to significantly cripple it.

Targeting Collateralisation Ratio

Traditional money markets set the borrow rates dynamically as a function of the utilisation rate. Credit protocols like Maker set borrow rates manually to fund operations and account for the risk of that specific collateral asset.

We took a different approach, a theme that will persist throughout these articles. We identified the main objective as autonomously keeping the collateralisation of a given synthetic asset at a stable and safe ratio — the goal being to strike the perfect balance between capital utilisation and safety.

In this system, borrowers pay a base interest rate according to their CDP’s loan-to-value ratio. A nonlinear adaptive controller then adjust your interest as a function of the global collateralisation ratio for a given synthetic asset.

Everything Is Dynamic

We’ve talked about isolated margin multi-collateral systems. Each position is opened against a single collateral asset, and the risk parameters are carefully selected by governance in order to minimise risk. Cross margin systems don’t have it quite as easy. Positions can be opened against a myriad of collateral assets at once. Furthermore, other risk parameters like close factor affect the way liquidations are processed which may result in liquidation cascades and, subsequently, bad debt.

To solve these pesky issues, create a more powerful system, and resist governance attacks, we committed to automating other aspects of the system. Let’s explore what they are.

Base Rate

“Why can’t Nero use the controller *just* to set the borrow rates? 🤔” We could. But that would be suboptimal. Should a person with a very healthy collateralized debt position, say 40% LTV, pay as much as a person who likes to live dangerously, at 80% LTV? We don’t think so. So Nero calculates your base rate as a function of your LTV.

LTV Threshold

By now you know that Nero allows you to open a credit line against a collection of crypto assets, but how is your liquidation threshold calculated?

Each token has its own LTV threshold. Nero calculates the weighted average of the thresholds of the tokens that your position is comprised of in order to determine what your threshold should be.

Liquidation Penalty

Gas isn’t free, so in order to incentivise liquidators to, well… liquidate, there must be a monetary reward. The status quo is to use a fixed penalty, typically about 10%. From a liquidator’s point of view, this is great. From a borrower’s point of view, not so much. Instead, when your position is eligible for liquidation and becomes increasingly more unhealthy, Nero will gradually increase this penalty from the minimum up to the maximum until it becomes profitable enough for a liquidator to pick it up. The benefit of this is that it functions as an auction and guarantees the best possible experience for our users.

Close Factor

The close factor is the percentage of your debt that can be paid off by a liquidator. Nero took the same approach as with the penalty, whereby the close factor gradually increases with your LTV once it is no longer healthy.

But that’s not all. Nero was designed for scale. Can you imagine what would happen if we had to liquidate $200M worth of ETH in one go? The slippage would be immense. Nero’s dynamic close factor ensures that liquidations will occur over longer periods of time, mitigating negative feedback loops and allowing the market to recover.

Now you know how Nero is going to be the most dynamic, autonomous protocol.