Know RWA, DeFi and AI in 4 Articles with Alpha.RWA | Part 2 of 4 Giveaway update: the first winner is @DannyP1333 , who receives $10. Another $40 goes to four more winners at the close of the series. Insightful quotes and comments keep counting. QUICK UPDATE: NYSA’s(@NysaFinance) FIRST MARKET GOES lIVE TODAY It runs on Solana, built on Kamino’s lending infrastructure, and it is the first market Nysa has launched. Part 1 covered what Nysa is. Part 2 follows one borrower through a Nysa market so you can see how it works. Every number in the example is illustrative. None of them are Nysa’s settings. WHAT NYSA BUILDS AND WHAT IT USES I. WHO RUNS THE LENDING ? Nysa does not write lending code. It uses established third party lending infrastructure and decides what runs on top: which collateral to list, how much can be borrowed against it, which assets can be borrowed, and how the market behaves when prices move. For the first market, that infrastructure is @kamino on Solana. Each asset issuer gets a market of its own. Nysa calls this counterparty isolated, which means a problem in one market stays in that market. Control is split. The admin keys, which can change a market’s settings, sit behind multi signature approval, with at least two of three signers required. Where the lending infrastructure allows it, the power to list new collateral stays with the infrastructure provider. Nysa keeps two things: the risk parameters and the interest rate model. Any change to those parameters goes through a 7 day timelock, a mandatory delay between queuing a change and applying it, and stays visible on the front end. A change queued on Monday cannot take effect for a week. THE FOUR NUMBERS THAT DEFINE A MARKET II. WHAT DOES A BORROWER DO? Meet Amara. She holds $1,000 worth of a tokenized fund share and wants stablecoins without selling it. She deposits the token as collateral. This market’s loan-to-value ratio (LTV) is 80%, so she can borrow up to $800. LTV is the borrowing ceiling, expressed as a share of the collateral’s value. Borrowing the full $800 would leave her with little room, and Nysa built Smart Loans for this moment. Smart Loans is a machine learning model, a program that learns patterns from historical data. In Nysa’s app, Amara selects her asset, opens Manage Position, chooses Borrow and taps Smart Borrow. The app sends her loan details to Nysa’s server, which returns a recommended amount and fills it into the box. The model studies more than 12 inputs, including recent volatility, how the asset moves relative to other assets, and simulated liquidation scenarios. It was trained on hundreds of thousands of historical position combinations and market simulations for the assets listed on Nysa. Say it returns $720. Amara reviews the number and signs. The model suggests, and she decides. III. WHAT HAPPENS WHEN THE PRICE FALLS? Every market has a second line above the LTV, the liquidation threshold (LT). Here it is 85%. If Amara’s debt rises above 85% of her collateral’s value, part of her collateral is sold to repay it. With $720 borrowed, her collateral has to fall to about $847 before that happens, a drop of roughly 15%. Had she borrowed the full $800, a fall of about 6%, to $941, would do it. The distance between the LTV and the LT is her reaction window, the time to add collateral or repay before anything is sold. How wide should that window be? Nysa measures the token’s volatility with EWMA, a method that gives recent price moves more weight than old ones, applied to the token’s price or NAV history. NAV, net asset value, is what the fund’s holdings are worth per token. Nysa then calibrates the LTV so that at least 95% of loans opened at the maximum ratio stay over-collateralized after 3 days, and at least 90% after 30. A calm asset needs less room than an erratic one. If liquidation does arrive, it works in pieces rather than all at once. Nysa calibrates the incentive paid to liquidators, uses partial liquidations, caps position sizes and sizes a reserve fund for losses that remain. The LT itself comes from backtesting against multi year price history, with each overnight or weekend price gap charged in full. LIQUIDATION,MONITORING AND THE FREEZE IV. WHY DOES A MARKET LIMITS IT COLLATERAL? A market sets a supply cap, a ceiling on how much of an asset it accepts. Suppose the fund share can only be redeemed for cash in limited amounts at any moment. A market holding too much of it could not sell the collateral fast enough in a crash. Nysa sizes the cap so that the amount redeemable instantly covers all of the collateral expected to be liquidated under stress. It does not size it for the extreme case where every position is liquidated at once. V. What does borrowing cost, and what do lenders earn? Amara pays interest, and lenders who supplied the stablecoins she borrowed earn it. The rate moves with utilization, the share of deposited money that has been borrowed. Nysa uses a kinked curve: rates climb slowly up to a target utilization, then steeply past it. Lenders earn the borrow rate, scaled by how much of the pool is lent out, minus a Reserve Factor that the Nysa treasury keeps. Take an illustrative curve: a 1% base rate, a target utilization of 80%, a first slope of 4%, a second slope of 60% and a 10% Reserve Factor. 1 ) At 40% utilization, the borrow rate is 3%. Amara’s $720 loan costs about $21.60 a year. A lender with $10,000 deposited earns about 1.08%, roughly $108. 2 ) At 80%, the borrow rate is 5%. The same lender earns 3.6%, about $360 a year. 3 ) At 90%, the borrow rate jumps to 35%. Amara’s loan would cost about $252 a year, and the lender earns about 28.35%, around $2,835. The jump past the kink keeps withdrawals possible. When almost everything is lent out, the high rate pushes borrowers to repay and pulls lenders to deposit. 4 ) What if the token’s price and its value drift apart? Suppose thin trading pushes the fund share’s market price 5% below its NAV, even though the fund holds the same assets. Nysa calls a gap like that a depeg. A price feed based on market price alone would show Amara’s collateral worth less than it is, and could trigger liquidations the underlying value does not justify. So where the oracle provider supports it, Nysa asks for price feeds based on NAV or redemption value. Each market also has deviation thresholds beyond which it can be suspended or frozen. A real-time system watches for on-chain anomalies and can freeze a market where the technology allows. Volatility readings feed back into the LT and LTV continuously, and the independent proof-of-reserve check on issuers runs annually. VII . WHAT TO WATCH With the first market live, the parameters behind it are public, and any change to them stays visible on the front end. The roadmap sets the next targets: at least 2 new markets backed by RWA collateral, and at least 3 markets live, each with 40%+ utilization against its supply cap. After that come the DAO launch, with a TGE, a public INO, a staking module and a stNYSA credit market, and a Skin-In-the-Game vault meant to protect lenders against default on a Nysa-curated market. Part 3 looks at where the value comes from and who uses it. Sources: Nysa Finance documentation (What is, Market-as-a-Service, Risk Management and its Market, Protocol and Credit Risk and Risk Monitoring pages, Interest Rate Models, Smart Loans, Roadmap).