Introduction and Motivation
The work of the Real-World Assets (RWA) workforce is currently financed by the Maker Foundation under the Collateral Onboarding Grants. Those grants have limitations and aren’t expected to last forever and, even if the exact ending date remains unclear, the DAO needs to start thinking about viable paths to finance these efforts and ensure it’s continuity. In the specific case of RWA, a lot of monitoring and coordination work needs to be done on an ongoing basis to successfully onboard these assets.
The objective of this document is not to provide a framework for the DAO to finance and structure the domain teams. Another working group is working on that. As discussed with the autonomous working group, the idea is to start somewhere in the continuation of the DAO operating budget (which is more targeted at unexpected things than financing domains). The discussion and the results of this experimentation will then feed the work of the working group.
On a personal note, having visibility of fewer than 2 months is not ideal. Especially as I have a client willing to take me back on a yearly basis for higher compensation. I also find it difficult to onboard people without being able to give them any visibility.
Those are the objectives that I think the team should address in 2021. Governance can change them and it sees fit if needed.
Increase RWA loans to 300M DAI
Currently, the amount of loans originated from RWA is zero. Considering collateral types that are quite advanced in the onboarding process (6S, New Silver and ConsolFreight), we might soon achieve $22M in loans (in Q1 2021). Once those first targets are hit, the processes ironed, and the Community finds itself more comfortable with these assets, those vaults can increase significantly (6S is targeting 100M after 12 months of operations, New Silver has room to grow as well). Moreover, we are already exploring new partnerships with various asset originators (PaperChain, BlockSquare, Uprets, Peoples Company, …). Therefore, $300M of loans seems like an ambitious but achievable target with the current deal flow momentum.
At a 3% average SF, that would correspond to $9M revenues on an annual basis. The number of vault types should be between 5 and 10.
Set up the monitoring infrastructure
RWA collaterals are a portfolio of assets managed by a third party: the asset originator. Being outside of any regulation, for the most part, we need to ensure that our assets (DAIs lent to asset originators) are used soundly.
We are currently working with asset originators to have access to their loan tape and building the infrastructure to monitor everything. We also expect to use third-party services to have independent updated valuation (real estate prices, credit ratings, …).
Find the best investment conduit for RWA
Currently, we are exploring the Centrifuge model and the Trust model. We also need to explore how we could set up a legal subsidiary.
Finding the best way to invest in RWA is of prime importance to scale safely.
Define a good solvency model for RWA
Currently, MakerDAO uses risk models at the collateral level and doesn’t define solvency at the MakerDAO level. The uncorrelated nature of RWA makes it a strategic advantage to be able to define solvency at the MakerDAO level. Indeed, if we want to compete with real banks, we need to have the same tools (and exploit some of their limitations). This is also including balance sheet manipulations.
Having a good solvency model will be a wonderful tool to earn more revenues while keeping the risk in line.
Diversify, diversify, diversify
Corollary to the solvency model, RWA aims to enjoy a strong diversification. RWAs are naturally somewhat uncorrelated to crypto-assets but, within RWA, we should also aim for diversification. The best target might be new kinds of assets, unserved by banks, with good growth prospects (Paperchain, for instance, that use big data and machine learning to safely advance money on a future invoice). Nevertheless, established asset types are more likely to provide higher debt ceilings in the medium term.
While it is likely that we will mainly work with small businesses/startups, we also need to work with more established players even if that means lower revenues.
To achieve this strategy, the following cost structure is proposed to MakerDAO. It is meant to be a max limit and not what will really be spent.
From a top-down approach (starting from revenues), we can expect 725k annual revenues with the first three collaterals at the initial debt ceiling (see table below). RWA should be profitable so we suggested that costs are hard-capped by revenues.
Moving forward, with RWA loans increasing, the cost ratio of servicing the RWA collateral should be decreasing. As the target for end 2021 is $300M of assets, that leads to a $9M annual revenue stream at a 3% stability fee. The aim would be to keep costs below 10% of the revenues (costs below $900k on an annual basis).
The bottom-up approach starts from the current cost structure and the need to define the cost structure evolution.
Current team composition
The RWA team is currently (or more precisely might be composed at the end of January) of:
Sébastien (application as risk facilitator): full time
William: full-time risk contributor
Phil: Part-time, our real estate expert
Jan: Part-time, our finance expert
James: Part-time, our farmland finance expert for P1-DROP
Rayan: Part-time, our finance laws expert
Adding some external cost (software, data feed for collateral assessment), we might end up with a $300k cost structure. My personal compensation would be $8k/month. I will let the working group define how compensations should be evaluated. Nevertheless $8k/month is the minimum that makes sense in my situation.
By merging the top-down and the bottom-up approaches, we end up with the following 2021 budget. Initial debt ceilings are expected to be filed on March 1st and the team cost grows from there.
|Month||RWA Revenues||Team cost||Cost/Revenues|
Based on those expectations, the 2021 budget line asked for by the RWA risk team is $400k (lower if Collateral Grant continues in 2021). This amount can be operated using the Keg or the suck function into a multi-sig wallet managed by mandated actors. All actual draws on the wallet will be documented.
Let me know what you think of it, so I can adapt it or move it to the next step.