[NS-DROP] Collateral Onboarding Risk Evaluation

Legal Disclaimer: This communication is provided for information purposes only. This communication has been prepared based upon information, including market prices, data and other information, from sources believed to be reliable, but such information has not independently been verified and this communication makes no representations about the enduring accuracy of the information or its appropriateness for a given situation. This content is provided for informational purposes only, and should not be relied upon as legal, business, investment, financial or tax advice. You should consult your own advisers as to those matters. References to any digital assets and the use of finance-related terminology are for illustrative purposes only , and do not constitute any recommendation for any action or an offer to provide investment, financial or other advisory services. This content is not directed at nor intended for use by the MakerDAO community (“MakerDAO”), and may not under any circumstances be relied upon when making a decision to purchase any other digital asset referenced herein. The digital assets referenced herein currently face an uncertain regulatory landscape in not only the United States but also in many foreign jurisdictions, including but not limited to the UK, European Union, Singapore, Korea, Japan and China. The legal and regulatory risks inherent in referenced digital assets are not the subject of this content. For guidance regarding the possibility of said risks, one should consult with his or her own appropriate legal and/or regulatory counsel. Charts and graphs provided within are for informational purposes solely and should not be relied upon when making any decision. The content speaks only as of the date indicated. Any projections, estimates, forecasts, targets, prospects, and/or opinions expressed in these materials are subject to change without notice and may differ or be contrary to opinions expressed by others.

Outline

  1. Proposed parameters
  2. Background
  3. Industry analysis
  4. Asset originator analysis
  5. Issuance platform analysis
  6. Implementation details
  7. Proposed covenants

Proposed parameters

Debt Ceiling: 5M DAI
Stability Fee: 3.5%
Collateralization Ratio (vault-based): 105%
Minimum Collateralization Ratio (pool-based): 108%
Underlying Collateralization Ratio (pool/DAI loan): 113%

Background

New Silver is an online real estate lender with a primary focus on “fix-and-flip” real estate, a large but fragmented market in the US valued at $60-100 billion annually. Most of their loans are for single family residence properties with terms of 12-24 months. Together with Centrifuge, they are proposing to use Tinlake pools to deposit into a Maker vault and generate DAI as a credit facility.

References:

Industry analysis

“Fix-and-flip” real estate is a large and growing industry in the US, accounting for roughly 7% of all real estate transactions. The strategy is to buy a property and improve it with high-value renovations in order to sell it at a substantially higher price in a relatively short time. As with regular mortgage lending, loans are backed by the property and the borrower must invest its own capital alongside the lender. New Silver focuses on companies that pursue fix-and-flip on a professional basis and non-owner-occupied properties, operating in an area known as “hard money” lending for real estate investors.

Traditional lenders have used FHA-backed construction loans in the consumer market, and new online services typically focus on retail consumer mortgages, such as Rocket Mortgage. Some fix-and-flip competitors include: Lending Home, RCN Capital and Lima One Capital.

LendingHome (https://www.lendinghome.com) is an online version of a “hard money” lender, offering bridge loans and longer-term rental loans. Bridge loans are often referred to as “hard money” due to their short-term nature, and generally come with a higher interest rate. Bridge loans also can be offered at a higher Loan-to-Value (LTV), sometimes up to 90%, and sometimes alongside other loans, and can take the form of rehabilitation loans. Like New SIlver, LendingHome performs rapid pre-approvals online. Founded in 2013, LendingHome has funded over $5 billion for over 23,000 projects, and provides statistics about the industry in 11 of the states where they operate (out of 28 states).

RCN Capital (https://rcncapital.com/) offers fix-and-flip, rental and bridge loans, and bridge loans can apply to multi-family or commercial properties. Most of its loans seem to be short-term, running for 12 months to 24 months, though rental loans can have 30-year terms, like traditional mortgages. Established in 2010, their scope is nationwide and in most US states. They also work with real estate brokers.

Lima One Capital (https://limaone.com) offers loans for a full range of investment property financing needs, including fix-and-flip, rental, multi-family and new construction. Its website includes a link to the public record of its mortgage banking licenses in various states. Interestingly, if the borrower owns another property, it can be used as collateral in lieu of a down payment. Lima One also works with real estate brokers, offering preferential service to the most active brokers.

New Silver uses advanced techniques for automated lending approval and deals primarily with businesses. Then a traditional process of property appraisal and loan servicing is used to issue and manage loans to maturity.

New Silver operates in a small number of US states (currently 6) where they are legally exempt from mortgage lending regulation (39 US states offer exemptions from mortgage license requirements). They utilize credit scores of company directors and Attom and Clear Capital for automated valuation models as part of the approval process.

The relatively short term of New Silver’s loans limits risk substantially, as changes in the real estate market generally take place over a number of years. Nevertheless, black swan events can occur and the recent one in 2008-09 provides a useful worst-case scenario. As Wall Street leveraged up subprime mortgages based on an obvious bubble caused by 10 years of money printing, the bubble burst and shoddy lending practices resulted in widespread foreclosures, collapsing real estate markets in inflated areas. The worst-hit markets dropped up to 40% over a period of 18 months to two years, and lending freezes made it difficult for buyers to step in.

Such events are exceedingly rare, but it provides a kind of stress test for downward price movements in real estate, which typically move up gradually over many years. This event demonstrates a price floor of -40% in some parts of the US. However, it is much more likely that a particular property will experience some problems, in which case that risk can be monitored and mitigated, as will be discussed below.

The process of recovering from loan defaults is well defined, and most options do not involve foreclosure, where the primary lender pursues a sometimes lengthy legal process for repossessing a property. Default rates were elevated during the subprime crisis, but the US market has recovered and for fix-and-flip or hard money loans, default rates may have been as low as 1% in 2008-09 or as much as 2% in 2013, according to this article. The current real estate market during the pandemic has been strong in non-urban areas, and some data backs that up. There are also negative economic impacts in some urban areas as people move away from them, and the economic fallout is still playing out.

New Silver estimates that only 5% of defaults end up in foreclosure, indicating a diversity of resolutions are possible. This diagram illustrates the default process:

Asset originator analysis

Founded in 2018, New Silver Lending LLC is a relative newcomer to the hard money lending industry in the US. It takes a tech-heavy approach to streamlining the loan underwriting process, using a variety of data sources and metrics to assess creditworthiness and suitability of a borrower and property. After providing a pre-approval within minutes, New Silver collects a variety of due diligence information prior to issuing the final approval, contingent upon an official property assessment by a third party. Credit scores of the individuals in a company are averaged together.

As mentioned above, New Silver currently operates in 6 US states - Massachusetts, Connecticut, Florida, Illinois, Texas and New York, though it aims to expand into a total of 39 states. Its loan portfolio consists mostly of purchase + renovation loans with short-term maturities of 12-24 months. The small proportion of long-term loans will not be included in the Tinlake pool and Maker credit line. New Silver is the only lender for a property and keeps the senior debt position, which is transferred to the SPV and indirectly to Maker via the mechanism described in “Issuance platform analysis,” below.

The management team consists of:

Alex Shvayetsky, Managing Partner: 25 years in real estate investment and management and property development, including 22 years as CFO of Udolf Enterprises, a large commercial property operator based in Connecticut.

Kirill Bensonoff, Managing Partner: an experienced entrepreneur in enterprise IT, starting his career at IBM Global Solutions and an active investor and advisor at over 30 startups and also active in private lending, driving the vision to enable a new user experience in private lending while using modern software to scale.

Alexey Shevchenko, CTO: full-stack developer with years of experience in real estate, blockchain and fintech, previously CTO of OpenLTV, which enables fractional investing in high-quality short-term debt backed by real estate with yields of 8 to 12%.

From 2019 to the present, New Silver provided 134 loans at a notional value of $34,510,864. In 2019, all loans were sold to a range of partners and serviced by New Silver and a third-party servicer. The average loan size is $195,000 with an average Loan To Value of 71%. The interest rate charged to borrowers is around 12% and varies up or down based on loan amount/LTV and creditworthiness. The goal going forward is to refinance those loans via Maker.

New Silver has never experienced a default by any of its borrowers. The default process is shown above, and there are numerous remedies that can be pursued before the estimated 5% of cases in which a foreclosure is required. Default can be triggered by delinquency of a monthly interest-only payment or the “balloon” payment at the end (which is the loan amount). In about 80% of cases, payments resume, and then in 80% of the remaining 20% of cases a work-out is implemented.

Foreclosure is the last resort and very undesirable to borrowers because of its impact on their credit standing, and is also more time-consuming than the other remedies. Work-outs are usually document-based, such as assuming a controlling role in the borrower’s LLC or a deed-in-lieu transfer, which can take from a few weeks to a few months - on the shorter end of the time scale if a loan is sold and on the longer end if the property is sold. Foreclosure can take 6 months or longer if it goes through the state court system, but it can also proceed directly as a distress sale, with a potential recovery of 50-90% of the market value.

It should be noted that borrowers will be very motivated to maintain good standing if they have a significant amount of equity in the property, and especially if their own funds are at stake. This kind of loan is backed by real estate as collateral, after all. This is the reason that Loan-to-Value or Loan-to-Cost percentages are kept within reasonable levels, usually 70-80%.

The automated pre-approval process includes a variety of checks that reduce risk and the cost of underwriting. These checks include:

  • FICO score of the borrower
  • Borrower experience (low: 0-1, medium: 2-3, high: 4+)
  • Automated Property Valuation
  • Average days on market in Zip Code
  • Standard deviation from median sales price in zip code
  • FHFA HPI Maximum Yearly Decrease 2006-2018 in County and State vs National
  • ZHVI (Zillow Home Value Index)
  • Median sale price for zip code to ARV %
  • Monthly sales count by zip code
  • Census data - average household income, town population

Before issuing final loan approval, additional information is collected, including:

  • Appraised property value
  • Borrower liquidity
  • Borrower and entity background
  • Entity Good Standing
  • Corporate docs
  • Verification of experience
  • OFAC
  • Flood zone
  • Insurance
  • If its a condo, HOA and master policy

While New Silver assesses these metrics at the time of loan origination, Maker will assess and monitor risk via an asset valuation model that incorporates the same valuation metrics and others, which will be described in detail below.

For most projects there is a second construction loan, on top of the primary loan for acquiring or refinancing the property. The construction loan is based on the initial assessment of After Repair Value, and is kept within a maximum level of 85% of ARV (usually less). As the construction proceeds, the borrower puts up their own cash at each step and then requests an inspection. A third party makes the inspection and reports if the proposed work was done, then that portion of the construction loan is sent to the borrower. In this manner, the construction process is closely monitored, and never exceeds the agreed level of ARV.

Fix-and-flip is growing and services like New Silver and its competitors are expanding the range of financing options for property investors. Even during the pandemic, demand for properties in some areas is growing and by selectively choosing the markets experiencing the most growth, while avoiding the markets that experienced the most declines in the bursting housing bubble of 2009, New Silver can use the Maker credit facility to grow its business and increase DAI generation.

Issuance platform analysis

New SIlver has partnered with Centrifuge, a tokenization platform focused on loans as assets. Unlike tokenization platforms designed to turn assets like real estate into fungible or non-fungible tokens, Centrifuge creates a bridge between lending in the real world and borrowing DAI. The asset pool it creates is called Tinlake and it uses two tokens to establish senior and junior debt positions, DROP and TIN.

The SPV can issue some DROP to be deposited in a Maker vault to borrow DAI, while TIN and other DROP are offered to external investors or purchased by the Asset Originator itself. The TIN tokens fluctuate based on a NAV (Net Asset Value) model and provide a safety buffer for DROP holders by taking portfolio losses first.

DROP holders have the equivalent of senior position in traditional collateral-based loans, though the main difference is that a lender with senior position can take possession of the underlying collateral in the case of default, whereas in the Centrifuge model, TIN holders create a buffer and the SPV that holds the Tinlake pool can exercise its right to repossess collateral on behalf of the DROP holders.

For Asset Originators using the Centrifuge platform, the expectation is that they must deal with defaults in their portfolio in the normal ways, as discussed above, and the Maker liquidation model must account for this, as discussed below.

Centrifuge is a startup based in Germany that has worked closely with Maker in POCs and community proposals and is extending its platform to integrate with Maker vaults and smart contracts. The Centrifuge platform was launched on May 26, 2020 and the company was founded in 2017.

Implementation details

The Centrifuge model is used in this case and several aspects of the model are worth mentioning here.

First, the issuance of DROP and TIN tokens is handled via Securitize, with AML/KYC procedures and compliance with US securities guidelines. Investors based in the US must be accredited investors (generally defined as having a net worth of at least $1 million). The SPV issues DROP tokens for the Maker vault, and Centrifuge will post legal discussion about potential implications for Maker.

Second, MIP22 provides a liquidation mechanism for the Maker vault, and the liquidation process will depend on the type of underlying asset. In general, the liquidation process is triggered by the TIN value falling below the minimum collateralization ratio set for the Asset Originator. There is no auction or keeper function to liquidate the vault, but Centrifuge can “lock” incoming funds to redeem DROP tokens and prevent further loans.

The collateralization model is illustrated here:

The risk mitigation process looks like this:

Centrifuge has its own mechanism to protect the safety of DROP, and that level has been set at 118% of the DROP level, i.e., TIN is 15% of the pool. When the Centrifuge NAV model detects TIN value moving below that level, the Tinlake pool will freeze the ability to invest in DROP and allow the Asset Originator to increase TIN investment or adjust its portfolio. If the TIN level gets above 15%, the pool reverts to normal.

The Maker liquidation mechanism is detailed in MIP22 and is an extension of the Centrifuge platform. The Maker liquidation results in removal of all DROP in the Maker vault, and funds coming into the Tinlake pool will redeem DROP. This is a drastic action, and will likely take time to complete. Therefore, it makes sense to monitor the pool and underlying portfolio and be proactive with the Asset Originator at the earliest warning sign. In addition, a violation of any covenant below can trigger risk mitigation.

Before liquidation is triggered, the Risk team will make an assessment and may propose an executive vote to decide on reducing the debt ceiling or activating liquidation. For New Silver, it makes sense to allow sufficient time to mitigate any issue in its portfolio. This time period could be one month to several months, depending on further discussion with New Silver and Centrifuge.

The standard NAV model in Centrifuge, and the NAV calculation framework, is modeled on invoice factoring, and over time we will suggest some changes for other asset types like real estate. For example, monthly payments of interest only carry some weight in consideration of default, rather than only lateness of the balloon payment at the end of a 24-month term. Invoice factoring expects one payment at the end of a very short term. Regardless, lateness in monthly payments will be monitored via the asset valuation model described below, and Maker will have its own safety buffers, as described below.

Proposed covenants

Allowed investments

The allowed investments by the SPV are real estate investment loans of 12- to 24-month duration, whether including a renovation component or not, and within maximum LTV (Loan to Value) or LTC (Loan to Cost) levels of 80% and 85% respectively.

Minimum collateralization ratio

The Minimum Collateralization Ratio is set as a floor by Maker to protect the collateral in the Maker vault. This is based on the Tinlake pool but below the TIN-adjusted Collateralization Ratio. The value of a potential default resulting in 50% loss of one loan (worst case of distressed sale) is $95K on average. With a vault size of $4 mill, the TIN level could decrease from 118% to 113% if two such defaults occurred. To give room for four worst-case defaults, the Minimum Collateralization Ratio is set at 108% of the overall DROP level.

We won’t provide 1:1 DAI for nominal DROP value. The initial parameter is to allow a vault collateral ratio as low as 105%. In other words, a loan of 100 DAI can be generated from 105 DROP value. See “Collateralization ratio proposed threshold,” below.

Collateralization ratio

The Risk team has created a tracking sheet that models the value of the underlying property portfolio, and calculates the loan amounts as the assets in a Tinlake pool, basically as a check on the Tinlake pool. Due to NDA, it is not possible to share this document. The model will evolve over time, but it starts with a total of the loan amounts and remaining pool value compared to the TIN amount shown in the pool, and monitoring of equity value in the underlying portfolio.

New Silver will provide a monthly update of their loan portfolio. This gives the Risk team a view of the underlying properties and any anomalies that may require discussion with New Silver. Over time, this view will be supplemented with external data sources that validate property values. In addition, New Silver has agreed to provide API access to their data so a Maker dashboard can access not only the loan tape but also underwriting criteria.

In general, this view of the underlying properties is an early-warning signal provider and ensures that Maker is confident about a number of potential scenarios:

  • Increasing the debt ceiling
  • Increasing the timeframe before liquidation, should a collateralization level be crossed
  • Reducing the Stability Fee

Of course, red flags that are not remedied could lead to actions in the reverse direction.

From the view of the underlying properties, the total loan value is used in a monthly snapshot of the balance sheet, using some values available in the Tinlake pool UI, such as total DROP supply and Pool Reserve. This provides a check on the TIN level as well as a view of the overall pool status.

The debt amount is the one on the Maker vault but also the amount subscribed by other co-investors.

Total DROP supply accounts for all DROP investments, not just the DROP in the Maker vault. We expect other investors to participate in Tinlake pools, either with TIN or DROP. New Silver has proposed to be the primary TIN investor initially. Other DROP investors can’t deposit into the Maker vault, but they have similar rights as senior debt holders. See “Stakeholder rules” below.

This part also contains the data that the SPV should provide on an ongoing basis to the risk team.

As mentioned above, New SIlver has agreed to provide monthly snapshots of its loan tape, and eventually API access to its data.

Collateralization ratio proposed threshold

New Silver has proposed a TIN proportion of 15% of the overall pool. That means a DROP proportion of 85%, and therefore a pool/DROP ratio of 118% (100%/85%), referred to as the TIN-adjusted collateralization ratio. This is based on the Tinlake pool itself. From the pool perspective, Maker proposes a minimum underlying collateralization ratio of 108% of all DROP, called the Minimum Collateralization Ratio.

The Maker vault receives DROP from the SPV and is likely to be the largest holder of DROP. The vault imposes its own collateralization ratio of 105%, thereby limiting the DAI that can be withdrawn and in effect creating a small discount for its DROP. This creates a slight buffer in the vault itself, and compensates for a slightly lower interest rate. The Stability Fee of the vault’s DAI loan is 3.5%, whereas DROP receives a 5% yield generally.

In addition to the TIN-adjusted Collateralization Ratio and the Minimum Collateralization Ratio, we also propose a pool/DAI loan or Underlying Collateralization Ratio of 113%. Even though there will be other DROP investors, it is clear that overcollateralization in the vault will have in addition at least 8% overcollateralization in the pool based on the MCR.

From an overall pool perspective, Maker gets four buffers of safety:

  • TIN-adjusted Collateralization Ratio, triggering Tinlake pool actions
  • Minimum Collateralization Ratio including TIN as well, triggering Maker actions
  • DROP co-investors, who share our interests
  • Vault collateralization ratio, for the Maker portion of DROP

In addition, another level of safety exists in the properties and New Silver’s borrowers. The loans represent only a portion of the overall property values, and the borrowers have their own equity at risk. Assuming the worst case of 85% ARV or LTC, there is a 15% buffer outside of the loans. Calculating this (1 / 0.85 x 1 / 0.85 x 1.05 with no DROP co-investors) reveals a “real-world” collateralization ratio of 145.33% (usually higher).

The initial Debt Ceiling is proposed at $5 mill, and New Silver estimates it could reach that level within two months of opening the Maker vault.

Concentration risks

The maximum loan amount for a single loan is the maximum of 10% of the pool or $500k.

The maximum exposure to a borrower is 25% (with a pool size of $5 mill or $1,500k) if FICO score is above 650, and 10% or $500k if FICO is below 650.

More important are geographic diversity and choice of markets where the chance of severe price drops and defaults is relatively low. Maximum loan concentration is 30% per state. New Silver must focus on areas with price appreciation, jobs growth and a robust sales environment.

Regarding the current situation with COVID-19 and the potential impact on default rates in some urban areas, New Silver has focused on suburban areas where demand is growing and prices are rising. While it is true that rising defaults in cities could impact markets generally, that also creates opportunity for its borrowers in the form of additional inventory to fix and flip. An additional covenant is to manage the location of properties to be in areas of price and demand growth.

Stakeholders rules

The Risk team has assessed that Centrifuge is financially sound and most likely will be in business in 12 months. This impression is based on an analysis of bank statements and internal forecast models from the company.

The Risk team has assessed that New Silver is financially sound and most likely will be in business in 12 months. This impression is based on an analysis of recent financial statements.

Details of the Centrifuge model and investors in a Tinlake pool were mentioned above. TIN investors are providing equity in the pool, and therefore it makes sense for the Asset Originator to be a primary investor in TIN. In addition, it could make sense to require a minimum level of DROP co-investors, as they could assist in enforcing contract rights, benefitting Maker.

New Silver should always keep an exposure of $40k in the TIN tranche of the pool.

The MakerDAO vault should not have more than 75% of the outstanding DROP.

11 Likes

Itshappening.gif

1 Like

Thank you for the evaluation @Philinje — just a question — how did you come with the SF of 3.5%?

I mean—hard money lenders usually average 11.5%—just trying to figure out why 3.5%?

2 Likes

For all RWA (currently at least) we have non-investment grade loan (which imply high interest rate).

The model is taken from RMBS and other ABS (Asset-Backed Security). You take a set of low quality loans, package them together and find someone to take the first losses. You end up quickly with something of investment grade quality. That’s because most of the risk is on the junior tranches.

You can see below the decomposition of the tranches of an ABS (which was just before the subprime so way too aggressive). AAA is super safe (like lending to the US government), A is a good industrial company, BBB is medium, BB and B are low quality.

The individual loans would be rated B probably (but obviously those are too small to be rated at all). We don’t have a model yet to define every threshold (that would take a team of actuaries).

My current reasoning is that we should aim at a BBB risk adjusted level (this is the kind of credit risk you have when lending to General Electric, AT&T or General Motors). There is already inherent risks by doing something new and on the blockchain so targeting a credit risk of A or more would be false safety.

Another issue we face is that yields are falling hard in the real world with BBB yield at 2.16%. BB are at 3.59%.

4 Likes

@ElProgreso, I think you are mixing financing cost that the borrowers pay with the interest rate the senior tranche gets. Financing cost include origination fees, servicing fees, a spread for the originator.

The interest rate that Maker is setting is on a portfolio of loans where the borrowers all put in 15% equity at least. Those are then bundled into a single security against which junior investors put up another 15% of equity. On top of that Maker is then asking for a 105% collateralization ratio. This means all in all you have an equity ratio of 45%.

When you look at senior bank line of credits, those are actually competitive to what Maker offered, actually sometimes even less.

2 Likes

Cool–this should open a gigantic window for lenders to use the Maker facility–the Rates are just :fireworks: – Looking forward to more RWA MIP-6 applications.

Now, if we can only get RWAs going… In the words of Mick Jagger:

“If you start me up
If you start me up, I’ll never stop”

Indeed, these are the first baby steps toward oceans of opportunity. And let’s walk before we run!

Just to add a few thoughts to other replies:

In lieu of a major change in the Centrifuge platform, we have taken a preferred DROP position by taking a slight discount in the cost of DROP and having a bit of a buffer in the vault. This was compensated for by a slightly lower interest rate, ie, Stability Fee. And as Seb mentioned, our position is in the same zone as BBB debt. Plus we have the ability to invoke MIP22 liquidation.

3 Likes

Not that I am complaining but 3.5% seems extremely low cost of borrowing considering the inherent risk with smart contracts. I think that real world interest rates will continue to plunge most likely going to zero adjusted for inflation. I don’t think that it would be wise to try to match such artificially suppressed interest rates. Close to 30 trillion dollars of sovereign debt has negative real yields which we cannot possibly compete with. Maybe there should be an interest rate floor (e.g. 4%) similar to the LIBOR floor on CLOs because at a certain point it just doesn’t make sense to take on the lending risk for an abysmally small reward.

1 Like

Regarding the rates, we are more than good regarding to the credit risk (obviously we will monitor that closely). I was a bit concerned regarding Covid and real estate. But data is good and price are increasing in the NewSilver markets. You never know what will happen.

Compared to CrescoFin, we don’t have insurance but are taking a senior tranche. We are targeting a BBB risk which I think is good. I think we don’t want too much risk currently as we don’t have the solvency tools in place to sustain losses.

On the smart contract risk, and I would say more broadly the operational risk, it’s hard to price that exactly and remains competitive. That’s why I don’t think we should go in higher credit quality because the decrease in credit risk doesn’t compensate the still present operational risk.

We can target pocket of good yield (mainly stuff that no big bank really care because it doesn’t scale and too complex to manage on their side). But after that, you are right, I have no idea what we can do. We can hope that the yield curve will improve when our size will be too big. 3% is, for me, the limit on the low side. Governance can signal they want more or less.

Happy to discuss if you have a source of CLO with a 4% floor and good credit rating (and I would add short maturity).

We might also want to invest in long maturity to avoid, as you say, the fall in interest rates (that I think will happen). But we have no idea how to manage interest rate risk.

Are you suggesting insurance is the solution? It might :slight_smile:

2 Likes

@SebVentures If CrescoFin is paying depositors a risk-free rate of return of 3%, then that means we are only charging a 50 bps spread for a lot more effort and going much further down the risk spectrum. This is far from optimal given the effort it will take to make sure RWAs don’t cost MakerDAO any money. When I said we should set a floor I meant that the floor for MakerDAO lending to RWAs that are more or less BBB-rated should be set at 4% which is in my opinion reasonable.

CLO debt tranches price their cost of borrowing at 1-month LIBOR plus a spread. However, they also make sure that no matter what the floor on LIBOR is 100 bps. So there is no borrowing at 0.25 plus a spread because it would be uneconomical. BBB-rated CLOs with a 3-year loan maturity are priced at around LIBOR plus 350 bps.
Screen Shot 2020-12-04 at 5.49.08 PM

This is in line with what I think we should be charging, which is far above what the BBB-rated corporate bonds are charging. If we want to compete with them, then we will be at a marked disadvantage. Instead of trying to import the yield curve that is manipulated by central banks, we should be working to create a DeFi yield curve that is priced by the free market. This is what Robert Sharrat of CrescoFin is trying to accomplish and I think we should be trying to work with him and Aave to possibly create a bigger pie rather than fighting for crumbs which hurts both protocols.

Since he owns 95% of the company, he intends on selling his equity in order to finance operations. I was trying to convince him to work with us to onboard wCRES tokens and use them to open a working capital line of credit. MakerDAO would have a super safe form of collateral (possibly use it as strategic reserves) and he wouldn’t be forced to sell his ownership stake given that he has built it from the ground up and financed it all. Since he has credit insurance experience, he has the infrastructure to invest in listed securities, and a passion for bringing RWAs to DeFi why not work with him rather than against him.

On the one hand, he can focus on onboarding institutional investors (i.e. KYC-compliant) and TradFi assets (e.g. asset-backed securities); on the other, MakerDAO can focus on onboarding DeFi assets and non KYC-compliant investors. It doesn’t have to be a zero sum game, it can be a win-win-win scenario.

3 Likes

I don’t want to derail the thread, but I’m really confused about how CrescoFin is pulling off a 3% risk free rate on what’s effectively a rolling 30-day CD. 30-day Libor is 0.15% so I think something is mispriced. Either the borrower is paying too much, Lloyds is underwriting for too little, or CrescoFin is subsidizing the rate. There’s no free lunch. Do you have more information on the mechanics of the product?

PS - It’s great that they accept Dai for the investments. If you need any help in convincing them to join MakerDAO I will gladly volunteer. Also…sorry to keep editing this… but how did I just buy their security token on balancer without going through KYC?

3 Likes

@ejbarraza the issue with BBB CLO tranches is that there is many tranches on top of it (I assume in your case). With a BBB corporate bond (or NS-DROP) you own the BBB tranche up to the AAA. This, I think, explain the difference in rates. The probability to have a credit even is the same, when on the CLO tranches it’s more often 100% of losses. That’s the difference between losing 1M of a 5M vault and losing the 5M. I don’t think that there is a free lunch.

Speaking of food, I would make the analogy between tranches and the movie The Platform (El Hoyo). A BBB bond is like eating from level 1 to level 60. A BBB tranche is like eating from level 50 to level 60. Not the same experience.

Let’s continue the wCRES discussion here to avoid polluting this one.

@ElProgreso I think the 3.5% has been chosen to match up with what 6s Capital is getting + a bit extra for the Oracle issues. This is not meant as criticism of the RWA risk work, it is just that there is (for now) most likely no way to measure the risk so we just start there and adjust later.

@ejbarraza you are most likely correct that 3.5% is far below what they would pay elsewhere, thank you for pointing that out. There is a measure called “Wall Street Prime Rate” which is what the largest bank charge their prime customers. Currently this rate is 3.25% so both 6s Capital and New Silver will be getting a deal they would otherwise have no chance getting due to lack of size and track record.

As I see it there are really two ways this could develop. Either Maker is a simply a better way of financing or Maker is underestimating the risk involved and will at some point be forced to increase rates due to losses.

1 Like

The 6S Capital SF was decided by Governance in the MIP13c3-SP4. The NS-DROP SF is not linked to that and there is no Oracle issue.

You can check the NS 1 Tinlake pool and see that the rate at which they lend is indeed at the 10% level.

We are lender of capital and not here to gamble by taking a big credit risk. NS-DROP yield is 5% APR (not APY like our vaults) and we are giving a way to refinance those tokens. There is an extra cushion on top of the NS-DROP token. We can be wrong, but we feel quite safe.

Quite sure we can find a way to make a NS-TIN vault that can yield a 20%+ yield. But the risk would be quite big and not serious.

If this experiment is working well, quite sure also that we would be able to ask NS for a custom pool that match our preferences.

@Planet_X I think it’s a third way. Maker has the ability to mint money and is super small. Which big bank would worry working of such work intensive loans? The legal setting would cost more than 20y+ of revenues and the size would not change a thing in their yearly P&L.

3 Likes

Yes but work intensive loans suck bleach. It will never scale. The less manual work the better. Preferably no human involvement, just pure protocol.

1 Like

CrescoFin is not investing in a CD, it is investing in invoices from BBB-rated companies. Robert Sharrat has said that CRES earn 7% and pays Lloyds 1% for insurance and depositors 3%. The depositors are the ones investing in a DeFI equivalent to a 30-day CD. You can buy the wCRES token in the same way that you can buy wBTC or wNXM. However, only the CRES token is entitled to dividends, voting rights, and convertibility to stable-coins. He has said iCRES was designed to be a stable-coin with composability with MakerDAO and other primitives. Even though they seem similar (yield 3%) the difference is that CRES is an equity stake with the potential for capital gains and iCRES is an interest-bearing stable-coin without capital gains.

I am suggesting that anything below 4% without insurance on principal is underestimating the risk. By bringing in the yield curve of TradFi (i.e. LIBOR and Prime) we are trying to compete with an infinite supply of negative real interest rates. They don’t care about credit risk because they can just print their way out of insolvency, MakerDAO cannot.

Actually there are also many tranches beneath which absorb defaults first which serves as a cushion against losses. Also, CLOs lend on an over-collateralized basis to further insulate against credit risk. In order for the BBB to lose any money all of the equity, B- and BB-tranche has to be wiped out. This would be EXTREMELY unlikely since BB-rated tranches have only sustained 1.5% defaults over 2 decades.

These are the kinds of securitizations that I hope to see onboarded to MakerDAO/DeFi. The interest rates on BB-rated CLO debt is in the range of 6.5% with the credit risk of a A-rated corporate bond. The same kinds of structures exist for other assets such as collateralized mortgage obligations, insurance-linked securities (i.e. catastrophe bonds), collateralized bond obligations, credit card asset backed securities, student loan asset backed securities, and so on. Only institutional investors can purchase these assets which is why I want someone like Robert to act as the bridge between TradFi and DeFi. MakerDAO can then be the source of long-term capital.

With a BB tranche you have a 8-11% buffer (in your infographic) and you have to pay all the above tranches first.

In NS-DROP, our loan is similar to the AAA tranche (31% of buffer on the property value vs 35-40% in the infographic). Your BB tranche is more or less the TIN tranche (the equity tranche being the borrower equity).

I made a quick analysis for a base case (1% default, 80% recovery on the property value), mild case (20% default, 60% recovery) and big stress (50% default, 50% recovery).

I don’t model price appreciation from the renovation which is why, borrowers are always losing money.

There is a bit of randomness, but Maker is only sometimes impacted in case of a big stress. The TIN tranche take a serious hit already in the mild case and is fully destroyed in case of a big stress.

The yield on the vault equity is a bit overestimated, but probably something to adjust moving forward.

image

The last line is the yield if you had invested in the overall pool (TIN + DROP), before being shared between TIN, vault equity and Maker Loan (with DROP = vault equity + Maker Loan).

Again, it’s something simplistic, maybe buggy and not fully accurate. But it gives the idea.

I have been benchmarking them against CLOs and that is an apples to oranges comparison. What is the LTV of the underlying property (not ARV)? I hope it is no more than 65%. This would give a 35% buffer assuming the property does not appreciate or actually collapses in value. Do they exclusively invest in non-judicial foreclosure law states? The closest benchmark that I could find is BRMK (LendCo) which does fix and flip loans (BorrowCo) using private investor capital (depositors). They pay a 7% dividend to investors which is far above the 3.5-5% of New Silver.

1 Like

Sorry I wasn’t saying it is a CD, I was saying that it’s like a CD because they’re marketing the returns as “risk free.” My question is how this is possible. Where are they getting these receivables that Lloyds will fully insure for 1% but the borrower will pay them 7%? That’s an arbitrage opportunity, not a calculated risk/reward. There has so be something more to this product that I’m not understanding or is not being advertised.

Also, I purchased their equity on Balancer. Do you know if there’s a catch to this? Generally as a US citizen I cannot purchase securities without KYC/AML and a proper registration from the issuer.

2 Likes

Thanks for this reference to Broadmark. They are kind of a blue chip in the hard money lending space, and I believe the point you are making is that the dividend on owning their stock is 7%, which is compelling. I suppose within the world of equities, they would be compared to REITs, which generally offer a high return, often north of 10% (but taxed as personal income, not dividends). In the risk evaluation, a closer competitor to New Silver is LendingHome, who are of a size similar to Broadmark but with an online “next-gen” approach. To answer your questions, the typical LTV for New Silver is 70%, and the max is 80%. That is similar to guidelines for banks, and it may be hard for a relative newcomer to demand as much equity as Broadmark. Regarding non-judicial foreclosure states, that was discussed at one point with New Silver and I will get a concise answer from them and report back. There are a variety of “work-out” methods before going to foreclosure, such as deed-in-lieu in states that use deeds, and I will summarize that related issue as well.

3 Likes