thanks a lot ser. with this report - just kept writing about all the things I wanted to understand in the prediction markets space. Followed that up by getting feedback from friends and AI about what ideas had been conveyed effectively versus where we can improve upon.

Agree that Melee is very interesting. you get a nice mix of prediction markets and pump like mechanics that rewards early participants on the bonding curve.

Lightcone is also cool. They're trying to bring the multiverse finance idea to reality (https://www.paradigm.xyz/2025/05/multiverse-finance)

The features that are available don't seem to be that radically different between the 2 but Hyperliquid definitely looks to be the more performant of the two in terms of throughput and latency (expected given the differential in the validator count).

Each validator on the dYdX Chain runs an in-memory order book locally that is not committed to the chain (i.e. each validator is not expected to have the same version of orderbook) which is interesting but also required dydx to work on some anti-MEV measures. Whereas with HL all validators maintain the same copy of the orderbook.

The defining feature for both to attract market makers was priority for cancel orders over the matching orders.

Megavault (27.6mn) is relatively small compared to HLP (297mn). Both vaults have seen massive outflows of capital over the last 3 months HLP down from 540mn and Megavault down from 80mn.


Both receive a portion of the trading fees from the protocols along with the PNL from trading. It’s difficult to benchmark the performance of these vaults against each other as the TVL deposited in both has been constantly changing over the last few months.



But Megavault estimates annual pay based on the previous months returns annualised (which is -7% right now) & using the same methodology HLP estimate would be (-9.2% - No thanks to the Hyperliquid whale)



I think the vault that outperforms is the one with more trading volumes over a long enough time frame. The protocol with more trading volume/TVL ratio gets to earn more from trading fees with the trading returns across both platforms converging over time unless the MMing strategy is significantly different (which I doubt it is).

Based post

Not exactly - I think both of those are somewhat independent from each other.

The capital rotation from SUI to SUI alts is a matter of when in my opinion. Best guess is when the next downtrend in bitcoin dominance.

But I definitely think that once a killer application is built and that alt offers major gains to users who invest early - we will see people with large unrealized SUI gains venturing into other ecosystem alts to catch that next token

It's not live yet. The team is still in the process of bringing on existing Solana validators to run the bot from day one of launch.

https://x.com/paladin_solana/status/1833213679321428323

The main hurdle for adoption is getting a sizable Solana stake to adopt the Paladin solution. After that the bot would need to prove its ability to capture Atomic and CEX-DEX arbitrage

I also think continued development funded via the evergreen fund can also become a hurdle if not executed well

Any Validator participating in the Paladin system - would ideally not engage in MEV practices like Sandwich attacks or front-running.

If caught doing so, their locked PAL tokens can be slashed heavily. These validators only capture MEV or value from backrunning arbitrage, liquidations (usually considered good MEV).

Paladin Bot also offers swappers a chance to improve order execution over Jupiter.

It is higher than the comparable pools on Aave, but for many of these top pools (like the LRT pools) the comparable just didn't exist on Aave up until very recently.

However, I think it is not exactly a fair comparison since the ideal utilization ratio and interest rate models are different in the Aave and Morpho pools. Hence, I compared the protocol-wide utilization rates rather than doing so at the pool level

Currently, the graphs are limited to the past couple of years of data. I agree that it would be much better to have the optionality of looking at the data from the beginning of it being recorded rather than being limited by the data source used

Not a deep dive on Gammaswap in particular but the team had covered the design of Infinity pools - https://members.delphidigital.io/reports/are-infinity-pools-the-next-big-speculation-primitive#infinity-pools-eli10

Gammaswap, Infinity pools, Panoptic, Smilee finance all have a similar mechanism utilising AMM LP positions for creating derivative like payoff s

yes, depositors cannot be expected to rebalance their assets between pools based on utilisation and demand for individual pools.

In protocols like Aave the aggregation happens due to 2 reasons:

1. The multi-asset pool design
2. yield aggregators like yearn

For protocols like Ajna, Silo, and Morpho where each pool is isolated - to efficiently allocate assets like ETH, stablecoins across pools this process needs to be abstracted away from the end user

The Euler team just released their vault kit codebase last week
https://twitter.com/eulerfinance/status/1783145801289519176

Once the audits and bug bounty program is done, we can expect the launch to quickly follow up.

But no exact launch date released yet - But seems like the launch will happen this quarter.