How Hadron Prices a Trade

spread = price_curve(input_amount)
       × inventory_curve(your_inventory)
       × base_spread

output_bid = input × spread × oracle_price
output_ask = input × spread / oracle_price

​

The quoting equation

fee    = amount_in × fee_ppm / 1,000,000
net_in = amount_in − fee

price     = price_curve(net_in)
inventory = inventory_curve(your_inventory)
spread    = min(price × inventory × base_spread,
                clamp)

out_bid = net_in × spread × oracle_price
out_ask = net_in × spread / oracle_price

then: triggers, min_out check, drawdown guard

• A final spread of 0.999 means 10 bps of spread around your oracle price, on both the bid and the ask.
• Oracle price is quote atoms per base atom in Q32.32, as are factors (1.0 = 2^32).
• The clamp caps the spread factor at 1.0 by default, so you never quote better than your oracle price. See Risk controls.
• The fee, and any token transfer fee, comes off the input first. See Fees.

​

Your curves

• price_curve(input_amount) sets the factor by trade size, so larger trades can be priced wider.
• inventory_curve(your_inventory) sets the factor by your inventory, so your price skews as your position moves without you sending an update.

• Stateless (the default). Every trade reads the price curve from the start, so identical trades get identical prices.
• Stateful (integrated). Trades use up the curve like a book and pay the average across the depth they take, so a burst of orders walks down it. It refills at a rate you set, or when your crank refills it.
• % mode. The inventory axis is your share of value in base, as a %. This way you can set how your quotes skew at, for instance, an 80/20 imbalance.
• Integrated. The inventory axis is your token balance, and the factor is averaged over the inventory move the trade causes.

​

Fees