Most altcoins are impossible to value because there is nothing underneath the price. Perp DEX tokens are one of the few corners of crypto where that is not true. These protocols take a cut on every trade, and that cut is real money moving through a contract you can read. So instead of arguing about narratives, you can do something closer to what an equity analyst does. You can line the tokens up next to each other and rank them on price-to-real-fees. The hard part is figuring out which fees are real, because a lot of what shows up in the dashboards is the protocol paying itself.
Start with fee revenue, then subtract the bribe
The first number everyone quotes is total fees. It is also the number most contaminated by incentives. A perp DEX can look like it is printing money while it is quietly handing most of that money right back to the traders generating it, in the form of token emissions, rebates, or points that convert to tokens later. If you take fee revenue at face value and ignore the outflow, you will rank a subsidy machine above a business.
What you actually want is net fee revenue after incentive spend. Think of it as the protocol's fees minus whatever it is paying out to keep those fees happening. If a venue earns some amount in fees but emits more than that in token rewards over the same window, it is running at a loss on a fully diluted basis, and the volume is rented. When the emissions taper, the volume usually leaves with them.
A rough workflow I run on any perp DEX token before I take it seriously:
- Pull fee revenue over a trailing window long enough to smooth out one wild week, so a month or a quarter rather than a day.
- Pull token emissions and any rebate or points program over the same window, valued at market.
- Subtract the second from the first. If the result is negative or barely positive, the fees are being bought.
- Divide the fully diluted valuation by that net number. That is your real price-to-fees, and it is usually far uglier than the headline multiple.
The gap between the gross multiple and the net multiple tells you how much of the story is subsidy. A protocol trading at a reasonable gross multiple that goes to something absurd once you net out emissions is not cheap. It just has good marketing on its dashboard.
Judge buybacks by float retired, not dollars spent
Buybacks are where a lot of perp DEX tokens try to look like they return capital to holders, and where the framing gets slippery. A protocol will announce that it spent some large dollar figure buying back its token, and that headline is meant to sound like a dividend. The number that actually matters is what share of the circulating supply got permanently retired, and whether it was retired or just parked.
Two questions cut through most of it. First, are the bought-back tokens burned or sent to a treasury wallet that can redeploy them later. A burn is real. A treasury buyback is a promise, and treasuries have a way of selling back into strength when the protocol needs runway. Second, and this is the one people skip, how does the buyback compare to new emissions over the same period. If a protocol retires some tokens through buybacks while emitting a larger quantity as incentives, net supply is still going up. The buyback is theater layered on top of dilution.
So the rule of thumb is to express buybacks as a percentage of float retired per year, net of emissions. A program that quietly takes a couple of percent of supply out of circulation every year, funded by actual fees rather than treasury reserves, is worth more than a splashy one that spends a bigger dollar figure while the token count keeps climbing. Measure the float, not the press release.
Check whether the open interest is real
Open interest and volume are the metrics most easily faked, because for a while the protocol was paying people to generate them. Points programs turned trading volume into a farmable reward, and the predictable result was wash-heavy activity that exists to harvest an airdrop and evaporates the moment the airdrop lands or the criteria change.
You cannot always prove wash trading from the outside, but you can look for the tells. Points-farmed open interest tends to cluster in the pairs and the wallets that maximize rewards rather than the pairs a real trader would care about, so a venue where a huge share of OI sits in one incentivized market is suspect. It also tends to run at strange volume-to-OI ratios, with enormous volume churning against thin standing positions, which is what you get when people flip in and out to rack up a metric rather than hold a directional view. And it is fragile in a way organic flow is not. If you can find a moment when a points program ended or its rules tightened, look at what happened to OI right after. Real users do not leave because the points stopped. Farmers do, all at once.
The cleaner signal is whether open interest survives the removal of the incentive. A perp DEX whose OI barely flinches when emissions step down has genuine demand for its order book. One whose OI falls off a cliff was never really that big. When I am unsure, I weight recent, post-incentive data far more heavily than the peak numbers from the farming period, because the peak was manufactured and the trough is the truth.
Putting it into a comps table
Once you have those three adjustments, ranking the sector becomes a fairly boring spreadsheet, which is the goal. For each token you want fully diluted valuation over net fee revenue, buyback as a percent of float retired net of emissions, and some estimate of what fraction of open interest is organic. The tokens that look expensive on net fees, retire little or no real float, and lean on incentives to hold their OI are the ones the narrative is carrying. The ones that look cheap on net fees, quietly shrink their float, and keep their open interest after the rewards dry up are the ones with an actual business under the ticker.
None of this tells you where the price goes next quarter, and I would not pretend it does. What it does is stop you from paying a premium for volume that only exists because it is being paid for. In a sector where most of the tokens are dressed-up incentive programs, being able to spot the few that are not is most of the edge.