Solana stake pools weigh yield against punishing sandwichers

Today, enjoy the Lightspeed newsletter on Blockworks.co. Tomorrow, get the news delivered directly to your inbox. Subscribe to the Lightspeed newsletter.

The Solana validator world is an interesting tangled web right now.

For starters, tips and fees on Solana transactions have declined along with the price of SOL from pretty stellar levels over the summer. As we covered last week, smaller validators are particularly hard-hit by this, especially since the Solana Foundation lowered its commission caps for validators receiving stake from its delegation program.

For those smaller validators, and all Solana validators in general, the entire ballgame is drawing more delegated stake. Profits scale with stake for validators — who are responsible for verifying Solana blocks — while costs remain roughly fixed regardless of size. Validators can grow their stake by marketing themselves directly to stakers, or they can try to draw delegation from stake pools, which combine staked SOL and dole it out to multiple validators.

The second-largest stake pool, however, just made this harder to do.

Marinade recently introduced a “stake auction marketplace” (SAM) that delegates its stake based on which validator bids the most SOL. This has boosted mSOLs native yield to 8.1%, higher than any stake pool with more than two staked validators, according to Solana Compass.

Some validators have said that the only validators able to afford Marinade stake are ones who are participating in unsavory behavior such as sandwich attacks, which are exploitative trades at the expense of unsophisticated users. DeezNode, which is widely thought to participate in sandwiching, draws a significant stake from Marinade, per Solana Compass.

Jito and the Solana Foundation have both taken aim at sandwiching, which is a toxic form of what‘s called maximal extractible value (MEV), but there doesn’t seem to be a silver bullet for the problem. I spoke to Marinade core contributor Michael Repetny to try understanding a bit more of the ethos behind SAM.

Repetny admitted that validators participating in its marketplace could “quite possibly” be sandwiching users, but added that a majority of Solana stake runs on an MEV client, and its impossible to fully prevent sandwiching without turning off MEV altogether. He added that he favors sandwiching being solved at the Solana core level, rather than at the validator level.

Repetny also argued that SAM is good for the Solana network because the rules are clear, and better yield could cause stakers to move away from large exchanges like Coinbase, which runs the third-largest validator.

In the end, Repetny‘s view seemed to be one of pragmatism — if you can’t solve the negative externalities of MEV, then you can at least pass the yield back to stakers.

The amount of yield given to stakers is proportionally up by quite a bit this year, actually. Stakers were taking home around 1% of transaction fees and tips this week last year, according to Blockworks Research. Today, stakers net 29% of fees and tips.

Not everyone agrees with Repetnys approach, however. Yesterday, the Solana liquid staking project AeroSOL announced a stake pool called Aeropool, with a likely jab at Marinade.

“Modern delegation strategies often prioritize achieving the highest APY, resulting in a concentration of stake among a select few validators across various pools,” which contradicts Aeros ethos, the project wrote on X.

Instead, Aeropool would delegate its 150,000 SOL to empower positive Solana contributors and attract new validators that could decentralize the validator set. In essence, Aeropool is willing to sacrifice some yield in exchange for rewarding desirable behavior. Repetny would likely argue that this is an impossible task.

It will be interesting to find out what the market thinks.