PUBLIC-PROXY RESEARCHv0.9.0

Pay a little.
Survive a lot.

A transparent laboratory for testing permanent, deep-out-of-the-money S&P 500 put protection—measured where it matters: the complete portfolio.

Run the retail backtest →

✳

Not the Universa algorithm. Public, inspectable strategy proxies only.

196 tests passing

27 strictly typed modules

2013—26 OPRA coverage

0 silent substitutions

SYNTHETIC ACCOUNTING TEST

Same hedge. Two regimes.

Convexity changes the ending

Complete portfolio value

SYNTHETIC · NOT HISTORICAL

The same transparent long-put proxy monetizes three times as the synthetic shock deepens, preserving capital while retaining protection.

Hedged return+57.50%

Unhedged return−15.98%

Hedged max drawdown−19.91%

Unhedged max drawdown−34.79%

Gross premium purchased$2,330

Option sale proceeds$75,057

Ask entry · bid exit · $5 commission · integer contracts · no leverage

STRATEGY SANDBOX

Set the rules.
Never predict the crash.

Configure a research proxy. These controls set the mandate—not a performance forecast.

Target put delta-0.0050

Signed put delta. 0.50Δ is not the target.

Target expiry60 days

45D120D

Monthly premium cap0.50% NAV

0.25%1.00%

MonetizationResearch proxy

EVIDENCE, NOT NARRATIVE

Built to fail honestly.

The research system preserves unfavourable results, blocks future information, and refuses to invent a quote when the historical record is incomplete.

01 / INPUTS

Fail-closed evidence

Missing quotes remain missing. Every proxy is labelled, timestamped, and retained for audit.

no silent substitution

02 / TIME

No look-ahead

Definitions, quotes, underlying observations, and rates must all exist at or before the decision.

timestamp ≤ 15:45 ET

03 / EXECUTION

Pay the spread

Buy at the historical ask. Mark and sell at the bid. Charge commission on every contract.

cash ≥ 0 · leverage = false

04 / NEGATIVE RESULT

T4 did not earn promotion.

Across 985 identical untouched-test dates, all five option-ranking selectors underperformed fixed strike/DTE controls. The permanent hedge stayed in place.

Best selector 0.367Fixed controls 0.567—0.793

RESEARCH PIPELINE

Ready for the real pilot.

AUDITED · 26 AUG 2026

01COMPLETE

Metadata & parent resolution

OPRA.PILLAR · definition + cbbo-1m · SPX / SPXW / XSP

442,600 records resolved

02READY

COVID pilot definitions

Four point-in-time roll dates · raw DBN + normalized Parquet

4 dates fixed

03READY

Candidate & parity quote plan

Broad surface once · exact raw symbols · no repeated full chains

Inputs locked

04COMPLETE

Historical T1—T4 evaluation

Fixed controls · chronological models · failed selectors retained

14.6m option rows

05LIVE

Retail barbell backtester

$100k portfolio · finite $10k sleeve · whole-contract SPX puts

Open T5A →