Chainflip liquidity

Grouped by swap: one arriving in one pool takes resting orders from the best tick outwards, and every fill it produced shares a block and a market. The window counts swaps rather than rows for the same reason — a row limit would cut one in half and leave its totals describing a fragment. Narrowing the market or direction narrows which swaps are counted, so a quiet market reaches further back instead of coming up empty.

Every provider's fills in the markets listed above — plain LPs as well as registered trading strategies, and range orders as well as limit orders, all marked as such — but only those markets. A swap takes whatever is in its way, and one filled entirely by range orders would otherwise leave this table blank while the market's earnings above kept counting it.

The two are different instruments. A limit order rests at a tick the provider chose, and its spread is what that choice captured. A range order has no tick: it holds liquidity across a band and earns the pool's fee on whatever passes through, so its rate is the fee tier rather than anything the provider decided. The rate column is fee-over-volume in both cases, which is why they can sit in one table — but only the limit rows say something about how well a provider is quoting. BTC, ETH and SOL pools are excluded for the same reason they are missing from the table above: no strategy here trades them yet. FLIP, DOT and TRX go with them, having no oracle feed to value a fill against at all.

Trades under a dollar are not listed. Below that a fill measures rounding rather than quoting: an eight basis point edge on a few cents is a fraction of one fine unit, so it floors to nothing, the maker captures none of it, and the fill lands wherever the two oracle feeds happen to sit. Those rows dominated this table by count while accounting for none of the money. Batch totals here cover only the fills shown.

Each leg is valued at the same block's oracle price, so both sides and the earnings between them are on one basis. The traded amounts are shown beside the earnings on purpose: a few cents on a few thousand traded is an ordinary stablecoin spread, and the same earnings on a few dollars traded would mean something was wrong by orders of magnitude. The figure cannot be judged without them.

Tick and spread are not the same measurement, and will not agree. A tick is a step of 1.0001 against the pool price, so it reads as whole basis points. The spread is measured against two independent oracle feeds, and USDC and USDT are not priced alike — the gap between them runs several bps and moves with every revision, which is why the spread is fractional where the tick is not. Four fills at tick −8 in one block measured 0.22, −0.08, −7.71 and −7.71 bps: the feeds were 7.71 bps apart, and the two smallest fills were dust — too few fine units for an 8 bps edge to round up even one — so they captured nothing from the pool and kept only the gap.

Every fill this account made in the window, grouped by the swap that caused it. Unfiltered, unlike the activity feed on the Strategies tab: that hides trades under a dollar because chain-wide dust dominates by count while accounting for none of the money, but a fill you made is not noise to you — and at these sizes the floor would hide most of the account's activity. The earnings above sum every fill, so this list and that figure agree.

Value is everything the account holds — free balances plus whatever is resting in open orders — priced at the oracle rate recorded with each observation, so a past point shows what it was worth then rather than what it would be worth now. Earned comes from the account's fills, not from the change in value: paying money in raises the value without earning anything, and an APY built on that would flatter the account at exactly the wrong moment. APY annualises those earnings over the time-weighted capital that produced them, and is withheld below an hour of history, where annualising a single fill produces a number in the thousands of percent.

Live from the chain's lending pallet. Supplied is what lenders (LPs) have put in; borrowed is out on loan now, with the pool's utilisation beside it; interest is the current borrow rate, which the chain sets from utilisation. Amounts are in each pool's own asset. Expand a pool for its lenders and loans.

Ticks, where one tick is one basis point. Mid is the oracle mid the bot prices from; bid and ask are what it is quoting right now, and — means it is not quoting that side at all — the JIT trigger only fires on a side with volume heading into it. Offsets are the configured distance out from the mid on each side.

The decision line is the last thing the borrowing rule concluded, refusals included — a rule that silently declines is indistinguishable from one that is broken. Observed is the margin the rest of the market has been clearing at on each side, which is what a borrow is judged against; it must reach the minimum before the bot will take on debt.

Net is the question the whole feature turns on: what the trades made on borrowed money, less what borrowing them cost. Earnings are marked to the oracle mid at the block each fill landed — the same measure applied to every other provider's fills — so this is a valuation, not settled cash, and it only means anything once every loan has closed. Volume alone is not success: at 1 bp, $5,000 of borrowed trading earns $0.50 against roughly $0.66 of financing.

Supplied is in the lending pools, earning — and it is also the collateral a loan is secured against, which is why borrowing capacity and trading size compete for the same funds.

FLIP has no oracle price feed, so it is kept out of the earnings and spread views. What a swap does have is a price: the USDC it moved per FLIP, the ratio of its two legs — shown for limit-order trades, which carry both legs. Most FLIP activity is the protocol swapping collected fees into FLIP (the burn): these arrive as range-order buys, tagged fee/range, and record only their USDC input, so they show a size and a direction but no price. The price line and range are from limit trades; the fee-burn figure sums the range buys.

Freshness comes from each price's oracle timestamp and the timeouts as ingested from the chain — governance can change them, so they are read rather than assumed. These are aggregate prices: whichever of Arbitrum or Ethereum the runtime found freshest.

These prices are what values everything on the Strategies tab. They are here rather than there because they are the input, not the subject: a liquidity or earnings figure is only as good as the price behind it, and this is where to check that the price was sound.