verifiable track recordpaper + live-testnet

Performance

Cumulative return of Sonar's rebalanced book versus a buy-and-hold baseline of the same index universe, both indexed to 100 at inception. Computed from the dated theses, their target weights, and the on-chain NAV snapshots Sonar persists each cycle. This is paper plus testnet performance during the buildathon, not a live-money track record. The credibility is the verifiability: every number traces to a cited thesis.

Sonar book

+17.95%

Buy and hold

+14.22%

Excess return

+3.73%

Win rate

52%

114 trade / 25 no-trade

Cumulative return vs buy-and-hold

Gold is Sonar's rebalanced book; the dashed line is a static hold of the same indices from inception. Both start at 100.

Per-thesis attribution

Each thesis links to the run that produced it. Cycle return is the book move over the cycle the thesis governed; notional is the USD the cycle executed, summed from one ledger (paper_trades) so a live fill is counted once. It is dominated by the simulated SSI legs, which are sized against the book, not by the venue-capped perp hedges.

Thesis	Mode	Notional	Cycle return	Result
2026-10-17 04:00Z	trade	$15,367	-	-
2026-10-16 04:00Z	trade	$27,357	-0.41%	loss
2026-10-15 04:00Z	trade	$41,930	+1.34%	win
2026-09-26 04:00Z	trade	$10,718	-9.46%	loss
2026-09-19 04:00Z	trade	$23,293	+4.67%	win
2026-09-18 04:00Z	trade	$33,647	+5.07%	win
2026-09-17 04:00Z	trade	$18,451	+3.36%	win
2026-09-16 04:00Z	trade	$19,651	+1.48%	win
2026-09-15 04:00Z	trade	$30,214	-1.96%	loss
2026-09-12 04:00Z	trade	$40,673	+0.50%	win
2026-09-11 04:00Z	trade	$35,446	+0.41%	win
2026-09-10 04:00Z	trade	$6,656	-1.30%	loss
2026-09-09 04:00Z	trade	$53,739	-1.05%	loss
2026-09-08 04:00Z	no-trade	-	+0.00%	-
2026-09-05 04:00Z	trade	$35,736	+0.84%	win
2026-09-04 04:00Z	trade	$19,498	-1.82%	loss
2026-09-03 04:00Z	trade	$18,290	+3.65%	win
2026-09-02 04:00Z	trade	$29,827	-0.42%	loss
2026-09-01 04:00Z	trade	$53,540	-1.56%	loss
2026-08-29 04:00Z	trade	$41,386	+0.91%	win

Delta-neutral book

rules-based carry

The second strategy's mark-to-market track: long the MAG7 basket, short a BTC perp sized to the basket's BTC weight. The short offsets the dominant BTC beta, so net exposure falls toward the smaller non-BTC residual while the book harvests the index-versus-perp basis.

Long$173.82Short$2.4kNet$-2.2kUnrealized$-369.75Rebalances64

Net exposure over time (long minus the BTC short, USD)