[PC-DROP] MIP6 Application: Paperchain DROP: Tokenized Music Streaming Invoices

Do you have any estimate on when this might take place, and any blockers you can talk about here to it happening soon?

Is there a way to make sure this happens before the position gets liquidated? Or are you hoping that your bots are faster than the keepers’? Will be interesting to hear more about which Debt Financing companies would be keen to be onboarded as keepers specializing in buying distressed loans - that is cool.

Also, it’s enormously cool that you are making use of Delaware SPVs - props to you all for that.


The main blocker is user comfort with the tech. The sharing of this data is very strict and users aren’t in a rush to make this data public. It’s possible for the data to be anonymized and validated by the community but there’s no timeline for that. If the community can accept a validated signature from a source participating in a network, that’s also possible, but again, no timeline as of yet.

We’re excited with Spotify’s involvement in Libra Coin (leaving aside criticisms and concerns of the project in general), because it shows a willingness to explore these types of systems.

We want to stay focused on adoption and minimizing concerns from users who don’t understand the tech, but just know that that they need a problem solved.

Spotify have released their Q1 earnings report.
You can read the shareholder letter here.

High level:

  • Despite COVID-19, performance exceeded forecasts
  • 5% increase in active users from last quarter (286m)
  • $1.8B in revenue (17% YoY increase)
  • $472M profit
  • Hiring has slowed but still expecting to increase headcount by 15% for FY20

After an initial decline in listening, demand has increased.

“It’s clear from our data that morning routines have changed significantly. Every day now looks like the weekend,” Spotify said.


I feel this does not really answer the question. Where (provide link) does the exchange of the DROP token occur? Or is it planned as an over-the-counter thing? Or is Paperchain DROP still so much in the planning stage that this has yet to be determined? If this is planned as a subscription model I think we need to have this detailed. This is not to detract from your project which I regard as highly interesting but it is still better to provide all information possible.

1 Like

The token is not live so the final list of venues where it can be purchased is not yet clear. The Tinlake contracts have a way to accept DAI in exchange for minting DROP. We are launching a Dapp that can be used to do these direct investments in pools. We’re releasing this Dapp next week. This will be the primary way to provide funds to the pool and redeem funds. It’s a direct exchainge of DAI for DROP with the pool. We would love to see DROP being available in other places as well but the full list of venues where DROP will not be known until it’s closer to launch.


Hi Paperchain,

Thank you for submitting this application. We understand how long these things take, and appreciate your team making the effort to submit everything.

I’ve been active as of late, commenting on several applications, including the other Centrifuge-related application ([CF-DROP] MIP6 Application: ConsolFreight DROP: Tokenized Freight Shipping Invoices) and more generally, I’m trying to bring up highly relevant and potentially problematic legal and regulatory issues where I see them in the sincere hope that MKR holders do not blindly wave tokens onto the approved collateral list. MKR Whales and all other MKR Holders, the big M signal has appeared in the sky! Into the breach once more my friends…

As is becoming a habit, we’ll need more color from the projects so the community can be informed. To that end, can @pcDan or @spin please address the following issues:

• Several application excerpts jumped out at me:

The asset type we are proposing for inclusion in MCD is slightly different to the majority of collateral applications: the Asset Originators will be using MCD directly as a line of credit to originate new loans against freight invoices (invoice factoring and reverse factoring).

Along with the necessary technical infrastructure to bring these assets into DeFi, the Asset Originator sets up a legal structure that provides the necessary support to ensure that anyone that owns a DROP token has a legal claim to the underlying assets. This is done with a legal structure very commonly used in the traditional financial system: The collateral for the individual loans are assigned to a legal entity, the “special purpose vehicle” and lenders get an ownership interest in the entire portfolio of this entity (with this entity the assets are in an a bankruptcy remote structure that is not influenced by the Asset Originators).

  1. How is the applying collateral type currently used?
    Centrifuge Tinlake has been in development since early 2018. We did a first technical POC where a pre launch version of MCD on Kovan was used to deposit an ERC20 token that was backed by an NFT in April 2018. Only a few months later we deployed the first version of Tinlake to Ethereum mainnet and worked with five different Asset Originators to facilitate loans in DeFi. In a joint effort with MakerDAO to develop a tool to bring real-world assets into MCD we financed four of those Asset Originators with funds provided by the Maker Foundation totaling a volume of 250,000 SAI. The Asset Originators included a residential bridge loan as well as a mortgage originator, a financing solution provider for music royalties and a logistics platform providing invoice factoring. This allowed us to gain first experience with the setup and iterate on the smart contract and risk architecture. The current version of Tinlake has been audited and tested with two running deployments.

  2. Does one organization bear legal responsibility for the collateral? What jurisdiction does that organization reside in?

Paperchain has incorporated Paperchain Pilot LLC, a Delaware (USA) limited liability company (the special purpose vehicle, “SPV”). This SPV has been formed to finance Paperchain’s assets.
This SPV structure creates a bankruptcy-remote entity whereby owners, debt holders or interested parties of this newly created SPV are left unaffected by the parent’s financial, operational and/or legal health.

  1. (Optional) List any parties interested in taking part in liquidations for the proposed Collateral type.

The way this collateral type is used varies from how standard vaults are opened: DROP tokens have a stable USD price and any small fluctuations in the loan portfolio performance should be covered by the insurance provided by the TIN tranche. This means that under normal operation, the Asset Originator would not see their Vault get liquidated. A liquidation would only occur if a large amount of defaults occur across the portfolio that the risk model did not calculate.
In case the Vault gets liquidated, the Tinlake contracts enforce a rebalancing of the pool to bring it back to its required collateralization ratio and will not allow issuing any new loans. Instead the Tinlake contracts are from this point on taking all of the cash flows generated by the borrowers and disbursing these to DROP token holders. Trade finance assets are usually short term assets and the instance of Paperchain the entire pool can be liquidated in <55 days (45 day term invoices + a grace period for late payments).

Overall this means keeper liquidations will be a much less common occurrence that only happen when the risk model worked out for a given collateral type did not perform. In the debt finance world there are companies that specialize in buying distressed loan portfolios that can be onboarded as keepers for MCD that would be ideal candidates for buying any DROP that MCD wants to liquidate before the underlying portfolio is liquidated.

First of all, the above description of the DROP mechanics describes what most people around the world would understand as a debt security, or rather a transferable instrument representing a legal claim against the future disbursements of an entity’s proceeds in satisfaction of a debt. The security status of DROP is further raised by the uncertain cash flows into the SPV, following Vault liquidation, from the originator and Tinlake “rebalancing”. This presents a whole host of issues for MKR holders – not least of which are the implications of forced transfers pursuant to government action. Have Paperchain or Centrifuge undertaken any securities law analysis of DROP and/or obtained legal counsel regarding its status, including a legal memo reflecting such analysis that can be shared with MKR holders?

Second, while not fully explained or stated, it appears that potential DROP buyers (addressed below) purchase DROP exclusively through the MCD liquidation/auction process. I have only a lay understanding of this, but most jurisdictions generally require a regulated intermediary to conduct sales of securities. Certainly, in the US, one has to generally sell securities through an SEC-registered broker-dealer. MakerDAO is not such a regulated entity and, as far as I can tell, there is no provision for a broker-dealer in your plan to administer auctions. Has Paperchain or Centrifuge analyzed this issue? If so, what do the applicants propose as a solution to this? Is there some reason this is immaterial? What is it?

@spin suggests in a comment above that DROP may be available for purchase on the open market. How is this possible when each DROP will be specific to the Asset Originator’s deposited collateral (music proceed rights) with the SPV (begetting the detailed discussion in the application section “The Infrastructure beyond the Tinlake Contracts”)? How will DROP get to the open market but for an MCD auction after Vault liquidation, which grants buyers the rights to the SPVs assets?

Were the pilots you all conducted with the Maker Foundation, MakerDAO or both? Please note that the Foundation and the DAO are separate concerns. Also, it’s stated that Centrifuge and MakerDAO “develop a tool to bring real-world assets into MCD” – aside from Centrifuge’s public blog posts and the answer to the above question, what did the Maker Foundation or MakerDAO do to develop Tinlake, if anything?

Since the pilots were directly funded by the Foundation, were you still able to test DROP deposits, auctions and DROP holder redemptions (on Kovan at least)?

Who owns the SPV? Is it a Paperchain or Centrifuge controlled parent? Please name the directors, members, shareholders and management of the SPV’s parent.

Will you disclose the names of each ultimate beneficial owner of the SPV’s parent?

If the SPV cannot liquidate its assets to pay DROP holders, what’s next?

What happens if the SPV fails to hold the underlying collateral and is unable to redeem DROPs for the collateral or cash? What then? Who can keepers (or MKR holders, for that matter) pursue?

How does the structure ensure that the SPV is bankruptcy remote? And remote to whose bankruptcy – the originator’s? Tinlake’s? Its own?

Do the SPVs have guarantors? Typically, with SPVs, although they are bankruptcy remote (if structured correctly), companies will establish a guarantor to ensure payment to creditors. Has Paperchain and/or Centrifuge considered such a structure?

Will Paperchain (or Centrifuge) guarantee redemptions or is it publicly going to disavow liability?

Will Paperchain (or Centrifuge) provide legal insurance surrounding the SPV for its guarantees?

Do Paperchain and Centrifuge envision other potential keepers for the Tinlake program aside from typical purchasers of distressed debt, i.e. hedge funds? Regardless, what happens if other entities do act as keepers? This suggests there may be some whitelisting/blacklisting for DROP keepers; is that so?

Are there ideal characteristics of a DROP keeper? What are they and why? Is that because it is implicitly understood that DROP would be a security and thus necessitate certain types of purchasers to avoid violation of the securities’ laws?

I’m presuming that Paperchain and/or Centrifuge has not interacted with any regulatory agency regarding TinLake and/or DROP, correct?

The Tinlake organizational graphic and application response to question 9 suggest that both an operating memorandum and loan agreement will be executed by the Lender, presumably MakerDAO. Please correct me if I am wrong.

9. Where does exchange for the asset occur?

The SPV enters into a subscription agreement with lenders who are receiving DROP from the SPV in turn for providing DAI. The DROP token can be redeemed against the cash flows of the underlying collateral directly from the SPV by any DROP holder. This is ensured by the Tinlake smart contracts and the primary way for interacting with these tokens.

There are a number of questions I placed in the Consulfreight application, but I will include them here for completeness’s sake:

This graphic is a bit confusing, so let’s clarify: the lender is MakerDAO; the SPV is the entity Paperchain(?) spins up; the Asset Originator is Paperchain; and the final borrower is the company that creates the invoices, right?

Where does the SPV’s parent sit? With this graph it looks like Centrifuge is the parent. That can’t be right, can it?

I’ll repeat the same question How are you going to have a loan agreement and an operating memorandum with MakerDAO – who can sign on behalf of “MakerDAO”? How can it be enforced by or against MakerDAO? Have you guys thought about these questions – they seem basic. And if there is no contract between the Lender – MakerDAO – and the other entities, how does your proposed arrangement work?

How dependent is the structure on off-chain legal agreements?

Back to the SPV, how will the system you describe ensure the preservation of SPV assets, and thereby value to DROP holders, i.e., by contract, bylaws, etc.?

Does your team have any suggestions on how auctions should be differently structured to handle this security collateral? I’ve reviewed the MCD documentation and it appears that the current auction system is not geared for this white listing feature. So, even if the community onboarded Paperchain, the system as currently designed cannot support its function without potentially violating securities law (admitting that DROPs are clearly debt securities). That seems like a real problem and one that a project hoping to be onboarded should have a possible solution for.

Ok, so that was incredibly redundant, save for one or two spots. As always, this may open a pandora’s box of comments and then questions and then comments and then questions, and so on. That said, I can’t be the only MKR holder with these comments and questions rattling around in my head. In the end, I hope these questions garner the information that MKR holders require to fully assess Paperchain DROP, and include it as collateral in MakerDAO.

Let me make two more points, however. First, I am not one of the dogmatic “only Eth and nothing more!” maxis. Ultimately, from my view, the DAO needs multiple collateral types, including securities and more exotic financial instruments. But it doesn’t strike me as wise to consider regulated products ripe for onboarding yet, given the system’s current structure.

Second, I also note Centrifuge’s responses in the Consulfreight post stating that the burden for any legal analysis related to these projects and onboarding sits with the “DAO” and “domain teams.” That might be the long-term case but Centrifuge has to realize that this process is still quite immature, and there are no legal domain teams able to do this analysis so arm-chair non-lawyers with time on their hands get the chance to ask questions. That’s not scalable though, so perhaps a more appropriate question is geared to Long for Wisdom and others – when will the push come for legal domain teams? If the DAO wants to onboard new collateral types and protect itself from predictable problems (imminent regulatory threats, etc.), a competent legal team should be an area of immediate focus. Or perhaps a big whale or group of whales can sell some MKR and pay a young, ambitious lawyer’s salary for a year to help the community through this next phase.



Collateral applications have so far come in two forms. The first is the tried and more or less tested type that has been in the markets for years, think ETH and the leading ERC20 tokens as examples.

The second type are the tokens that have yet to gain any significant market, or have yet to be launched, but where Maker have been involved on a bizdev level to test concepts. DROP being one such example I remember off the top of my head , but there could be more. In these cases the token in question could, if approved, be launched as a collateral-ready token. While potentially being more risky, the payoff in form of new markets, bridging to non-crypto industries or public perception could make the effort well worth it. The downside is that the applications could be let’s say vague.

You keep asking Paperchain to add color (adding information) to their applications. This additional information might simply not exist yet. Paperchain are for example not able to answer where, how or to whom DROP will be sold so it it safe to say the product still slightly conceptual.

Also, you keep asking questions regarding the legal and regulatory aspects of the DROP token. This is all very well but keep in mind that the legal aspects of crypto in general are still unchartered. Paperchain could spent 10 USD or 10 million USD on legal advice but the answer would still be “We don’t know”.


Imo the show stopping issue with both centrifuge applications, as well as DMM, is that we have no legal recourse or ability to collect if the underlying assets have an issue. This is no longer collateral based lending (the token is not collateral in a traditional sense, its unsecured debt). So while I’m super excited about centrifuge and believe DROP could be a great asset for MCD long term, I don’t think we’re nearly ready to accept it.

If the centrifuge team has any suggestions for how Maker could protect its interests in case of default (by underlying borrowers, paperchain, or centrifuge itself) this may be helpful to move this forward.


Present DROP design:

DROP is sold by Paperchain through a subscription model, not on the market
-> not on the market, no price
-> no price, no chance of Vault liquidation
-> no Vault liquedation, works 100% until business for some reason non-performs
-> on non-performance Maker takes 100% of the hit. Partial compensation possible through legal action.

Needless to say this is highly suboptimal for Maker and due to risk involved the solution will not scale.

Alternative design:
Maker and Paperchain agrees on a MKR staking contract.
-> Paperchain stakes MKR.
-> Paperchain sells DROP however they want.
-> DROP can be used as Vault collateral up to the limit set by the staking contract.
-> upon DROP non-performance the MKR in the staking contract is sold to cover for the Vault.
-> DROP transferred back to Paperchain.

Exact parameters for the staking contract (time lags, trigger rules, warning rules, collateral ratio, stake buildup etc etc) can be worked out by Maker and Paperchain.

Advantages of the staking model:

  1. Keeps Maker out of any need for legal action upon DROP non-performance.
  2. Largely eliminates legal issues. Only MKR is sold, not DROP.
  3. Scaleable. Paperchain needs to expand business? Just stake more MKR.
  4. Smart contracts can be used as a model for any other accounts receivable financing business.
  5. Allows the Maker system to always be overcollateralized.

The downside for this is that Paperchain barrier to entry increases as they will have to stake MKR. I suggest that initially some Maker fund stakes the MKR and then Paperchain takes over if the business gains traction.

@spin and everybody else - what do you think?


Planet_X, thanks for your proposal. These assets already come with one layer of insurance that acts quite similar to what you are proposing: The TIN token holders invest along with DROP token holders guaranteeing that up to a certain amount of losses it would not result in a default for losses. You can read about how the two tranches work in our documentation. This effectively insures the collateral for default up to the percentage that TIN investors invest (Paperchain will launch with a 10% TIN investment ratio).

I would tend to disagree. Centralized assets and decentralized assets both have risks that can lead to a complete default on the asset class. The stability fee that MCD charges (minus the DSR) needs to cover this risk. If there was no risk of default at all then why would there need to be stability fee charged at all?

On Black Thurday we’ve seen that even ETH can backed Vaults can have losses that MKR holders have to carry. A complete collapse of a crypto asset like BAT, REP etc. could lead to such a big crash that no keeper would want to bid on these assets or an undiscovered vulnerability in TBTC could lead to someone taking all of the BTC out of escrow and running away with it causing a loss for MKR holders.

The point I’m making is: both real world assets as well as “pure crypto assets” have an inherent default risk for MCD and if we can price this appropriately then the interest paid by the Vault holders will make up for a potential default.


In most other assets, effective liquidation (and penalty) can help defray the impact of risk on the stability fee. Risk is ideally managed through the collateralization ratio, the liquidation penalty and the stability fee.

Is it really not possible for the system to liquidate DROP tokens? If this is the case then the entire risk needs to be balanced by the stability fee and collateralization ratio, which would be higher than they otherwise would need to be than if there was a liquidation option.

On a separate note, can you provide some clarity on how DROP would behave in the event the stability fee on the Vault was raised above the projected income from the underlying loans? My assumption here was that DROP would just be removed from the Vault (because it would no longer be economical to borrow against it, however in the last collateral meeting, you implied it would result in liquidation, and I’m not sure I understood that.

1 Like

Is it really not possible for the system to liquidate DROP tokens? If this is the case then the entire risk needs to be balanced by the stability fee and collateralization ratio, which would be higher than they otherwise would need to be than if there was a liquidation option.

I was criticizing @Planet_X’ “Alternative Design” without actually correcting this in the their description of the current design. So maybe let me do correct that:

It is very much possible to liquidate DROP tokens of course. We are talking to keepers who are be interested in participating auctions and there is a market for DROP token holders. Taking ConsolFreight as an example, there are 8 parties that bought DAI 250k worth of CF-DROP a few weeks ago because they are interested in directly investing in these assets.

@Planet_X suggested that because there is not a liquid market there was no way to price an asset. I think that is what led to the confusion. You can value a portfolio of loans and give it a price, this can be used to derive a DROP token price. This is similar to getting your house appraised to refinance your loan, the bank won’t ask you to actually sell it to discover the price of it.

1 Like

Ahh, I misunderstood, my apologies.

Do you have any input on my second point?

Counterargument to this:

Until recently, I worked for a consumer lender which obtains funding for loan origination by issuing tranched securitizations (structurally very similar to DROP and TIN tokens). The securities are traded over the counter, so they have some liquidity but not really a fully liquid asset like government bonds, stocks, or mid/large-cap cryptos.

During the market selloff in March, the class C notes (first to experience losses from defaults - similar to TIN) were selling for 20% of face value or less, and even the class A notes (last to experience losses) were selling at only 80% of face value. New issuance became impossible in these conditions , and a lot of existing holders who had levered up their exposure were facing margin calls which further impacted pricing.

Illiquidity makes it much more difficult to value assets even in stable markets. During times of market stress, there’s likely to be more pricing volatility for illiquid assets.

Also worth noting all of the securitizations were rated by credit rating agencies (Kroll, Moodys, etc).


I guess interesting would be at what price the A-shares are selling at now? While the market crashed everything went down but if you held these notes to maturity you would likely get more than 80% of face value.

Your example focuses on consumer lending. With unemployment going up the way it is right now of these loans are heavily affected. Paperchain’s counterparty risk is Spotify and Apple. @pcdan just shared their earnings report:

Using structured products doesn’t remove all risks and any this does not negate the need for the risk team to assess these parameters. That is something that the domain team still needs to do.

1 Like

The second type are the tokens that have yet to gain any significant market, or have yet to be launched, but where Maker have been involved on a bizdev level to test concepts. DROP being one such example I remember off the top of my head , but there could be more. In these cases the token in question could, if approved, be launched as a collateral-ready token. While potentially being more risky, the payoff in form of new markets, bridging to non-crypto industries or public perception could make the effort well worth it. The downside is that the applications could be let’s say vague.

These are good thoughts, and I agree with you. There is a payoff for pushing edges, no doubt.

You keep asking Paperchain to add color (adding information) to their applications. This additional information might simply not exist yet. Paperchain are for example not able to answer where, how or to whom DROP will be sold so it it safe to say the product still slightly conceptual.

Also, you keep asking questions regarding the legal and regulatory aspects of the DROP token. This is all very well but keep in mind that the legal aspects of crypto in general are still unchartered. Paperchain could spent 10 USD or 10 million USD on legal advice but the answer would still be “We don’t know”.

I agree that a lot of Defi/crypto happens in an ambiguous or vague regulatory space. No argument there. But DROPs are debt securities, as Lucas acknowledged in responding to my comments in the Consulfreight channel. From reading the public docs, MCD does not have whitelisting in auctions, and the DAO is not a registered broker-dealer. Auctions involving securities, like DROP, require whitelisting and a regulated broker-dealer. If the system does not have those components yet, doesn’t the potential introduction of DROPs as collateral seem premature? And shouldn’t the technology provider and backer of this proposal, Centrifuge, come to the DAO with some framework on how to make DROPs legally palatable?

1 Like

h/t for focusing on the “lack of recourse/ability” to collect if there are problems with the underlying assets backing DROPs. That problem is ubiquitous with all these (Consulfreight/Paperchain, etc.) applications to date.

@Planet X, very creative thinking on the staking model. Perhaps that is a new avenue to explore for the DROP auctions? Beyond me to say one way or the other, yet it’s something to analyze.

1 Like

Is there any fundamental reason why there couldn’t be a liquid secondary market for these tokens? I think this is absolutely necessary for us to use these token types as collateral. We need some way for price discovery.

This is my interpretation of the MIPS process: MIPS6 is not the final decision whether to add some collateral to MCD or not. This is the first step and a starting point for the domain teams to work on integrating them. There are assets that are simpler to integrate into MCD and one could make the case for integrating them first but the downside is that a lot of these assets (LEND, KNC, MANA, and really any other top 20 ERC20s) are so correlated with ETH and see so little usage (cf BAT’s DAI 580k debt) that these assets will hardly add the diversity in collateral and borrow demand necessary to scale Maker.

Bringing real world assets into Maker will require adapting the process the DAO goes through to onboard new assets. We’re happy to support in any way we can.

I’ve outlined how the DAO could decide to treat these tokens in the ConsolFreight thread:

Our current assumption is the following: DROP is considered a security but only if held by a legal entity. A DAO or smart contract is not a legal entity and therefore as long as theser tokens are held in a Vault controlled by the SPV, the SPV is not actually issuing a security. Only when the Vault goes into liquidation and keepers want to bid on it will the SPV issue the security to the highest bidder. Therefore while keepers need to get KYCed, not every single DAI holder or entity interacting with MCD has to.

This is the basis that we propose for MakerDAO to pursue and investigate.

We plan to elaborate further on that and develop guidelines for how to safely onboard securities into MCD with the legal domain team. Procuring a legal opinion as to how exactly this will work for MCD at this point in time would be premature and something that would likely be a significant expense. This is something we should tackle after the first set of polls that merely determines the ranking of different asset types.

The security status of these tokens means that while they are transferable (and redeemable with the SPV at any time) they are restricted: you can’t sell them to anyone on the public market. This means that the market for these tokens while available will be a lot less liquid than likely required for an efficient price discovery.

This is not something that is new to the credit world however: Credit funds frequently securitize and sell credit portfolios without there ever being a liquid market for them but instead by agreeing on price model that looks at the underlying loans and calculates a NAV (net asset value) based on a risk model agreed upon by the buyer and seller. The Tinlake contracts calculate the NAV based on the underlying loan data on chain to determine at what price to issue & redeem DROP tokens when investors join or leave a pool.

Not unlike other risk parameters the risk team needs to propose for MCD, we imagine that they would review the NAV model chosen by the asset originator, perhaps contribute to setting some of the parameters and use that as a basis for pricing the collateral in MCD. We’ve published a first article in a series on pricing here: Tinlake: Pricing and Valuation Series Part 1

We are hosting another collateral onboarding call tomorrow (Wednesday) at 19:30 CET // 10:30am PST // 17:30 UTC (Collateral Onboarding Call (MIPs #6-12): Discussion and Review of Collateral Onboarding Process). Feel free to join and discuss these topics!