Update (August 2026): the vault has since been redeployed. The address below is superseded — the current contract on Base is
0x36C81d7E1966310F305eA637e761Cf77F90852f0. The original post is left untouched; a four-month follow-up is at the end.
I recently deployed a PCV Vault on Base (0x38DcDB3A381677239BBc652aed9811F2f8496345).
We built a highly optimized atomic internal router (buyGBLIN) to mint the token exactly at NAV, bypassing the intentionally thin Aerodrome DEX liquidity. My architectural assumption was that MEV searchers would aggressively arbitrage the spread between the DEX spot price and our internal quoteBuyGBLIN view function to keep the peg.
However, watching the current L2 meta on Base, it seems the ecosystem is almost entirely monopolized by off-chain Solvers (CowSwap, UniswapX, 1inch Fusion).
My question for the Research collective: Is building native internal routing for PCV vaults practically obsolete on L2s now? Are active searchers completely ignoring direct smart-contract arbitrage opportunities unless they are bundled through an intent-based solver network?
I’d love to hear how you are currently evaluating these on-chain spreads versus just plugging into solver APIs. If the meta has shifted completely to intents, we might need to rethink the gas overhead of native routers.
Four months on (August 2026). Answering my own question with what we measured since: direct contract-level arbitrage against our internal spread stayed rare. Flow kept routing through solvers, and the peg was held mostly by our own rebalance path rather than by external searchers.
We kept the native router anyway, for a reason that was not obvious to me when I posted: it is not only an arbitrage surface, it is the mint path at NAV. Its value turned out to be deterministic issuance, not the searcher competition I was expecting.