The fee buys
the pool.
Most protocols pay outside suppliers to provide their liquidity. The trading fee goes out as those payments, and when the payments stop the liquidity goes with them. dotpi holds part of the fee back and buys the liquidity it trades on.
Scrolling this page turns the loop once.
Rented liquidity
leaves when the
payments stop.
A protocol pays outside suppliers to park assets in its pool. Those payments are rent. They hold the assets there for as long as they keep arriving, and once they stop the protocol has nothing left to show for them.
The pool itself works perfectly well. It is the cost of keeping it full that never goes away.
Every trade
pays a fee.
Trading a token through a pool charges a fee. In a rented pool that fee is an expense, routed straight out to whoever is being paid to stay. It can equally be read as income, since it is the one thing a protocol earns without having to ask anyone for money.
dotpi reads it as income. Nothing else in the mechanism needs money from anywhere else.
Hold part of it
back, and buy
with it.
Some of each fee stays in the treasury instead of going out. That share buys both sides of a pair and places them in a rangeLiquidity placed between two prices rather than spread across every price. It earns fees while trading happens inside that band., which is where the trading actually happens.
This is the only decision in the mechanism. Everything after it follows on.
The position is
a token the
treasury holds.
A range like that exists on chain as a token of its own. It records which pair it covers and the two ticksThe discrete price steps a pool is divided into. A position is defined by the two ticks at its edges. it sits between. Anyone can trade against the pool, while the fees that range earns belong to whoever holds the token.
Holding it still carries risk. The assets inside can fall in value, and the price can move outside the range the position covers.
Where the return
comes from.
One part is the share of each fee that was never held back, which goes out the way it always did. The other is the fee income the owned range earns whenever someone trades through it.
Some of that second part buys the next position, and that is the step that brings this back round to where it began.
Ownership of the range does not depend on either. Income from it does.
Renting compared
with owning.
Both models charge the same fee. What separates them is where that fee ends up.
The last row is in the table deliberately. A comparison without one would not be much use to anybody.
Everything the
treasury holds is
published.
Owning something only matters if somebody can check it. Every field below is read straight off the chain rather than reported by hand, and that includes the two that make the numbers look worse.
The two flagged rows stay in. A treasury that only published the figures flattering to it would not be worth checking, so the ones that count against it are published on the same page as the rest.
The last step
is the first
one again.
Part of each fee buys a position, the position earns fees of its own, and some of those buy the next one. Whatever it does gets published, including the parts that went badly.
or keep scrolling02
- The fee
- Charged on every trade and split at the moment it is collected.
- The treasury
- Holds the positions it buys. Withdrawal is not part of the design.
- The ledger
- Positions, fees and balances are read off the chain and published.
- The loop
- Fees buy positions, positions earn fees, and the treasury compounds.