What happened this month.
May was a quietly profitable month. We took advantage of three separate basis dislocations — one driven by ETH ETF outflows on the 9th, one by DAI's PSM rebalance on the 16th, and a slow CEX-DEX spread that re-emerged once funding flipped negative on Hyperliquid late in the month. None of it was dramatic, which is the point.
Where the return came from
Perp basis contributed +1.32% (gross), which is roughly in line with our 12-month average. The CEX-DEX latency book sat dormant for the first two weeks — spreads simply weren't there — but came alive once funding rates inverted on May 22nd. It clipped +0.46% across nine trades, all of which closed inside their target half-life of 4 hours. The DAI/USDC peg arb produced +0.41% from a single sustained event around the PSM rebalance.
The v4 LP rebalancer book had a soft month — +0.18% gross, after accounting for IL on three out-of-range positions we let drift. I think the regime favored full-range over concentrated this month, and we leaned the wrong way for about 8 days. Adjusted on the 24th.
Mistakes and what I'm changing
Two things worth flagging. First, the tail hedge book cost us −0.12% in put theta with no triggering event — the standard cost of carry, but worth naming. Second, on the morning of May 18th our solver routed a $4.2M stable arb through a thinner-than-ideal Curve pool; slippage was 7 bps higher than expected. Net impact small (~$2,900), but I've tightened the venue allowlist for trades over $3M and we'll re-audit Curve's depth feeds in June. The fix is already deployed (commit 4a8f1e2).
— kira.eth · published from block 23,418,221