A standard and an exchange for on chain loans.

The problem today

Right now each lending app keeps its own score. It values its own loans and reports its own losses. So anyone lending into it is trusting the borrower's own bookkeeper to mark the book.

When a loan goes bad, the app itself decides when to call it, and it often waits to work out a quiet deal first. So the books can look fine right up to the day they do not.

And there is nowhere to sell down the risk. The money is locked in the loan, and there is no contract a lender can buy to cover itself if losses rise. The market prices its risk on faith.

What we build

We supply the three things this market is missing, and we sit in the middle of the money.

The first three are the market. The fourth is the seat with the real weight in it. Because every contract is funded in full, the cash is in our hands the whole time, and the moment a default is called the cash moves. Holding the money and moving it is the deepest seat in any market. It is the spot the whole market has to pass through, and it tends to outlast the trading floors that feed it.

Why we can hold the money and still be trusted

It looks like a conflict to hold the cash, call the default, and pay it out, all in one place. It is not, and here is the plain reason.

We never take a side in the trade. We make the same fee no matter who wins. So we do not care which way the money goes, and that is exactly what makes it safe for us to hold it and to make the call. A clearing house moves huge sums every day and does not care who wins, because it earns on the flow, not the outcome.

The payout runs on its own. Once a call is final, the contract moves the cash by code. No person here picks the winner by hand.

The call can be challenged. Anyone who thinks we got it wrong can put up a deposit and contest it before the money moves. We decide against a rulebook that was written down before the deal traded, not on a whim, and a bad challenge costs the challenger their deposit, so nobody contests for sport.

Why this is hard to copy

The math is easy to copy. Anyone with public data and a bit of code can compute a loss number. That is not what protects us, so we do not pretend it is.

What protects us is being the one everyone agrees on. A settlement number is worth something only because everyone uses the same one. The first trusted, neutral number that contracts settle on becomes the one the whole market points at, and a second copy that nobody settles on is worth nothing.

We can be neutral, and most others cannot. The lending apps, the lenders, and the trading venues all have a side. A number you can trust has to come from someone with no side, which rules out almost everyone with the means to build it. Few can hold that seat.

It grows stronger with time. A record of honest calls across good years and bad, and a clean history of losses, are worth more every year and cannot be rushed into being.

And the money runs through us. Once the trades clear and settle in our vault, that is the point the market cannot route around. That is the hardest seat to take away once it is ours.

Where else this works

The same machine is not tied to one app or one kind of loan.

Any on chain loan. The same number, the same call, and the same venue fit any lending where the lender can lose money to a borrower, from loans to trading firms to loans to real businesses.

Loans against tokenized funds. As stakes in private funds move on chain, loans against them move on chain too. They need the same three things, a trusted loss number for the collateral, a rule for when the loan is in trouble, and a place to hedge it. That is the market this company was built for.

Markets that turn on a fact. The part of us that settles a dispute is the same engine that could settle a bet on whether something happened, the kind of market where people lose when it is called the wrong way. That is harder, because those questions are fuzzier than did the money arrive, so it is a later step, not the first one. We say that plainly rather than claim it is solved.

What is true today

We will not pretend the venue is already running. Here is the honest state.

• The pool of loans that can actually go bad is still small and lopsided. At a sensible size cutoff it is about five names on chain, and one of them is roughly four fifths of the total. A broad basket that is hard to game needs more names than that, so the tradable venue waits on the market growing.
• Most of the loan detail we need sits inside each app, one level below what the public trackers show, so counting it is real work.
• The recovery on a bad loan is partly worked out off chain, so that one piece is not fully on the chain.
• The loss record so far is bunched in the 2022 stress, on the order of tens of millions, drawn from public reports and not yet checked deal by deal.
• Running a place that holds money, calls defaults, and pays out is the most watched job in finance, so the legal build is the real one.

The plan

Now. Publish the loss number, free for anyone to read, and honest about how thin the pool still is. That plants the standard and earns the trusted seat.

Next. When there are enough names and the legal piece is in place, open the venue, hold the money, and settle the first contracts.

The one thing that sets the clock. Count the real loans inside each app. That single number tells us how far off the venue is, and it is the next thing we do.