Personal research project — paper-testing & tuning, not raising capital.

Investment Strategies

One Framework. Three Risk Profiles.

Same deterministic risk engine. Same non-overridable gates. Different allocations optimised for your risk tolerance. All APY figures are variable and not guaranteed.

Lower Risk
Higher Risk
Paper Tracked · evidenced

Conservative

The desk's real book — a live paper track since June 22, 2026

Target profile

~6% target APY

Target band up to 6%. This IS the evidenced paper-tracked book; the realized track-to-date rate is rendered live from the snapshot, never typed. Variable, not guaranteed.

Lower Risk

Tier 1 protocols only. No looping, no leverage. Higher cash buffer. Stricter protocol eligibility. Evidenced paper track since June 22, 2026 — this is the book the desk actually runs.

Who it's for

Allocators prioritising capital preservation over yield optimisation.

Yield sources

Aave V3 USDC Compound V3 USDC Spark sUSDS
View Strategy Details
Research / Paper — tail shown

Balanced

Research book — being rebuilt onto observed rates

Target profile

~5% target APY

Target band. Research/paper only, not live-allocated and not part of the go-live track. Measured 2026-09-09: coverage by observed rates was 0.0% — the book accrued on literals. It is being rebuilt onto observed rates with mark-to-market (ADR-292); until that lands no realized number is claimed here.

Medium Risk

Tier 1 + selected Tier 2 protocols. Deterministic risk-gated rebalancing. Research/paper — the evidenced June 22 track belongs to Conservative, not to this tier.

Who it's for

Family offices and individual allocators seeking systematic, transparent DeFi yield.

Yield sources

Aave V3 USDC Morpho Blue USDC Compound V3 USDC Yearn V3 USDC Euler V2 USDC
View Strategy Details
Research · refused for live

Aggressive

Advanced mechanics for higher yield tolerance

Target profile

~15% target APY

Target profile. Research/paper only, refused for live capital by design. Measured 2026-09-09: coverage by observed rates was 0.0% — the book accrued on literals. Being rebuilt onto observed rates with mark-to-market (ADR-292). Possible loss of capital.

Higher Risk

Levered PT carry, points / incentive farming, and unhedged directional restaking. Real liquidation-cascade and depeg risk — paper-tested in the Aggressive Lab, refused for live capital. The tail is shown, not hidden.

Who it's for

Experienced allocators accepting higher volatility for yield upside.

Yield sources

Optimised Aave positions Advanced Morpho vaults Selected higher-yield pools
View Strategy Details

How new strategies are vetted

Candidates compete in a tournament backtest → paper → live. Only credible-data winners promote — the fail-closed gate refuses on degenerate data. See the live leaderboard and promotion ladder.

Open the tournament →

The tiers, honestly — with the tail

Conservative targets up to 6% (evidenced, fundable; ~5.0% realized (measured 2026-09-20) track-to-date). Balanced (up to 12%) and Aggressive (up to 20%) are research/paper, shown WITH their real dated tail and refused for live capital. The higher the tier, the bigger the tail — measured, not hidden.

Open packages →

APY figures are variable and not guaranteed. Past paper performance does not guarantee future results. Possible loss of capital.

Strategy Lab — research sleeves

What we test beyond the three product tiers

This is a paper backtest bench: research sleeves run through the same deterministic risk engine. The sleeves below are research, advisory, and not live. They are not offered products and are not part of our paper track record.

Honest finding

Crypto (ETH/BTC) yield sleeves are diversifiers, not floor-beaters: the neutral books are low-vol but don't clear the ~3.4% RWA floor; directional books take the full crypto drawdown. Only the stable engines (Engine A/B/C + RWA) clear the floor.

Backtest · advisory · not realized capital
RWA floor (risk-free bench): Backtest window: source: offline
Loading research sleeves from the live feed…

All figures are paper backtest on historical data, variable, not guaranteed, and not investment advice. β(ETH) is the estimated sensitivity to ETH returns. "Killed" means the deterministic risk gate closed the sleeve during the backtest (kill-switch, or fail-closed on missing data).

Personal research project in paper validation — not investment advice, not a regulated service, not raising capital; results are simulated. Full disclaimer & risk disclosure →