A systematic Solana-native fund applying institutional risk frameworks to early-stage liquidity venues. We extract structural alpha from launchpad mechanics, post-graduation order flow, and cross-venue arbitrage — with the discipline of a multi-strategy desk.
Each strategy targets a distinct inefficiency. Combined exposure is constrained by a portfolio-level risk budget that decorrelates strategy outcomes from underlying SOL beta.
Systematic capture of price discovery inefficiencies in the first 4-hour window following bonding curve graduation. Edge derived from order book imbalance, holder distribution telemetry, and venue-spread arbitrage.
Mean-reversion across PumpSwap, Raydium, Meteora, and Jupiter routing layers. Exploits transient mispricings caused by router latency and asymmetric LP positioning across AMM curves.
Systematic short-vol exposure on liquid Solana memecoin perps where implied funding consistently overshoots realized variance. Risk-managed via dynamic deltas and per-name max-loss circuit breakers.
Directional SOL and SOL-correlated basket exposure during high-confidence BTC→ETH→SOL rotation windows. Filtered by cross-asset momentum, dealer positioning, and on-chain liquidity inflow signals.
Continuous-cost tail hedging via OTM put spreads on SOL and ETH, sized as a function of book gross. Designed to monetize during regime-shift events that disable other strategies' liquidity assumptions.
Tokenized money market and on-chain T-bill exposure for unallocated capital. Maintains operational liquidity for opportunistic deployment and same-day strategy rebalances without forced position exits.
Selected notes on Solana market microstructure, launchpad mechanics, and the broader on-chain liquidity landscape. Long-form work intended for institutional readers.
Pump.fun's bonding-curve-to-AMM graduation is functionally an auction with deterministic clearing rules. We argue it should be modeled as a structured price-discovery event with predictable post-clearing volatility decay — and walk through the empirical signature across 14,000 graduated tokens.
Allocations are dynamic, rebalanced weekly against a risk budget rather than a target exposure. Each strategy contributes a constrained share of portfolio variance, capped by historical drawdown profile.
Capital weighting fails in regimes where strategy correlations spike. We instead allocate by ex-ante variance contribution, constrained so that no single strategy contributes more than 30% of expected portfolio variance under normal conditions.
Each strategy carries a capacity ceiling defined by the venue depth it trades against. We do not scale post-graduation flow capture beyond the level where our own order flow would meaningfully move the price we are trying to capture.
Strategy-level stop logic is automated. A strategy hitting its 30-day rolling drawdown threshold is reduced to 25% allocation pending desk review. Re-engagement requires explicit governance sign-off, not algorithmic recovery.
| Period | Strategy Return | Benchmark (SOL) | Spread | Sharpe | Max DD |
|---|---|---|---|---|---|
| — Q1 (illustrative) | — | — | — | — | — |
| — Q2 (illustrative) | — | — | — | — | — |
| — Q3 (illustrative) | — | — | — | — | — |
| — Q4 (illustrative) | — | — | — | — | — |
| — YTD | — | — | — | — | — |
Our risk framework is structurally separated from the trading desk. Limits are defined ex-ante and enforced algorithmically. The framework below describes design targets, not realized historical performance.
Team positions intentionally left as placeholders. A real fund would list real people here, vetted and disclosed under applicable jurisdictional requirements.