DefiLlama has been a dominant force in crypto data for some time, and now they’re using their positioning to settle bets, literally.
On Oct. 20th, DefiLlama announced it was partnering with Trueo, a permissionless prediction market protocol on Base, to launch markets that “settle based on DefiLlama data,” covering protocol revenue, DeFi exploits, and some high-profile chain metrics.
— (@)
Revenue is the meta today, and the revenue markets work like crowd-priced earnings estimates for onchain businesses. Conversely, exploits remain DeFi’s biggest specter, and the exploit markets put a live price on that risk.
It’s still early, so the size of these venues is still modest for now to be sure. Regardless of that, I think it’s obvious that these opportunities will, or at least have a shot to, become much more popular. Let’s break down what I mean.
How the markets work
The current slate of offerings spans six monthly revenue lines (Hyperliquid, Pump, Uniswap, Aero, Papertrade, and Collector Crypt); October exploit or bad debt markets for Aave and Morpho; and weekly markets on whether any +$1M or +$5M hack hits crypto, among others.
You trade in these markets with TYD, Trueo’s USDC-backed stablecoin. One TYD mints a YES share plus a NO share, and the winning side redeems for 1 TYD.
Notably, YES and NO each trade in their own Uniswap pool, so you can understand a YES price of 0.82 TYD as an 82% implied probability.
DefiLlama’s data is the source of truth, naturally, but Trueo’s optimistic oracle does the resolving, with bonded proposals, a 12-hour challenge window, and an escalation path for disputes. DefiLlama earns creator fees as each market’s curator.
But like I said before, these are small markets at the moment, with roughly $10k to $25k of liquidity apiece, so many of them may come with compelling mispricings.
Earnings estimates for onchain businesses
Back in August, I wrote about holders revenue as crypto’s new dividend metric, i.e. the market increasingly values crypto projects like businesses.
What crypto hasn’t had is the rest of the earnings ritual, where analysts publish consensus estimates and stocks move on beats and misses. These DefiLlama revenue markets are a crowd-priced take on that, though onchain “earnings” print in real time, and each market page draws the run rate right next to the odds.
So are the odds tracking well here so far? I lined up October revenue through Oct. 20th against what each of the tracked protocols needs over the final days of this month:
As you can see in the graphic above, a few trends jump out.
For instance, favorites are currently trading cheap.
Collector Crypt needs ~$200k per day to clear $15M, and its worst full day of October was $218k, yet its market opened at 62%. This dynamic undoubtedly prompted DefiLlama CRO Patrick Scott to later quip that “They’re selling dollars for $0.94.” Likewise, Hyperliquid has averaged ~$2.15M per day against the ~$1.54M it needs, and the market is still giving it an ~18% chance of missing presently.
On the flip side, longshots are trading rich.
For example, Uniswap needs ~$477k per day to hit $10M, i.e. more than its best day of October, every day, for 11 straight days. Aero needs ~90% of its best day, sustained through Halloween. Both trade at 17–18% odds at the moment. Both did clear $15M in September, to be fair, but a repeat burst would have to start now and run furiously.
Enjoying this article?
Subscribe to Bankless or sign in
Papertrade is seeing more harmony with its activity, however.
The 1,000x exchange booked ~$61.8M from its Oct. 10th launch through Oct. 20th, so a naive average says $85M is a lock. But launch day alone was ~$25.2M in revenue, and daily revenue has slid from $7.5M to $1.45M since Oct. 15th. At its recent pace, Papertrade finishes near $80M, and the market saw that coming, having slid from 68% to 19.5% in under two days.
Why Papertrade Won't Let PAPER Trade on Bankless
In a house pool exchange, the token is the house. That's why Papertrade won't let it trade.
Bankless
Interestingly, those first two trends are demonstrations of favorite-longshot bias, documented at racetracks for decades. Bettors generally tend to overpay for longshots and underpay for favorites, especially in thin, new markets. That’s why Patrick Scott noted on launch day that “There’s a lot of opportunity to arbitrage these markets right now.”
Putting a price on hack risk
Now, onto the exploit markets.
Another thing that Scott’s noted is that if the market prices an exploit at 1%, spending that 1% on YES lets you “fully hedge your position.” The two kinds of exploit markets here behave very differently, though.
The weekly markets are priced to history. With ~5 days left, the +$1M hack market sat at 82.7% and the +$5M market at 50.6%. In DefiLlama’s hacks database, 82% of all five-day stretches over the past 12 months had a +$1M hack while 47% had a +$5M hack. The market is pricing the trailing base rate almost exactly!
The catch is that 2026 has run hot. DefiLlama logs 111 hacks of +$1M so far this year, versus 70 in all of 2025, and over the past six months, 92% of five-day stretches had one. If that regime holds, the weekly YES sides look a bit cheap.
In contrast, the protocol markets are priced like a premium. With ~10 days left in October, Aave’s YES side traded at 6.8% and Morpho’s at 11.9%. The Aave market pays on a +$1M exploit or +$1M of bad debt.
That latter pillar is arguably the lower-hanging fruit, since Aave’s biggest recent scare wasn’t a bug. When Kelp DAO’s rsETH bridge was drained in April, the attacker borrowed against the stolen rsETH on Aave V3, leaving Aave modeling $124M to $230M in potential bad debt.
As for cost, plain insurance is still far cheaper for now. Nexus Mutual, DeFi’s best-known cover protocol, quoted ~$6.75 to cover $10,000 of Aave V3 for its 28-day minimum term. The same protection on Trueo currently means buying ~$680 of YES at 6.8%, and that would only cover the ~10 days left in October (before slippage and with liquidity permitting).
In fairness, though, these aren’t the same product. Nexus requires proof of your own loss, carries a 5% deductible, and likely wouldn’t have covered a Kelp-style collateral blowup, while a Trueo YES share pays out regardless of whether you lost a dime.
Some of the premium here buys that broader trigger, sure, but most of the gap is likely the favorite-longshot bias again. YES on a rare hack is the ultimate longshot, of course.
As for who sells the hedge, buying NO is functionally underwriting. BOB founder Alexei Zamyatin suggested big DeFi apps could “supply treasury into these markets,” though what if a protocol sells NO on itself and then gets exploited? A healthier version could perhaps look like reinsurance, i.e. diversified capital spread across many markets.
The hacker problem?
A reasonable question at this point is, well, can this system be abused by a hacker?
For his part, Scott countered this concern with a good point, namely that “abusing the PM would increase their chance of getting caught” while offering very little compensation compared to what a blackhat would get from just literally running whatever exploit they’re sitting on.
— (@)
Zooming out a little further here, in my recent Papertrade piece I argued that a liquid “house pool” token is a ready-made short for anyone who can drain the house. An exploit YES share is essentially that same short, albeit listed for any protocol.
The hedging pitch here needs these exploit markets to grow big enough to cover real deposits, and at that size, YES could become a second payday for a particularly greedy attacker. As such, the ideal hack market is big enough to be informative but not big enough to be a major attraction.
We’ll have to see how things play out here, yet in the meantime expect more to come, both in terms of liquidity (e.g. Trueo is migrating to Ethereum) and offerings. Scott has said he hopes “we can launch many more [markets] if the reception to these [first ones] is good,” so keep your eyes on this front accordingly.