Why Funding Rate Scanners Are Missing The Most Important Variable.

There is a category of crypto tool that is now almost universal among retail perpetual futures traders. The funding rate scanner. You know the kind I mean. A grid of assets across the top, a list of exchanges down the side, every cell showing some number with a percent sign, and at the top of the page, sorted in big bold text, the headline arbitrage opportunities. Long here, short there, projected return in three digits, click here to trade.

I have been building one of these myself for the last few weeks. Not because the world needs another funding rate scanner. The world has plenty. I have been building one because, after using the existing ones for actual trades, I noticed that every one of them is doing the same thing wrong, and the thing they are doing wrong is not subtle. It is the central variable in whether any of the numbers on the page mean anything at all.

That variable is time.

What the scanners actually show you

When a funding rate scanner displays a rate next to an exchange, the rate it shows is the rate as of the last funding snapshot. Most exchanges run on either a 1 hour, 4 hour, or 8 hour funding cycle. The rate gets locked in at each snapshot, paid out, and then a new rate gets calculated for the next period. What the scanner shows you, in real time, is whatever the current period’s rate happens to be at the moment you look.

That number is not what gets paid. What gets paid is whatever the rate is at the next snapshot, which could be five minutes from now or seven hours from now depending on which exchange and where you are in the cycle. The current rate is a leading indicator, sometimes a useful one, sometimes a misleading one. On thinly traded markets it can swing fifty percent in either direction over the course of a single funding period. The number on the scanner is a snapshot of an evolving system, presented as if it were a fact.

This is bad enough on a single exchange. It gets much worse when you start comparing rates across exchanges, which is the entire point of these tools.

The cross-exchange problem

Here is the trade the scanners are designed to surface. Asset X is paying positive funding on Exchange A and negative funding on Exchange B. The spread between the two is some impressive number. You go long on the exchange paying you to be long, short on the exchange paying you to be short, lock in a delta neutral position, and collect the spread. In theory.

The theory works only if the rates you saw at trade entry are roughly the rates that get paid at the next snapshot. And the rates that get paid at the next snapshot have only one chance to be close to the rates you saw, which is when both exchanges’ next snapshots fire close enough together in time that nothing meaningful can happen in between.

If Exchange A is on a 1 hour cycle and Exchange B is on a 4 hour cycle, your trade has a serious problem. Exchange A’s next snapshot might be 47 minutes away. Exchange B’s next snapshot might be 3 hours and 47 minutes away. Three hours of price action sits between when you collect from Exchange A and when you collect, or do not collect, from Exchange B. The number that locks in on Exchange B in three hours is not the number you saw on the scanner. It cannot be. The market does not stand still for three hours so that your trade can work.

If Exchange A and Exchange B are both on the same cycle but their snapshots are offset, you have the same problem in a smaller form. If Exchange A fires at the top of the hour and Exchange B fires at the half hour, you have thirty minutes of drift between the two locks. Not as bad as three hours, but still enough to turn a clean spread into a bad trade.

The displayed spread on every scanner I have looked at ignores this. It just subtracts one number from the other, multiplies by the number of funding periods in a year, and slaps the result at the top of the page in green. There is no field anywhere that tells you when each leg’s next payment fires. There is no warning that the two legs are on different cycles. There is no acknowledgment that the spread you are looking at is not the spread you will actually capture.

This is not a small omission. It is the omission. Time alignment is the single variable that determines whether a displayed spread is real or fictional, and the entire scanner industry has decided not to show it to you.

Why this gets ignored

A few reasons, all of them honest in the sense that they are commercial reasons rather than malicious ones.

The first reason is that showing time correctly makes the headline numbers smaller. If a scanner currently displays “1,500% APR” as the projected return on a misaligned trade, and the honest version of that scanner shows “this spread is not capturable because the two legs do not fire within an hour of each other,” the user does not get the dopamine hit of seeing the big number. The scanner that surfaces fewer opportunities loses the click. The scanner that lies to make the number look bigger wins the click. The economics of building scanners reward the version that misleads.

The second reason is that most scanners make money through referral arrangements with the exchanges they list. Every time a user clicks through to open a position, the scanner collects a fraction of the trading fees. The scanner does not particularly care whether the trade you put on is profitable to you. It only cares that you put it on. Surfacing the time alignment problem reduces the rate at which users put trades on. The honest scanner is worse for the scanner’s revenue.

The third reason is that doing it right is harder. Every exchange’s API returns funding data in a slightly different format. Some return native rates per period, some return annualized values, some return rates pre-multiplied by their own conversion constants for “comparison purposes.” Building a scanner that correctly normalizes all of these into a unified display, and then layers on the time alignment information, takes more engineering than copying the existing pattern. If you are building a scanner to capture referral revenue rather than to be useful, the engineering investment in honesty does not pay off.

I am not saying the people who built these scanners are bad. The incentives lead them where they go. I am saying that if you are using one of these tools, you should know what they are, and you should know what they are missing.

What the honest version looks like

The version I am building does a few specific things differently.

Every cell shows the funding rate, the venue’s native funding interval, and a countdown to that venue’s next snapshot. The countdown is not decoration. It is information that determines whether the rate in the cell is going to be the rate you collect.

Sort options include not just rate magnitude and spread, but time to next snapshot. Click that sort option and the assets where snapshots fire soonest come to the top, regardless of whether the displayed spread looks impressive. Because the displayed spread is only meaningful if you can capture it, and you can only capture it if the snapshot fires soon.

A planned filter takes this further. Show me only the assets where two or more visible exchanges have their next snapshots firing within the same sixty minute window. Because that is the actionable subset. Everything else is informational; that subset is tradeable.

The cross-venue spread sort is still available, because traders want it. But when it is active, a banner appears reminding the reader that spreads sorted by magnitude are usually not the most tradeable spreads. The two are different things, and conflating them is the mistake the entire industry makes.

None of this is technically hard to build. Every piece of the data exists. Funding intervals are public information. Current UTC time is, well, current UTC time. Computing time to next snapshot is a one line function. The reason this is not standard across funding rate scanners is not engineering difficulty. It is incentive misalignment. The tools that surface this information cost their builders revenue.

What this means for actually trading

If you are running real money on cross-venue funding plays, the practical implications are these.

Before you put on a trade, look at when each leg’s next snapshot fires. If they are not within roughly an hour of each other, the spread you are looking at is not the spread you will collect. The misalignment buys time for the rates to drift, the prices to diverge, and your delta neutrality to break down.

Prefer trades where both legs are on the same funding interval. A 1 hour exchange paired with another 1 hour exchange is a much cleaner setup than a 1 hour paired with an 8 hour. Both exchanges fire at the top of the hour, both rates lock at the same moment, both pay out at the same moment.

If you are going to trade across mismatched intervals, time your entry to when both legs are close to firing. The same trade entered at 59 minutes before each exchange’s snapshot is much different from the same trade entered with three hours of waiting on one leg. The first version can capture the spread. The second version is open exposure dressed up as arbitrage.

And generally, treat any “max APR” or “best opportunity” surfaced by a magnitude-based scanner as a starting point for research, not as a trade. The scanner has done one thing for you, which is show you that something is interesting. The next step, the part where you figure out whether the something is real, is on you.

What I am actually building

The site I am building is a directory and editorial publication for retail perpetual futures traders. It tracks funding rates across the venues that retail traders actually use. It surfaces the same data the existing scanners surface. It also surfaces the variable that the existing scanners hide, which is time.

The grid is live and you can see it for yourself. The assets I track are a curated set of 90 priority symbols, and the venues I track include the major decentralized exchanges and the major centralized ones for comparison. Every cell shows the rate, the interval, and the countdown. Sort by snapshot timing and you will see, immediately, which opportunities are time aligned and which are not.

The point is not that this tool is going to make anyone rich. The arbs surfaced by even an honest scanner are small. The point is that the tool is honest about what it is showing you, which is the bare minimum a tool you trade against should be doing.

If that sounds boring compared to “1,500% APR,” that is the whole argument. The boring version is the real version. The exciting version is the one that costs you money while showing you a good time on the way down.