How can we make rates anonymous and permissionless without sacrificing efficiency?

Part I: The Problem

Interest rates are the demand for money. By this logic, utilization rate should be the perfect calibration in which to determine the rate at which a dollar (or other asset) is borrowed at. As the amount borrowed relative to supply increases, so does the rate. At its basic level, this would mirror a supply and demand curve that only theoretically exists. In crypto, this is executed with a bonding curve and various adjustments to encourage certain behavior, that is, at high capacity rates exponentially rise to discourage borrowing and at low to medium capacity the rate remains relatively constant. Top to bottom, this construct makes sense mathematically. So why does utilization rate fail the sense test?

A simple example to illustrate. Let’s say I am an uncreditworthy individual with a history of not paying my bills and I want to borrow $100k. Apple on the other hand is a massive $2T+ market capitalization public company with billions of dollars in cash piling up and Apple wants to borrow $400k.

By the laws of utilization rate, Apple will be paying a far higher rate on it’s borrowings than I will all things equal. This is one point of failure where the math, equation, logic, however you want to characterize it, breaks down. On-chain, there is no or minimal identify to differentiate me from Apple. We will discuss on-chain credit scores, bringing further assets on-chain, etc. in part II.

Suggestion I: Apple could put up more collateral into the pool to drop borrowing rate. They could do that, depositing additional cash collateral however makes this trade inefficient from a capital allocation perspective and why would you deposit your own funds at 2% to borrow that same cash at 5%? What if they did need all of this money anyway?

Suggestion II: Apple could go to multiple protocols and get better rates. As of this writing, there is approximately $20B in total unused liquidity between Compound, MakerDAO, AAVE, and Solend. This is likely an option but a headache from an administrative, timing, and cost standpoint. Assuming Apple had adequate on-chain collateral to deposit into each of these protocols, they would need to collect assets and repay interest in multiple locations, this option is certainly doable though not ideal

This problem is only exacerbated as we have more money market forks across chains that are replicas of Aave and Compound with a tweak here and there. While each of these forks are materially the same, they all require different collateralization levels, have different liquidation mechanisms, altered rate constructs and the list goes on. In reality, the fractured liquidity in crypto looks something more akin to the image below with 20+ additional buckets.

Some of this is solved by bridging liquidity and thoughtful traders and arbitrageurs but generally, the market for liquidity within crypto remains wildly inefficient creating both opportunity for improvement and profit.

Switching the above example a bit, if I would like to borrow $1 billion my rate should be higher than Apple who wishes to borrow $400k. The point here is that credit decisioning, while completely imperfect, is not made off of math alone and includes a series of exogenous and borrower specific factors.

“Democracy is the worst form of government, except for all the others” — Winston Churchill.

Fractured liquidity, imperfect supply and demand dynamics, and the inability to manufacture effective on-chain identities cause the aforementioned imbalance…Utilization Rate is the best we have. The quick solution is that if more liquidity comes into the system we will have a more efficient system as the deposited liquidity trickles through one chain then another and so on. While this is likely true it ignores the obvious math problem here, no pun intended, and their mechanics.

To complicate factors, all but a few protocols have truly been battle tested. Even fewer can really handle material inflow anyway without sacrificing on rate materially. Goldman Sachs may salivate over the idea of 3% yields on cash but as soon as they drop $1B in capital into AAVE, where a 3% yield is aspirational, it evaporates because they did it. $1B in some protocols would be a 100%+ increase in a single transaction, a full turn increase in liquidity they are already unable to allocate effectively. Suffice to say it doesn’t seem like anyone is ready for “material” liquidity the size of which could fix the fractured nature of crypto money markets.

Until protocols homogenize further or liquidity aggregators emerge to efficiently allocate these funds (hello DAOs), adding more dollars to the system does not solve this problem. In traditional markets, it would be rare for a single company to shoulder a full loan of this size simply for risk management purposes. Goldman Sachs might originate a loan of this size and fund it initially to get a deal done and then farm our SNCs or sell off pieces to reduce risk, ultimately the risk is borne by a diversified set of lenders and rarely a single one. Depositors can break up capital administered by Aave but it is only Aave that shoulders this liquidity burden. If we look at debt in the traditional capital markets we see it goes a step further, slicing up bond issuances and letting them trade on the secondary markets at premiums and discounts. On a large, read national or global scale, utilization rate does make sense. In crypto it just does not work this way. Eventually this will become a reality, eerily similar in the way loan-to-deposit ratio of banks trail Fed rate bumps/cuts but that is years to decades away.

The thirst for liquidity is not quenched until rates are sustainably better in Crypto. This then means uses of capital in Crypto also need to be better, not simply reward emissions and trading volatility for yield. Utilization rate can bring institutions onboard but there needs to be a huge demand for capital before that happens which means we are likely in a space where DeFi and TradFi battle for the foreseeable future arbing out differences in one another until they are indecipherable from the borrowers point of view.

In Part II we will explore how on-chain identification, fixed rates, and liquidity aggregation constructs could creatively solve this problem and provide a further gateway for TradFi into DeFi.