From my humble point of view, the most interesting piece of financial innovation in circulation today is what a small group of decentralized finance startups is doing: trying to bring real world assets on-chain.
What these companies are doing is fascinating, and yet extremely challenging at the same time. They are trying to migrate trillions of dollars of assets and billions of daily transactions to new financial rails, built on a new technology which incorporates a different set of values. The assumption behind this migration attempt is that the new rails will be simultaneously more efficient but also fairer to the ordinary user.
Among these knights of financial decentralization, trying to convert the financial world to the new paradigm, one of the most fascinating projects is Centrifuge.
What is Centrifuge?
Founded in 2018, Centrifuge is a very familiar name for people in DeFi as it was one of the first established protocols in the space . What Centrifuge wants to do is to “unlock liquidity for real world assets”.
In simple terms, the company is building a P2P lending product that connects lenders (crypto LPs, mainly DeFi investors or protocols) to borrowers (SMEs generating real world assets like invoices, real estate and royalties). In reality, as described later in the post, the product is more complex than that: Centrifuge doesn’t directly link investors and final borrowers, Centrifuge links investors to alternative lenders that will then re-lend to the final SMEs.
The value proposition on both sides is the one typical of P2P alternative lenders: on the investor side, Centrifuge is promising good returns, in this case diversified and non crypto-correlated. On the borrower side, Centrifuge is offering SMEs cheaper funding costs.
How does it work?
Centrifuge’s core product is built in an application running on the Ethereum chain: Tinlake – that is now in the process of being migrated to a dedicated chain.
The product is articulated into a few key components:
- Tokenization layer: assets like invoices or royalties are presented as collateral by the borrower to get a loan. The collaterals are wrapped in an NFT which becomes the physical collateral that protects the lender from future defaults
- Securitization layer: the loan is then securitised and tranched into two tranches: TIN and DROP. DROP is the senior tranche and TIN is the junior one. If you are interested in a more comprehensive analysis of the securitization approach, my friend Luca wrote a beautiful post that I recommend reading.
- Liquidity layer: LPs will fund the liquidity pool and get TIN or DROP in return for their stable coins, namely DAI. Centrifuge has built a direct integration to a MakerDAO vault which acts as a senior tranche investor, but future further integration to other liquidity gateways is in their roadmap
In this process, Centrifuge does not hold any underwriting responsibility nor any legal direct responsibility in the asset origination.
Going into the details: Centrifuge doesn’t really originate any loans directly, its product is ultimately a legal-tech solution, composed of a Special Purpose Vehicle (SPV) linked to a network of technological and commercial integrations.
On the borrower side, Centrifuge doesn’t underwrite or originate any loans, as traditional P2P platforms do. Centrifuge provides a blueprint to establish a SPV but doesn’t establish the SPV itself. The SPV that will perform the securitization is created by the Asset Originator, so Centrifuge’s customer on the borrowing side, that will then re-lend the DAIs to the final borrower (SME). It is the Asset originator who performs the Underwriting of the final borrowers, sets the terms for the pool in which LPs will invest and is responsible for verifying the genuine nature of the collateral deposited.
On the investor side, Centrifuge doesn’t distribute any repayments. DROP and TIN ERC20 are bought by the investor directly from the SPV, after an appropriate KYC is performed.
Once the SPV is set and the loans originated, as in any securitization process, the repayments flow is fully separated from the Asset originator. If the Asset Originator collapses, the SPV is insulated and the financing is not impacted: repayments will flow to the investor through the Tinlake pool.
In case a borrower defaults, Centrifuge has a liquidation mechanism in place. As soon as the default is triggered due to a missing repayment, the origination of new loans from the pool is stopped, redemption of TIN token is frozen and any incoming repayment is directed to DROP holders.
The elephants in the room
Centrifuge is a brilliant technological solution, built with scalability and decentralization in mind. But I’m not sure this is enough to achieve mainstream adoption and become the de facto standard in DeFi lending.
Centrifuge’s success will ultimately depend on one key question: can investors get superior returns and borrowers get lower financing costs? The beautifully architected system presented above is not going to be used at scale if the tool cannot provide economic benefits to its users. And this is not a trivial achievement to reach.
The DeFi lending space is built on very little risk management sophistication and a lot of capital buffers here and there. The vast majority of loans are overcollateralized and this brings a lot of capital inefficiency. As analyzed by my friend Luca in his post, in one of Centrifuge’s securitization – New Silver – MakerDao’s vault receives 3.5% return while borrowers are paying 10-12%. This is far away from the capital efficiency that is reached through banking financing in Traditional Finance, in which together with more sophisticated risk management tools that allow to reduce the capital buffers, there is also a systematic use of leverage.
For as much as mortgage loans can be incredibly low margin for banks, their regulatory treatment allows banks to lever their RoE up to 10x. The risk-adjusted returns for equity holders are incredibly attractive. How could an under-collateralized DeFi lender compete?Luca Prosperi – Dirt Road #10
Centrifuge’s bet is that through automation, disintermediation and organic growth, DeFi will deliver lower financing costs and higher returns to both sets of customers. And that when this happens they would be strategically positioned in the lending infrastructure space to capture the opportunity. Are they?
Centrifuge’s approach so far has been very product-oriented: they have built the most ambitious product in terms of vision, but they haven’t posed the same attention to their commercial proposition, particularly on the borrower side. While similar companies like Goldfinch and Credix have a less comprehensive and sophisticated product, they do have a much clearer go to market strategy, especially on the origination side.
I’m specifically referring to the borrower side, because that is usually the weakest link in P2P lending platforms: it is traditionally harder to find good companies to lend to than investors who want to lend, or at least this was true in the low interest rate regime we used to live in.
Taking Credix as an example, they are building a product to provide a cheaper and more convenient access to markets to emerging markets’ alternative lenders , and all their efforts are going in that direction. Centrifuge is building an agnostic product which could be very promising in the long run, but less so in the short term: my experience in startups tells me that building for everyone, usually means building for no one, and this is a risk that Centrifuge faces if they don’t laser focus on a segment of borrowers to start working with and eventually dominate.
Centrifuge’s team has done a terrific job in building the most sophisticated and comprehensive lending platform in DeFi. Their upcoming Centrifuge chain comprises very innovative elements that bring decentralized lending to new standards.
Similarly, their latest partnership with BlockTower demonstrates that the project is the best positioned project in the DeFi space from an ecosystem perspective, given their semi-symbiotic relation with MakerDao on RWA.
They have all the instruments to become a massive success, but technology and positioning alone are not enough, and to become a success you ultimately need customers.
I do believe that a stronger focus on distribution, particularly on the borrower side which usually proves to be the bottleneck in P2P platforms, will be key and necessary to help the team achieve the ambitious vision they have.