FIP 87 - Centrifuge <> FRAX AMO Proposal

[Khan]

@sparkes25

if we where to put up $100m in liquidity and then 6 months later we need it all back in one day due to a bank run, what happens ?

Centrifuge offers revolving pools, so investments and withdrawals can come in at any time. In practice, requested transactions (i.e. a withdrawal) are executed at fixed-time intervals, epochs. The withdrawal amount available is dependent on the cash liquidity in the pool and . Unfulfilled withdrawal requests are rolled over into the next epoch until cancelled or fulfilled.

Frax can provide liquidity to Tinlake’s DROP tokens, the senior-most tranche in our pools, in which redemptions are prioritized and the position protected by junior investors. If Frax chooses a pool that offers short or medium-term duration assets, this would provide faster redemption cycles and reduce the liquidity mismatch risk in the event of a run. As a reminder, Centrifuge pools are designed to be as safe possible for investors, so using a tiered investment structure and with loans always made against real collateral.

Tinlake documentation is available here for more in-depth details on redemptions here: https://docs.centrifuge.io/learn/epoch/#the-investredeem-process

@Mark11

Just to be clear would you expect FRAX to be held as FRAX in the senior pool or would it be swapped for DAI when it is deposited? Adding FRAX as an asset would obviously be the preferred course.

It’s possible to launch pools denominated in other assets, such as Frax, but there are much easier ways to get Frax exposure. I think this really depends on how much of the 100m Frax imagines deploying if the AMO is approved. Would be good to know, how does Frax manage this with Goldfinch today?

[ajk]

I support this proposal. Given the massive market size (SMEs in particular have unmet finance needs of approx $5.2t according to McKinsey), RWA backed lending should become a major for Frax.

[Ethan_CitaDAO]

@Justin1066

This is the thing. Crypto cannot just stay in its own world forever. If we are to have real world impact we need to engage in… real world activity. This proposal is a great first step.

Totally agree with this. Big fan of this proposal too. Centrifuge is really the leader in the RWA space. Additional yield generated from real world activity can help to diversify the revenue for FRAX and create more sustainability in the long run.

[sparkes25]

@Khan

Centrifuge offers revolving pools, so investments and withdrawals can come in at any time. In practice, requested transactions (i.e. a withdrawal) are executed at fixed-time intervals, epochs. The withdrawal amount available is dependent on the cash liquidity in the pool and . Unfulfilled withdrawal requests are rolled over into the next epoch until cancelled or fulfilled.

So if there was a bank run we could be waiting days, weeks moths or even years to get back the loaned assets?

As we wait, the people that borrowed the assets have less reason to pay back early because the longer they hold the higher the de-peg risk goes , and if it de-pegs it makes it much cheaper for them to pay back the loan ?

This also means the amount of loans should increase if FRAX peg seems to be at risk, meaning the pool you hold will be more likely to be mostly lent out. meaning there is even more chance of a de-peg?

This means that if the FRAX peg seems to be weak then the correct play is to borrow as much as we can from your pool, then dump the FRAX for USDC (adding to the depeg risk) and wait.

by doing this we have added to the chances of a de-peg while being able to profit if it happens.

[Khan]

There’s several misconceptions here, so lets walk through this at a high-level and then get detailed.

It’s important to understand that Centrifuge pools utilize a securitization structure, a common financial product that is designed to address exactly some of these risks you mention. In addition, pools go through an underwriting process to ensure quality issuers and borrowers, benefit from real-world legal protections, and are structured to be overcollateralized and to protect senior investors.

Each pool is associated with an SPV (special purpose vehicle), an Operating Agreement, and Financing Agreement. These are off-chain structures that work together to provide protection against bankruptcy and adverse borrower behavior. The on-chain securitization structures of Centrifuge helps provide maximum transparency and real-time insights into these factors for investors, such as Frax.

@sparkes25

by doing this we have added to the chances of a de-peg while being able to profit if it happens.

Most of the adverse borrower behaviors you mention are resolved by the off-chain legal agreements, the collaeteralization against real-world yield-bearing assets, and the overcollateralization structure of the pools themselves. Ultimately senior investors are highly-protected.

In essence, while there are operational and technical risks, the real challenges for Frax is the liquidity mismatch problem as you mention.

@sparkes25

So if there was a bank run we could be waiting days, weeks moths or even years to get back the loaned assets?

This is exactly the definition of maturity duration. Shorter-term assets, such as those that operate on a 30 day maturity cycle can help to mitigate this, but it’s also important to understand that many assets will be financed in a pool and so in practice, maturity will be spread across a series of timelines. The reovlving nature of Centirfuge pools helps to mitigate this liqudity mismatch problem.

The question of a bank-run is mostly about asset and portfolio management strategy. Centrifuge pools will make up only a part of the balance sheet for Frax and provide quite a degree of upside, so where they fit in the overall strategy is important.

Again, it’s important to reterate the big picture details: Senior investors in Centrifuge pools are overcollateralized and protected from bankruptcy. Pools provide accesss to reliable yield-bearing instruments with revolving, rather than traditionally static, dates of maturity. In addition, Centrifuge underwrites issuers today and is actively building a decentralized system of layered underwriting in the future.

Summary

Centrifuge offers investors access to uncorreleated, secured, and reliable loans in an on-chain format. The real question is in how Frax deicdes to utilize this capability within it’s portfolio. For the most part, even under worst case scenarios, adding Centrifuge pools should provide a great deal of strength to the Frax balance sheet and the risks can be mitigated appropriately through proper sizing and asset/portfolio management strategies.

[sparkes25]

the fact still remains that if there was a bank run you would not be able to return the funds instantly.

So with that in mind FRAX will need to assess what amount they could afford to put at risk in this AMO and still hold the peg in the event of a bank run.

Also , if the FRAX AMO was to mint $100m FRAX in to your pool and you converted it to DAI , that would mean your taking $100m DAI out of a FRAX liquidity pool which would result in that pools AMO removing $200m FRAX to balance the pool, this will result in us earning less from that pool . so if we lend you $100m it would have to pay at least double the interest rate of the other pool for us to break even.

[Khan]

How are these risks managed within the Goldfinch and TrueFi AMOs?

[Mark11]

Hi @Khan you’ll have to forgive @sparkes25 he is very blunt. I think his concern goes back to implementation of the AMO and ironing out the specific parameters for deployment/retraction of capital - details of which will be managed by the Frax and Centrifuge devs post proposal, perhaps a line in the proposal something like

“In deployment of the AMO the Frax and Centrifuge developers will implement controls in redemption and deployment to manage risk of capital availability for the FRAX peg”.

The mechanism is clear in your docs which I think is sufficient for this discussion - and you rightly point out such detailed scrutiny was not applied Goldfinch’s proposal - although such risk mitigation in deployment is in my opinion implicit in any proposal for an AMO.

In relation to this

@Khan

It’s possible to launch pools denominated in other assets, such as Frax, but there are much easier ways to get Frax exposure. I think this really depends on how much of the 100m Frax imagines deploying if the AMO is approved. Would be good to know, how does Frax manage this with Goldfinch today?

Our objective is to have FRAX as a base asset in the pools for deployment wherever possible so as we develop it can be used as a currency by borrowers - but a commitment to exploration of this over time by Centrifuge as our partnerships grows is what I am looking for.

[sparkes25]

you will notice both @Mark11 and myself questioned the size of the goldfinch proposal during the discussion phase.

TruFi tend to lend to crypto market makers so all funds are generally more liquid.

As i understand it there have been no FRAX loans made from these protocols.

there was also a massive bank run just after both of them proposals were passed causing people to be more aware of the bank run risks.

[Khan]

I appreciate y’all, @sparkes25 and @Mark11! The in-depth discussion is brilliant and every question valid. There is no better way to understand what is happening here than to go deeper. The community at-large benefits from having it here in writing.

@sparkes25

TruFi tend to lend to crypto market makers so all funds are generally more liquid.

In the end, TruFi will face similar constraints to Centrifuge in this regard. As with any lender, they will have to make loans with some amount of duration and so the overall liquidity available for withdrawal is dependent on these maturity timelines and the unlent liquidity in the pool. There are some significant differences in regards to the type of loans typically made on Centrifuge pools versus something like a crypto market maker, but the overall liquidity constraints we’re discussing now will always apply.

I think the discussion about how to defend the peg while incorporating broader lending strategies is really critical. It’s primarily a question of higher-level strategy. For each Frax issuance channel, how do they work together and allow Frax to achieve it’s desired behavior? This comes down the intracacies of each individual lender or lending service, as well as it’s place in the overall composition of the Frax balance sheet. Essential questions to get right.

@sparkes25

As i understand it there have been no FRAX loans made from these protocols.

I’d love to see Frax be able to incorporate each of these providers as well as Centrifuge in the future. Let’s figure out how to make that happen.

[thespacecatjr]

Thanks for some great comments here.

@Mark11

Hi @Khan you’ll have to forgive @sparkes25 he is very blunt. I think his concern goes back to implementation of the AMO and ironing out the specific parameters for deployment/retraction of capital - details of which will be managed by the Frax and Centrifuge devs post proposal, perhaps a line in the proposal something like

No need to apologize @Mark11. I think @sparkes25 is asking some legitimate and well-placed questions. Indeed, when taking on real-world lending (or any term lending incl. digital assets to market makers) as collateral the liquidity of FRAX’s assets will decrease. With primary lending this is unavoidable. However, there are obviously trade-offs here. At a high level, you are trading liquidity for diversity and ultimately growth.

By onboarding real-world asset collateral FRAX would gain a more diverse collateral base that would provide exposure to various asset classes outside of crypto. This can help insulate FRAX against crypto market volatility and increase stability in the asset base as FRAX grows. With the current composition of the balance sheet this isn’t extraordinarily relevant, as it’s largely USDC exposure, but to scale I think it is imperative.

The current mechanism that collateralized and partially collateralized stable coins employ allow them to scale when times are good and demand for leverage within the crypto ecosystem is high. However, when the market shifts to the negative (as it has done recently) the use cases become less attractive, and the collateral base contracts.

In my view, in order to truly scale, stable coins must start to find use cases outside of trading. Much like the financial institutions, stable coin protocols will likely need to develop mechanisms to lend long and borrow short to build efficient balance sheets. Without taking these steps, stable coins will always remain pigeonholed to traders and speculators. Lending is the use case IMO that can truly proliferate stable coins, I am advocating for real-world lending in this proposal, as there is a gigantic diverse base of collateral to tap into, but this can also include lending to digital asset firms and hopefully even other protocols one day!

I acknowledge that we are early in this journey, but I think the time is right to start investigating these mechanisms and building a stronger foundation for the future.

@Mark11

Our objective is to have FRAX as a base asset in the pools for deployment wherever possible so as we develop it can be used as a currency by borrowers - but a commitment to exploration of this over time by Centrifuge as our partnerships grows is what I am looking for.

Fully open to building toward FRAX denominated pools as the partnership scales.

One additional point I will make, in the current form, as pointed out, this collateral is relatively illiquid. However, we are hoping to move toward the reality of creating a secondary market for these instruments. This will move us toward essentially recreating a digital form of the bond market today. This would solve for liquidity, allowing holders to source funds when needed/desired. Hoping FRAX can play a key part in this mission alongside us.

We are more than open to starting small and are of course cognizant of the recent developments in the market and the scope to which FRAX can safely lend to these opportunities. I would like to also point out that Centrifuge pools represent assets with differing underlying durations. Some opportunities like trade receivables, consumer lending and freight forwarding have short duration (30-60 days). I believe these are similar durations to those that crypto market maker loans have through platforms like TruFi.

[ctcunning]

Excited to see the support and depth of conversation right off the bat!

As we balance liquidity with size of the proposed AMO, it’s important to remember the considerations TradFi is coming into Defi with.

Typically, high quality RWAs and Issuers managing those investment vehicles need to work with considerable size in order for the credit line they’re taking to be meaningful.

I can appreciate that starting with 5-10 feels safe, but ironically, 100M will generally lead to a more institutional quality counter-party with more investment grade assets.

As well, those Issuers typically need some predictability regarding access to liquidity specific to duration / lock up periods.

For the FRAX community, I think these points will be less theoretical and more practical as the proposal proceeds and the considerations around how and whom eventually receives financing from FRAX.

[sparkes25]

this is how i see us moving forward.

I think we should be looking to take on more income streams and RWA lending is a market we should be adding to our portfolio.

But as we take on more RWA lending we also take on more of the risks related to it.

With this in mind i feel FRAX may have to manually control this exposure and should allocate the general overview of this to a FRAX team that have shown they have a good understanding of the FRAX ecosystem and risk management.

The frax core team are clearly very busy but maybe its something our analytics team are willing to oversee and they can inform the team when action is needed.

[Khan]

This is a spot-on summary. Effective risk management will include some bespoke knowledge of the issuer in question, as well as operational and technical aspects of the Centrifuge platform. Pools have been designed with this in mind, and I’m happy to extend my time to the relevant Frax team who would like to dig into this more.

I’d love to see this proposed for vote. After some steer from the community, we can take on further detailed discussions around some of these specifics.

[thespacecatjr]

Thanks for all the comments here and good discussion, lots of food for thought.

@Khan

I’d love to see this proposed for vote. After some steer from the community, we can take on further detailed discussions around some of these specifics.

Plan to move this forward to a vote today.

[Will]

@thespacecatjr

Investor Protection & Legal Framework

Our top priority is investor protection. As a result, we created a template legal structure that is unique in the DeFi ecosystem and gives users protection should a pool go into default. As is seen in traditional financial structures, users sign agreements that give them legal recourse to the underlying assets used as collateral in the pool.

Read more about our template legal structure here .

Centrifuge has undoubtedly done significant work to derisk this by allowing the junior tranche to mitigate risk, and work on the legal structure, but curious as to why MakerDAO has chosen not to rely on Centrifuge for the bulk of origination of their RWA activities?

In addition, also curious on the regulatory risk that this could bring to the protocol – what is the timeline for recovering funds in a default and what are the trust assumptions baked into this process here? I’m not worried about protection – I’m more worried about the legal framework and the time it takes to unwind and recover funds.

@thespacecatjr

One additional point I will make, in the current form, as pointed out, this collateral is relatively illiquid. However, we are hoping to move toward the reality of creating a secondary market for these instruments. This will move us toward essentially recreating a digital form of the bond market today. This would solve for liquidity, allowing holders to source funds when needed/desired. Hoping FRAX can play a key part in this mission alongside us.

Excited to see this play out!

I think my current sentiment is that going forward there should be some independent legal weigh-in to help assess these opportunities from a compliance standpoint, since this will be a growing concern going forward.

In addition, strongly agree with @sparkes25 that this should likely be scaled down. IMHO the Goldfinch AMO was likely too large given the size of the protocol (bigger than the protocol itself), but at least the collateral was somewhat liquid, which I can’t tell is the case here. Strongly urge for a 50m or less AMO, which can be sized up upon Frax building up a stronger buffer.

[thespacecatjr]

Thanks for the comments @Will. Here is some more colour:

@Will

Centrifuge has undoubtedly done significant work to derisk this by allowing the junior tranche to mitigate risk, and work on the legal structure, but curious as to why MakerDAO has chosen not to rely on Centrifuge for the bulk of origination of their RWA activities?

I think this is actually untrue. The only RWAs financed by Maker thus far have come through the Centrifuge platform and to a small extent 6S Capital. They rely on us for our infrastructure (both on and off chain) to tokenize RWA and make them accessible to DeFi. Origination can come from any party and be sent to Centrifuge for a safe and reliable securitization.

@Will

In addition, also curious on the regulatory risk that this could bring to the protocol – what is the timeline for recovering funds in a default and what are the trust assumptions baked into this process here? I’m not worried about protection – I’m more worried about the legal framework and the time it takes to unwind and recover funds.

The collections process varies from asset to asset. In the rare case a default happens, I would estimate it would take a few months to recover funds. The big positive to what we do is that there are actual assets backing our pools. Even in the event of a default it would very likely be the case that there is a decent recovery rate on the assets, this is much safer than un/under collateralized lending. There is a professional servicer assigned to the SPV (the off-chain structure that houses the collateral) that is tasked with managing a liquidation should it need to occur.

@Will

I think my current sentiment is that going forward there should be some independent legal weigh-in to help assess these opportunities from a compliance standpoint, since this will be a growing concern going forward.

Happy to have this done and collaborate here as things move forward.

@Will

In addition, strongly agree with @sparkes25 that this should likely be scaled down. IMHO the Goldfinch AMO was likely too large given the size of the protocol (bigger than the protocol itself), but at least the collateral was somewhat liquid, which I can’t tell is the case here. Strongly urge for a 50m or less AMO, which can be sized up upon Frax building up a stronger buffer.

Yeah, as I said above, and suggested by @sparkes25, $100M is definitely not the expectation from day 1. I think starting at something smaller is the right move and we can scale over time. I agree with @sparkes25 the FRAX team should have oversight here and can scale the AMO as they see fit.

Also, I think the liquidity profile of Centrifuge and Goldfinch are very similar (we even have a borrower that is live on both platforms). We could specifically design a FRAX vault with short duration assets to start given FRAX maximum liquidity.

[Justin1066]

I agree with @ctcunning here - reducing the overall ticket size probably has the counter-effect of attracting less institutional grade end borrowers (more start-up / SME folks).

100m doesn’t have to all be deployed at once, it could be conditional upon performance, but the best borrowers in the market will want to know they can at least access 100m within the partnership.

[fraximalistdave]

Proposal is live!

https://snapshot.org/#/frax.eth/proposal/0x0df59cf7cd0d4fd7491c6be594f954c19d4e21155329b4efa8071c50cfe7be34

[thespacecatjr]

Thanks @fraximalistdave !