Their first (and best, probably) metric is actually quite similar to the third criterion Troy and I used when putting our real volume report together last year: ftx.com/volume-report-. But theirs misses something key.
Conversation
Comparing just to the regulated US exchanges does achieve a bit of a guarantee that you’re comparing to *real* volume -- but it’s gonna be biased toward real *U.S.* volume. We adjusted for that by rerunning against established Asian exchanges, too.
1
This is a pretty important adjustment to make -- derivatives exchanges mostly ban U.S. customers, so biasing towards them unfairly hurts derivatives exchanges.
1
The U.S. exchanges also basically don’t have leverage, which creates an effect where the times they’re busiest, they have WAY more volume than the U.S. exchanges can -- and that can fuck with the simple correlation, too.
1
2
Their other two metrics are plausible, too, but neither has anything to do with analyzing trades / order books the exchanges are actually reporting. For comparison, all six of our metrics analyzed those things, as they’re basically the whole story.
1
1
I’m not claiming that our metrics are either perfect or complete -- but they yield results that gel with our experiences actually doing a ton of volume on every exchange where that’s possible, and I think that’s a giant update that our methodology makes sense.
1
1
At some point we’ll likely update our methodology to include more diverse metrics (I do think there’s something to web traffic, FWIW) -- but for now, people should at least know that OKEx does not deserve a 0/3 score on basically any legit volume test. ftx.com/volume-monitor
3
1
6
My working hypothesis is what can be easily percieved as wash trading is in reality teams trading against themselves to obtain a better fee tier. If that's true, then all okex needs to do is to cramp down on these activities harder
1
Hmm does this work? I wouldn't think the extra fees you pay doing this would make this make sense but have not checked.
1
My guess would be, team negotiating a fee holiday of maker-taker difference of 0.5bps and then utilizing it. But like I said, it a hypothesis. We are traders, not research outfit, so dont care much to investigate. vast majority of volume on okex is legit, but wash trading exists
1
I think you need something else at play to make it work, something like:
a) exchange says if you hit X volume by Z time you get Y fees forever
or
b) someone's maker + another's taker < 0
Still might work, i.e. demonstrating the necessary volumes for one month to get the discount for the next three months might just make it. But it would be silly, I agree, just the only other explanation for (limited) wash trading would be okex doing it themselves...