*I’ve been drafting this for the past week, so worth sharing now given the discussion has started.*

*I think we should stick to the current agreed schedule of 4 hr until 100k and then 1/day post 100k, but with wider changes to get us to the long run situation.*

At present the MCR Floor value is increasing by 1% every 4 hours. If enough capital keeps flowing in to keep the MCR% above 130% then the MCR Floor will reach 100,000 ETH in August sometime. At this point the MCR Floor Value will revert back to increasing once-per-day as decided in the previous governance action. This post is to start the discussion on what we should do next.

Stepping back for a moment on the MCR mechanics. The MCR is actually the larger of two values:

MCR = max [ MCR Floor, f(cover amount) ]

At the moment the MCR Floor value is dominating and the more complex risk capital calculation, f(cover amount), is not being used. Without diving into the details of the risk capital calculation it can be approximated as 1/6 x Active Cover Amount, or a gearing factor of 6x.

Longer term the goal is for the risk capital calculation to drive the MCR, the MCR Floor is really just a short term measure to scale capital up to a critical mass so we can meet demand for cover. At scale we want the capital requirements of the mutual to be specifically driven by cover outstanding.

So we have a balance to strike, we want enough capital for a critical mass to meet demand, but we don’t want excess capital, as this is inefficient and could lead to issues maintaining existing capital. Discouraging excess capital also allows greater opportunity for the token price to move into the exponential section of the token price curve.

On a closely related topic, I strongly believe we should replace the complex risk calculation (currently performed off-chain once per day) with a simple factor of Active Cover Amount. This will allow the calculation to be fully on-chain and decentralised as well as being much more robust.

A more complex risk capital calculation can then be run off-chain at any point (by anyone) and parameters can be updated via governance as and when required. NB: any parameter changes would be done in a gradual fashion over time to prevent any shocks to the token price.

In general the gearing factor should increase as the mutual grows and sells more diverse range of covers. As a reference point, at massive scale AIG has a factor of around 60x and I believe we should be aiming for a factor of around 10x once we have $150m of capital and multiple products, quite a bit of judgement/art involved here though.

Taking all this into consideration I believe once we hit 100,000 ETH we should move to a more capital efficient approach and look to have cover amount drive MCR sooner rather than later. Ideally within 6 months of reaching 100,000 ETH.

So my starting point for discussion is as follows:

- Once MCR hits 100,000 ETH revert to 1 increment every 24 hours (no change from current)
- Develop smart contract changes to replace the complex risk capital calculation with a simple factor approach
- Start with a gearing factor of 4x to begin with, so 25% of active cover.

This means at an ETH price of $400 the mutual could sell $160m of cover before the MCR starts increasing again, with a maximum amount on any one system of $8m = 20% of MCR.

This implies to achieve growth the mutual must start diversifying and selling cover on a wider range of protocols. Which is fundamentally how insurance works best. The initial factor needs to be:

- High enough so that it doesn’t drive the MCR calculation when first implemented (ie Cover Amount x factor is still below MCR Floor);
- Low enough that it gives confidence claims can be paid (anything 6 or below should be fine for now); and
- Low enough that within a reasonable time it does start driving the MCR calculation so the mutual starts growing. At 4x, $8m of cover on at least 20 different risks is required.

I’m most certainly open to debate on factors and approach, I thought it valuable to propose something tangible to initiate the discussion.