Systematic Strategy

ProteanAI
Market-Neutral Systematic Engine

A fully autonomous long/short equity strategy with zero human intervention in the decision loop. From data ingestion to signal generation, conviction scoring, portfolio construction, and trade execution — every step is machine-driven, every decision algorithmically derived.

Overview

End-to-End Machine Intelligence

ProteanAI is not a tool that assists a human trader. It is the trader. The entire pipeline — from raw market data ingestion through signal generation, conviction scoring, regime classification, portfolio construction, risk management, and execution — operates autonomously with no human in the decision loop.

The system continuously ingests data from multiple sources, decomposes market structure across multiple proprietary analytical layers, scores conviction through an ensemble of proprietary signal layers, and constructs a beta-neutral portfolio from a select universe of single-name equities and liquid ETFs spanning fixed income, commodities, and alternatives at varying concentration levels.

A machine-driven beta overlay dynamically manages portfolio-level market exposure, tilting directionally only when the system's own signals confirm — maintaining near-zero average beta across full market cycles without human override.

Why Fully Autonomous Matters

No emotional interference — the system does not hesitate, second-guess, or capitulate. Conviction is mathematically derived and positions are sized accordingly, regardless of narrative, fear, or bureaucratic noise.

Deep signal consensus — conviction requires agreement across independent analytical layers before any position is taken. The machine sees structure a human analyst cannot.

Continuous regime adaptation — the system classifies the prevailing macro environment in real time and adjusts its behavior algorithmically. No committee meetings, no lag between observation and action.

Precision position sizing — no binary in/out decisions. Capital allocation reflects the machine's confidence in each opportunity, naturally concentrating exposure where conviction is strongest.

Architecture

Signal-to-Execution Pipeline

Six autonomous phases. Zero human decision points. Every layer is algorithmically defined, mathematically derived, and systematically executed. The human role is limited to system design and oversight — never trade-level intervention.

Phase 01 — Automated
Macro Context Ingestion
Machine-driven regime classification across an array of proprietary dimensions. Automated geopolitical and policy risk assessment. No analyst interpretation required.
Phase 02 — Automated
Adaptive Signal Generation
Proprietary signal engine scans a curated investable universe continuously, generating raw directional signals through a proprietary evaluation process. No human pattern recognition — purely systematic.
Phase 03 — Automated
Conviction Modification
Raw signals pass through a cross-domain modification layer that integrates positioning intelligence, sentiment analytics, and structural market data. Every adjustment follows defined quantitative rules — no discretionary overlay.
Phase 04 — Automated
Conviction Calibration
A proprietary calibration framework synthesizes macro momentum conditions with derivatives market structure to refine conviction scores and position sizing. Options-derived intelligence informs directional confidence without being traded directly.
Phase 05 — Automated
Portfolio Construction
Proprietary scoring-weighted allocation with dynamic risk calibration. Beta-neutral target with 1x gross leverage. No portfolio manager discretion.
Phase 06 — Automated
Execution & Rebalance
Systematic rebalance cadence with real-time structural monitoring. Full trade audit trail with SHA-256 data provenance. Machine-generated, machine-executed, machine-verified.
Performance

Backtest Results

February 2022 through April 2026. 1,044 trading days across a bear market, recovery, rate cycle, and geopolitical shock regime.

177.1%
Total Return
1.82
Sharpe Ratio
3.02
Sortino Ratio
-9.4%
Max Drawdown
27.9%
CAGR
-0.042
Beta to SPY
54.5%
Win Rate
1.40
Profit Factor

Comparative Performance

Metric ProteanAI SPY Buy & Hold
Total Return 177.1% 63.5%
Sharpe Ratio 1.82 0.44
Max Drawdown -9.4% -22.1%
Sortino Ratio 3.02 0.63
Beta -0.042 1.00

As of April 10, 2026. 207-ticker universe. 5bps round-trip transaction costs, 5.5% annual short financing, 1x gross leverage, weekly rebalance.

Risk

Risk Management

Risk control is embedded at every layer of the pipeline, not bolted on as a post-trade check.

Structural Controls

Beta neutrality — dynamic long/short allocation targets near-zero portfolio beta regardless of individual position performance.

Conviction gating — algorithmic entry and exit criteria govern all position changes. No discretionary override.

Complexity filtering — a specialized market microstructure measure that dynamically attenuates signal strength during periods of extreme trend exhaustion, when reversal probability is elevated.

Leverage discipline — 1x gross maximum. No derivatives leverage or synthetic exposure.

Audit & Provenance

Every signal, trade, and rebalance is recorded with full attribution. Input data is SHA-256 checksummed for provenance verification. The backtest is independently verifiable through thousands of trade records with no-lookahead confirmation.

Terms

Strategy Terms

Strategy
ProteanAI
Benchmark
S&P 500 (SPX)
Manager
Autonomous Systematic Engine ~16,000 lines of proprietary Python across signal generation, conviction scoring, portfolio construction, and execution infrastructure. Zero human decision points.
Management Fee
2% of AUM
Performance Fee
20% of profits Subject to high watermark. Performance fees are only charged on new profits — if the fund declines, prior losses must be fully recovered before performance fees resume.
Target Sharpe
> 1.0
Target Sortino
> 1.5
Roadmap

Platform Roadmap

ProteanAI is designed to extend into derivatives and volatility strategies as the fund scales — applying the same signal architecture to convexity instruments, cross-asset macro, and structured positions.

Contact

Learn More

For detailed documentation or to discuss ProteanAI's prospective integration into a portfolio allocation, contact the team directly.

investors@proteanasset.com