RWA Strategy Update - Being more strategic

TL;DR: Moving forward, the RWF CU will use strategic value for work prioritization and no longer focus on the number of MIP6 onboarded.

This post is repurposing and expanding on the ideas of the last RWF report. This was also discussed with the RWA Committee and the team.

The previous strategy for RWA was:

onboard as many collaterals as possible using all those that are deemed interesting by Maker Governance. The objective is to learn as much as we can and be a leader in RWA financing within the DeFi space. We know that the solution is not yet perfect, but we focus on gaining market share.

Onboarding a lot of collaterals is also safer on the credit risk as we increase diversification.

On one hand, one could argue that not much progress was made. We moved from 2 vaults to 6 and only 2 are used for financing (New Silver and FortunaFi). Reasons are both internal and external to RWF. Yet, we maintained the lead in RWA, learned a lot, and improved the team. The more important part is that we now have deals coming not from non-crypto enthusiasts but from traditionals companies (Peoples Company, 120dB, Tinka, Société Générale, a major alternative asset manager coming soon).

The previous strategy was crafted of a few assumptions:

  • The main task for RWF CU is the Risk Assessment
  • Diversity is key for learning and reducing risk

The first assumption is completely wrong as we spent a significant amount of time understanding how to set up and create the lender structure (RWA Foundation). We are also spending a lot of time on how to structure RWA.

The second is fairly valid, but is not going to move us where we want to be fast enough.

We are spending a lot on dealing with the MIP6 pipeline and not much on creating the good structures and strategy. It’s obviously not black & white. But assuming we were working 80% on MIP6 collaterals and 20% on working on what RWA means for MakerDAO (including RWA Foundation, the short-term ETF work). The new strategy will be to change the numbers from 80-20 to 20-80 (or 50-50, only the trend is important here).

Being more strategic/focused

The Pareto law applies in DeFi. Excluding stablecoins, ETH is around 80% of the collateral and adding WBTC brings us close to 90%. Yet, we onboarded a lot of collateral over the last 18 months. Tinlake pools of Centrifuge are in a similar situation where 3 out of 9 are 80% of the TVL.

The same might apply to RWA, we use the same process, focused on collateral to onboard, pick one, work on it, onboard it, deal with it. If you look at the MIP6 it’s usually startups in alternative assets. To steal from @Planet_X : “this will not scale”. Moreover, the strategy laid out in the report every month was already clear. What wasn’t clear was the path.

While the long-term strategy is to simply invest in senior tranches of structured products rated by credit rating agencies, we recognize that it’s not currently possible in DeFi.

There was a problem with this sentence. I used the words “structured product”. Even after SocGen MIP6, I’m not sure Citigroup would return my call to buy RMBS, nor do we have Bloomberg Terminals. Yet, the debt capital market is a $119T market, so we don’t need to start with an arcane product.

We had discussions about that since June within the team. After a few months letting this idea mature (day to day issues are a great way to forget about being strategic), we now have a more strategic/focused approach proposal.

What does that all mean?

The best way to focus is to bring scarcity. We will limit RWA (under RWF CU) to 20 collaterals for the foreseeable future. We have already onboarded 7 assets (⅓ of the slots).

Those 20 collaterals will be split in 3 buckets:

  • Off-chain lending:
    • This is where we are doing private deals by converting DAI into $ in lender-side entity to lend those dollars. Those simply don’t move the needle.
    • Objective: flagship projects
    • Slots: 5 maximum, 3 remaining
    • Current collateral approved: 6S, SolarX
    • The RWF CU team propose to publish a RFP to help source something big
  • Off-chain investments
    • Same as off-chain lending but using off-the-shelves products from financial institutions. This is not to solve a borrower need, but solve a MakerDAO need, at scale.
    • Objective: Getting rid of the excess of USDC, managing the balance sheet.
    • Slots: 5 maximum, 5 remaining
    • Ideas: Short-terms ETF, investment mandate (ESG?) for reputable asset managers, asset manager as a Core Unit?
  • On-chain capital markets:
    • The future of finance is on-chain and MakerDAO is well positioned as having a big balance sheet. Currently it’s mainly Centrifuge (structured finance) and SocGen (bonds).
    • Objective: making DAI the base currency of the future of Finance.
    • Slots: 10 maximum, 5 remaining
    • Current collateral approved: New Silver, ConsolFreight, HarborTrade Credit, FortunaFi, Peoples Company

The numbers are a bit random, but that gives a guideline.

We are therefore asking Maker Governance to use each slot strategically. Each onboarding should raise the bar in terms of:

  • Strategic value
  • Counterparty quality
  • PR
  • Scale
  • Geography
  • Diversity
  • ESG
  • Risk/Yield

We are currently working on 9 risk assessments that we will finish. Some other great MIP6 are coming. Choosing will not be easy. It’s less than one per month that should be selected. But not choosing is making the wrong choice. Moving forward, we will provide less Risk Assessment as the monitoring of the current investment will take some time so we will focus on the strategic ones. Collaterals will be a means to an end. Maker Governance can always finance another CU to onboard RWA or to set some MIP6 as Core Unit (see my discussion about that).

PS - How RWA can be in 10 years

In the future, we might come to a point where all RWA are on-chain with standardized ERC-XXX with ratings coming from credit rating DAOs with economic incentives. MakerDAO will provide a buy/sell spread and some repo facilities for all those assets based on ratings and the internal ALM policy (maybe some on-chain machine learning who knows). Fully automated, fully permissionless, fully trustless.

It will probably not be like that exactly, but moving toward the future is better than doing more of the present.

21 Likes

Your top-line message: Strategic Value on-boarding > As-many-as-possible on-boarding makes good sense. Equally your 3 buckets seems right way to go. It looks like your team and you have a very good handle on how to project MakerDAO interest into RWA relationships and as such have something robust to allow scaling on. That is great place to move into a more strategic action set from. Well-done indeed.

Now I understand that long-term vision is large scale deployment into senior, structured, rated notes (and I understand why that is not easily available today), but I do wonder if there isn’t room/scope for a more ‘active’ RW asset/collateralization management strategy to pursue long-term? One should be able to design and deliver a portfolio that delivers a safe minimum absolute return and potential for alpha generation.

For example something like this sort of ‘working strategic asset allocation assumption’ one sometimes have to write when starting out a new asset manager/manage an asset side (and take no note of the individual factors - this is just a very crude example for illustration)

  1. RWA will expects to be responsible for deploying 500M 2021/5B 2022/15B 2023
  2. Duration of the portfolio will be below 72M and average 24M across the portfolio with variance max of z
  3. The portfolio guideline composition will be:
  • 30% immediately liquid/money market assets with 3rd party providers
  • 30% Private Debt/Bond assets - 50% from own direct origination and 50% via 3rd party providers - x% hereof publicly trade-able
  • 20% Project Financing - portfolio managed via 3rd party providers within a strategic project focus (renewable energy, social housing etc)
  • 10% Real Estate something - own origination team
  • 10% Others/Strategic - managed directly
  1. Across portfolio we will require a geographic diversification of: xx% US /ww% Asia/Y% EU
  2. Across portfolio we will acquire and implement a partial hedging mechanism between portfolio currencies and DAI/USD
  3. Portfolio will be under minimum ESG requirement standard X and Y% will be required to be Impact Capital under requirement Z.
  4. Each asset class participant must be set to a PME target and managed accordingly
  5. All asset class participant must have max ‘xx measure of risk’ for breaching a minimum absolute return of y%
  6. etc etc.

Sort of like the asset side of modern pension fund/insurer I suppose - needs safety and close to guaranteed minimum annual returns, but need some scope for alpha generation also. At say 5B you could have quite a decent mix I’d think.

Or is the thought that this sort of portfolio is simply too difficult to tie up with a collateralized currency peg mechanism? And simplicity/clarity on the collateralization side is much more to be desired than any return/alpha generative capabilities of the portfolio? Or that with secured, senior, rated notes, when you crack it, can scale very fast and in a simple manner? Or perhaps the thinking is that building such an active portfolio is quite simply too difficult as a ‘crypto funder’ at this moment in time?

Forgive me if this has been the subject of former posts/discussions. I just want to understand the thinking.

8 Likes

This is very interesting @Allan_Pedersen,

But there’s another, in my opinion, fundamental question underlying all those points: is Maker’s main ambition that of reaching a reasonable risk-adjusted return to compensate liability holders (like a pension fund), or is it that of expanding the DAI footprint with extreme conservatism and benefit from the negligible cost of capital and implied operating leverage (a bit like central banks)?

@Porter_Smith from a16z had a (way more extensive) post on this a few days ago and, not sure you’re aware, we have opened a Real-World Sandbox project exactly to discuss those issues in the open. We have been all following and supporting the heavy work RWA CU has done, and in the spirit of decentralisation we are trying to expand and build alongside the existing effort.

I’m leading such project and would truly appreciate your participation in designing Maker’s lending guidelines for the future.

Luca

9 Likes

There was an indication of “investment mandate (ESG?) for reputable asset managers, asset manager as a Core Unit?”

I would say it’s really more about focusing on the macro level instead of trying to squeeze some alpha.

The stage is not about thinking of a strategic asset allocation yet. We are more at trying to figure out the building blocks. Allocating assets means understanding our liabilities. Your portfolio guideline seems credible. I think it is important to see how the space will evolve, especially on the on-chain stuff.

4 Likes

Got it! Thx

I also suggest that we need to really understand our liabilities and structure our investments to manage our liability structure. I believe that all of Maker liabilities are an on demand structure? In a nasty stressed market, what will be Maker’s liquidity needs over say over 1 day, 1 week, 1 month, 3 months…?

At this point Maker has far more down side from not being able to meet a liquidity run compared to the upside of getting a few more percent in yield from a outperforming investment portfolio. So lets manage it accordingly at least in the short run

8 Likes

I have been saying this for almost the past 1.5 years. That you can’t get to 1B 10M at a time (adding various tokens vs. being focused and improving ETH and wBTC deposits - which still has not been a focus), nor can you get to 1T easily at even 1B at a time.

  1. Primary goal I think for the protocol is to get as much DAI out there as possible. This along with PEG management is priority #1 (at least from the CUs I talk to).

  2. Earning a decent return for the risk taken on is probably priority (#2). This isn’t to burn MKR but really to build up a war chest reserve to deal with any issues trying to grow, and to back expenditures (this also really has not gotten much attention).

To do this requires a laser like focus which as far as I can tell still has not happened.

Right now I think a reasonable projection for RWA is .5B at end of year (even that is aggressive). 5B at end of 2022 also probably aggressive.

I am of the mind Maker needs a better plan to expand deposit TVL and DAI borrow in key asset classes.

If I had to pick on the RWA strategy I would pick.

  1. Scale - pick the things that are going to grow us as fast as possible, that are easy to scale.
  2. Reasonable diversification across RWF assets, and in geography
  3. Risk/Yield

PR should come on it’s own and ESG should be given some priority in the above 1-3 if possible.
Honestly should develop a project ranking system and have the DAO rank order the following:

The rank order should not be a 1-8 btw but we should be allowed to have 100 or 1000 points to allocate so we can get a signal as to strength (if possible) on these.

Key points to take away from my posts and views.

There is only so much productivity each DAI can buy, we need to use it in the most effective manner. This means focus should be on doing work that brings in scale, and yield. Why? Because we can use yield to expand manpower, and if scale is considered in every one of these projects (as the #1 component) we will be able to scale when needed. The last strong component should be a geographic spread. We don’t want to put all eggs in a single geographical or regulative/legislative region.

Everything else at this point comes in as lower priorities for me. Get to scale, get income and people and then we can easily broaden horizontally the structure to incorporate all the items above and expand more along the lines of the vision presented by @rune without growth and massive DAI expansion what Maker does won’t do much when trying to address real problems of systemic human endeavor sustainability.

Which really should be another key component of the above strategy. How sustainable are these RWF projects going to be. We should try to avoid any flash in the pan stuff and focus on the brick and mortar type building of the DAI financial superstructure.

7 Likes

I have been talking about things like this since I came to Maker now over 2 years ago. You can work for when times are good, but you better plan for when times will be bad. Maker today doesn’t even have 1% SB against outstanding. Lets just hope if the markets decide to dump that we find the keepers and liquidation system as well as the tx network itself can keep up. BT caused a 4% protocol loss. That was 6M, today the same loss would be 240M.

MKR was actually better capitalized then at 600M vs. now at 2.4B relative to outstanding at the time (150M vs. 6B now) and it still had to pause MKR flops to get 6M in capital ready to bid.

3 Likes

Great post, @MakerMan. Not sure if you’re aware, but @SES-Core-Unit is incubating a Core Unit to take care of everything liquidations related.

Here’s the video from last week: 🚀 Core Unit Launch Pod Sessions ||| Session #20: Sidestream Auction Services

3 Likes

100% agreed on this one. No amount of PR can really save Maker from loss rates, if they come (and they will). We need to be way more Risk Management focused. Leave the PR for Growth. Focus on quality, only that from here. When you have institutional grade credit quality and attract solid collateral the PR will be done by itself. No need to be a focus. Well, not from a RISK team.

Maker should be extremely happy with a 1-2% yield (on avg). But be absolutely intolerant against having PD exceeding 0.5-1% (tops) in the most stressful of markets.

The core mandate is the stability of a world currency. We lose the trust on this mandate, we lose everything.

15 Likes

I think this message has been lost in a lot of the recent discussions around yield farming, going out on the risk curve, and legal structuring.

I really think this needs to be the number one — and perhaps only — mission if we actually do decide to revisit core ideals.

6 Likes

I couldn’t agree more with you and @williamr on this.

2 Likes

@Eumenes speak about liquidity risk not about solvency risk.

The best way to avoid a liquidity run is to hoard a lot of regulated stablecoins (USDC, USDP, GUSD). We are quite safe on that front currently but it will change going forward. You can read more on the second part of my article on liquidity.

There is a tension between the wish to get rid of stablecoin exposure and the liquidity risk. Investing in short-term ETF is a part of the solution but that doesn’t change the problem completely, it’s just a different mix of regulatory risk and liquidity risk.

If the Maker community wants to be seriously risk-averse, then let’s only invest in money-markets type of products and get rid of the rest. You should vote against anything that promises more than 2% and is not managed by an institutional asset manager.

I recognize the problem of the Surplus Buffer as well. But governance voted against an increase. Not speaking that you should focus on profit-making to fill it (assuming issuing MKR to fill the SB is out of the discussion).

The strategy laid out in this thread is based on feedbacks from the governance (through what they voted, discussions, and comments on the forum). Therefore, I encourage governance (through delegates, MKR holders and the whole community) to have discussions.

1 Like

Thanks for formulating this strategy. It’s going to be a helpful document for community members & partners to use as a starting point for onboarding more RWA. Of course though the part I am most excited about is your 10 year view:

For DeFi to become the financial system and not just stay on the fringes we need to find a way to bring the real world into this system. Just passively buying assets and holding them in some legal entity is not going to replace the financial system in the long run. It’s 10x harder and 10x slower. I voted yes on onboarding ETFs and bonds (with a focus on ESG & gren tech) as a stop gap measure for that reason and putting this in a framework with target allocations for different asset classes I think is a good way to make sure we work on both.

My personal opinion on aside, Centrifuge of course is focusing on pioneering on chain securitization.

1 Like

I was not precise is referring only liquidity risk as opposed to solvency risk as they are closely linked. Typically when the market believes a bank has involvency risk, there is a bank run which causes liquidity issues and potentially insolvency.

My question is what percentage of assets does Maker put into ‘risky’ assets versus low risk/high liquidity assets. And how does Maker construct its asset portfolio given its liabilities risks (liquidity and solvency related)? Said differently what percentage of assets need to be very low risk and liquid to provide liquidity when needed? What level of risk can Maker take in its asset portfolio and maintain actual solvency (& perceived solvency by the market)? These are difficult questions but are essential to managing the Maker balance sheet.

Its also difficult to size the liquidity/solvency risk of Maker and the overall crypto-bank space (though more research can help here), so lets see how US banks managed their liquidity/solvency risk over the last decade (per Sebastien’s article on liquidity). The banks had the following liquid assets and other protections:

  1. Averaged around 10% cash reserves though they averaged about 15% since the COVID era
  2. Averaged around 17% US Treasuries though they increased to about 22% since the COVID era
  3. The FDIC provides deposit insurance up to $250K
  4. The Banks are regulated and required to maintain a certain quality of assets
  5. The Central Bank provides liquidity to banks in emergency situations (Lender of Last Resort)
  6. Banks typically have a sticky retail deposit base

So the typical US bank had around 30% of its balance sheet in (arguably) riskless assets in addition to providing FDIC deposit insurance and having regulators oversee its asset portfolio quality. Does this provide any guidance ?

2 Likes

Interesting @Eumenes,

I agree with most of your points. Having worked in banks my whole career, I can testify that solvency and liquidity are two correlated issues, although different philosophically and managed by different departments - liquidity under CFO/ Treasury and solvency under Risk Management.

I also agree that the commercial banking model is instructive and one to be looked at.

There are, however, a few fundamental differences I would like to highlight:

  • Commercial banks benefit from a very strong base of (retail) customer deposits; those are extremely sticky in nature and offer a fantastic fallback of cheap and stable funding - there is not even a shadow of such stickiness in DeFi and the underlying dynamics of each asset should be assessed carefully
  • Commercial banks benefit from the lender of last resort window with the central bank - when things go bad, and a bank has complied to solvency requirements, there is always a central bank ready to step in - that is why the bible of banking analyst is the Pillar III report of a bank, where liquidity risk is almost overlooked
  • Commercial banks have a cost of equity of c. 10-12% and average cost-to-income ratios in the >50% range - this means that their pressure to generate returns from climbing the risk ladder is very high (and that is why most successful banks are moving away from the lending model and going more and more into the advisory one - look at MS and JP as examples in the US), this is not the same for Maker where the cost of capital is negligible and the operating leverage very high
  • Commercial banks are highly regulated and centralised, this is bad in some way but also good in the sense that a lot of behaviours are enforced centrally - when interacting with TradFi, Maker naturally needs to stretch a bit and give a bit more trust than ideal to some individuals without the enforcement power of traditional institutions

The point I am trying to make is that there is, on the one hand, no need not to be ultra-conservative for Maker given its incredibly high operating leverage - growth has more benefits than risk, and on the other that the risks of making a mistake are much higher as highlighted by @williamr and @PaperImperium - because DeFi has a much more liquid nature than TradFi and there is no central pillar backing everything up.

We could even push ourselves to imagine that it could actually be Maker that central back-up pillar for the whole DeFi ecosystem.

2 Likes

Thanks Luca.

I added your two other key differences between traditional banks and Maker in terms of liquidity/insolvency risk.

I definitely agree with you that Maker shouldnt take much of liquidity/solvency risk as Maker has ‘incredibly high oeprating leverage’!

In terms of top risks for Maker to manage I think the following are essential:

  1. Liquidity/Solvency Risk - if Maker loses market confidence is the value of DAI, Maker is likely done. DAI value and liquidity must be preserved at all costs.

  2. Cautious/Moderate Growth Risk - I believe that the stable coin market has definite economies of scale and economic network effects. Thus at scale there will likely only be a handful of huge stable coin players and we all want DAI to be one such super player, or perhaps only one such player. In Highlander terms:

image

Thus fast growth is essential so Maker can achieve the economies of scale and network effects early. Slow/moderate growth is a high risk in this type of market structure as other stable coins can reach scale first and outcompete Maker.

2 Likes

Agree in full @Eumenes,

I think that the key will be choosing the risks ‘worth taking’ and those that are not. Profit optimisation might not be the best risk-reward path.

I think decentralisation helps naturally in this direction.

P.s. loved the quote

Regarding the growth strategy, I discussed that a stablecoin proposition is mainly a solvency/liquidity risk vs deposit interest rate (the risk-free, or no risk-added version, like DSR). You can have any point in this space to satisfy the need/want of different classes of customers.

This was also my conclusion:

How do you manage your balance sheet to meet the constraint of solvency and liquidity while optimizing the spread between funding and investment rate?

The winner will be a trillion-dollar company.

For low solvency/liquidity risk, the market is already full of “soon to be regulated” stablecoins. Only in crypto we can have insolvent stablecoins issuers and/or illiquid ones which provide rewards in governance token issuance. Therefore, I would say the sweet spot is somewhere in between as the ability to provide a deposit interest rate (assuming regulation cleans up) will be key to boosting growth and moving beyond the fiat-backed stablecoins competition. Growth has some great partnerships that will kick in only if DSR goes above 2% (I don’t see how we can achieve that in the current context).

I believe this discussion has some merits @SebVentures, although I see it decoupled from the RWA front (even if there are obviously connection points).

I believe that the treasury and liquidity management should be managed by a CFO-like actor within the protocol, not one that is at the same time engaged in improving yield on (a subset of) the asset base, and in mitigating specific (not institutional) credit risk in the meanwhile.

I think this is a natural part of the maturing process of Maker, but not one to be mismanaged.