We contributed to Lido DAO, P2P.org, =nil; Foundation, DRPC, Neutron and invested into 150+ projects
Currently, we already have an onchain lending industry with almost $50B TVL and $25B active loans. The protocols run on a simple basis:

But the TAM for existing lending protocols is relatively small compared to what we have in TradFi. These protocols assume that:
Today, stablecoin lenders generally pay more interest than the actual interest rate we have on USD because the onchain stablecoin supply is still relatively small compared to the demand. But where does the demand come from?

Today everyone is happy, leverage seekers are paying high interests to suppliers and suppliers make more than interest rates. But is this going to keep like this, when stablecoin acceleration will speed up?

CitiGroup recently shared an article about their forecast for stablecoin total issuance: $1.9 trillion base case (previously $1.6 trillion) and $4.0 trillion bull case ($3.7 trillion). They think that there will be trillions of stablecoins sitting on-chain. As a result, in the future there will be a lot of idle stablecoins, so it is not difficult to guess that collateralized lending will be much cheaper. But will there be enough yield that can pay such a good yield to this amount of stablecoin supply? In today’s context, no.

This is where undercollateralized credit will enter the market. Because collateralized lending is safe and relatively risk-free (bad-debt risk is almost zero if the protocol runs as programmed), there will be a lot of stablecoin holders ready to give credit to a more risky group: loans based on a credit score, future cash flow, holdings, and so on. This is much more different than what we have in the on-chain lending market today, and it is a lot riskier. Even if they give credits to people with 20% of the real interest rates, it will be a completely new and big market. So, what are our expectations to make this real?
In the past, undercollateralized credits have been tried by many protocols; most of them gave credits to market makers and investors whom they thought would pay back. But unfortunately, as in the example of TrueFi, Goldfinch, and Maple Finance, those protocols can go underwater and never pay back their debt. The reason: the loan takers were taking too much risk and got access to credit that they could not get on other markets; and mainly the issue was they did a lot of leverage and risky investments. Furthermore, some of the protocols focused on giving undercollateralized credits to retail users, but they also didn’t become successful because crypto retail users have one bad habit: if they can exploit something (even for $1), they will, and they think this is their right. We saw this with reward programs, affiliates, and so on. Therefore, there should be good ways to push people to pay their debt back. So, what can enforce people to pay their debt back:
1. Cutting the relationship with the product: Protocols can start from small amounts (like $10-$50-$100) and use zkTLS-like methods to be sure that they have the money to pay back. And if they don’t pay, the protocol can ban them and never give any more debts. This sounds reasonable, but I can promise that there will be people who will use their families’ identities to exploit the system and never pay back.
2. Onchain Auction of the Debt: We can first ensure that they have money, a credit score, and/or future cash flow; and then run an auction for debt collectors to buy the debt from the protocol if the borrower does not pay. For this to work, debt collectors should be able to run necessary legal enforcements and make the borrowers pay their debts back. This system can easily scale to big players, including HNWIs, mid-level credit takers, and even crypto degens.
I don’t think one solution fits all scenarios here. But this is a good solution to think of; I feel like this is the way to go to market and gain traction.
The stablecoin supply is growing, and it has been growing much faster than the broader crypto market over the last four years. Because of that, it is clear that stablecoin yields will decline in the coming years. As a result, crypto will increasingly become a place where businesses access cheap credit. But given the nature of crypto, this system cannot function in a fully permissionless way, because the under collateralized credit system is not built on crypto rails. We will need more verifiable credentials onchain, and private-credit projects will have to find mechanisms to issue credit legally and enforce obligations when borrowers fail to repay. And please don't give credits to motorcycle companies in Nigeria :), or at least do the research better.
This is extremely difficult to operate at a global scale, and it inevitably breaks some of the seamless onchain composability we have today. Still, I think the abundance of free stablecoin liquidity will create obvious scenarios where businesses tap crypto rails simply to earn incremental yield on top of T-Bills.
Thanks @_vshapovalov for the inspiration!