| Spread threshold | Per settlement | On $10,000 per day | Per year, roughly |
|---|---|---|---|
| 5 bps | 0.05% | $15 | 55% |
| 10 bps | 0.10% | $30 | 110% |
| 25 bps | 0.25% | $75 | 275% |
| 50 bps | 0.50% | $150 | 550% |
These are the same sets applied with one click in the screener settings. The values below are exactly what gets filled in.
from 12 bps · top 100 by turnover
Logic: only the top 100 pairs by 24h turnover, with high liquidity floors — open interest from $10M, turnover from $25M. Rate gap from 12 bps. The gaps are more modest here, but the pair has a deep market on both exchanges.
from 12 bps · every pair
Logic: every available pair, gap from 12 bps, open interest from $3M, turnover from $5M. The gap must also hold for three minutes straight — on thin pairs predicted rates jitter minute to minute. More hits than any other preset; the liquidity floors drop pairs where trading on one of the exchanges has all but stopped.
from 40 bps · top 100 by turnover
Logic: the threshold is raised to 40 bps — 0.4% per funding settlement. Rare hits, only on clearly visible rate gaps. Top 100 by turnover, open interest from $5M, turnover from $10M.
from 2 bps · SOL, DOGE, WIF, SUI, FARTCOIN
Logic: five popular coins at a 2 bps threshold. On coins like these exchange rates nearly match, and a higher threshold would never fire. 2 bps means "a gap has appeared", not "the gap covers the fees". Edit the list in the settings; open interest from $5M, turnover from $10M.
Any set is a starting point, not a recommendation. Once applied the thresholds are yours: move them and the screener starts counting the new way immediately.
Below 10 basis points the gap is usually eaten by the fees for opening and closing two positions, so 10–15 bps makes a sensible first setting. The slider runs from 1 to 100 bps; the higher the threshold, the rarer the alerts and the more room you keep for costs.
How many minutes in a row the spread must stay above your threshold before an alert is sent. Predicted rates jitter minute to minute, and a threshold read from a single minute catches one-minute spikes: by the time you open the terminal the spread is gone. Zero means the check is off, and the maximum is 30 minutes. Minutes that failed to collect are skipped during this check, so a collection glitch on one exchange does not cancel the alert.
The default kicks in — 8 hours, not “no pause”. On this screener the spread is a state rather than an event: without a pause the message would arrive every minute the spread holds. There is no per-coin alert cap in this form at all, so nothing else takes the pause’s place.
The widest spreads almost always show up on obscure coins — exactly where there is nowhere to open the opposite leg at the size you need. Narrow the scan with the pair universe (top 100 by turnover, or your own list) and with the open-interest and turnover floors.
Usually one of four reasons: the coin has no data from at least two enabled futures exchanges; the spread does not reach your threshold; it does not hold for the required number of minutes; or the pause on that coin has not expired. A separate case is “my list only” with an empty list — there is then physically nothing to fire on.
No. It answers one question: where the spread is right now. Fees, the cost of moving collateral, and the risk of one leg being liquidated are yours to count — which is why the threshold is worth setting with room to spare.
It converts the current spread to a year at an eight-hour accrual interval, so it is arithmetic and not a forecast: 10 bps per settlement works out to about 110% a year. A spread that size lives for minutes or hours, not a year — the message itself carries that caveat.