Systematic Strategies for Digital Asset Markets
We research, build and operate directional and market-neutral trading strategies, and the platforms that consolidate and monitor them: built in-house, validated out of sample, run around the clock.
Trading Strategies
Four independent strategies, each built for the structure of its market and each with its own source of return
OPUS ALPHA
Proprietary machine-learning models trading the most liquid crypto perpetual futures.
- Separate long and short models, trained on a broad set of price, volume and volatility features
- Exits and position sizes scaled to each asset’s own volatility
- Protective orders rest on the exchange; cost-aware execution on every entry
- Pre-registered research, held-out validation and regular retraining; live signals continuously checked against research
OPUS DELTA
A dollar-neutral long/short portfolio across the most liquid crypto perpetual futures, built from proprietary systematic factors.
- Multiple low-correlation factors combined into one diversified book
- Systematic rebalancing with per-asset, liquidity and net-exposure limits
- Proprietary execution algorithms designed to minimise market impact
- Models go live only after passing held-out validation with statistical confidence tests; every research trial is recorded
OPUS RHO
Cross-exchange funding-rate arbitrage: long and short the same asset on two venues, collecting the funding spread.
- Dollar-neutral by construction, so returns come from the funding differential, not price direction
- Continuous funding-rate monitoring across the major derivatives venues, with periodic re-selection of the best spreads
- Coordinated two-leg execution with leg-balance checks and automatic recovery on partial fills
- Independent position reconciliation against each venue’s own records
OPUS OMEGA
Systematic trading on Polymarket across six verticals, priced exclusively from the live order book.
- Six verticals: crypto, sports, weather, politics, multi-outcome and cross-market
- Low-latency infrastructure: dedicated order-book feeds and native order signing
- Independent data sources per vertical: exchange prices, weather models, sports and polling feeds
- Every fill and settlement reconciled against on-chain records
Platform Services
Cross-platform services that consolidate, monitor and report across everything we run
OPUS ATLAS
One consolidated view of a family’s entire balance sheet: banks, custodians, private assets and liabilities.
- Statements from any bank or custodian are read automatically and checked against their own totals, security identifiers and FX rates
- Private assets (real estate, private companies, funds, collectibles, digital assets) and liabilities (mortgages, Lombard loans, credit lines)
- Performance rebuilt from independent market data, in any reporting currency
- Risk analytics: VaR, drawdown, correlations, liquidity profile and historical stress scenarios
- Each client’s data held in its own isolated store; read-only, with no orders and no custody
OPUS HERMES
A private AI assistant over the whole platform: every strategy, fund and wealth view, by voice or text.
- One live view across all strategies, the tokenized fund NAV and consolidated wealth
- Natural-language questions and commands in English and Italian
- Private by design: language models run on-premise, so data never leaves the operator’s infrastructure
- Read-only and keyless: it cannot place trades or move funds; speaker verification, passkeys and an append-only audit log
- Next: daily briefings, anomaly alerts and capital-allocation monitoring
How We Build
One research discipline and one shared infrastructure behind every strategy
Signal Generation
A dedicated signal engine for each market structure: proprietary machine-learning models for directional trading, systematic factors for market-neutral books, funding-rate differentials for carry, and order-book pricing for prediction markets.
Research & Models
Machine-learning and factor models built on a research process designed to resist overfitting: pre-registered hypotheses, results averaged across many random seeds, every trial recorded, and data held back from tuning for the final verdict.
Risk Management
Volatility-scaled sizing, per-asset and net-exposure limits, daily loss halts and exchange-resident protective orders. One-click flatten on every strategy, and live edge monitoring that flags a model whose performance drifts from its validated profile.
Validation
Walk-forward testing, held-out periods and sealed lockboxes, block-bootstrap confidence intervals and stress tests. Live signals are checked against the research engine, so the model in production is exactly the model that was tested.
Execution Infrastructure
Maker-first and liquidity-aware execution algorithms, coordinated multi-leg orders with partial-fill recovery, and continuous reconciliation of every order, fill and fee against the venue’s own records.
Operational Resilience
Supervised processes that restart automatically and survive reboots, safety checks before every restart with open positions, and automated test suites that gate every release. Full audit trail of every configuration change.
Infrastructure
Multi-venue connectivity, multi-region deployment, security in depth
Connected venues: Binance, Bybit, OKX, Bitget, Coinbase, Kraken, Deribit, Hyperliquid, Aster and Polymarket
Real-time market and account streams with automatic reconnection, on a private encrypted network with automated backups
Isolated databases per account and per client, with a full audit trail and continuous position reconciliation
Latency-optimised hosting close to the venues: prediction-market execution in London alongside Polymarket’s infrastructure, digital-asset strategies in Singapore near the major Asian exchanges
AES-256-GCM encrypted credential vault; trade-only API keys, never withdrawal rights
Automated daily, weekly and monthly performance reports with full P&L attribution
Passkey (WebAuthn) and Argon2id password authentication with JWT session tokens and rate-limited login
TLS 1.3 transport with HSTS, automated certificate renewal and a full security-header policy
Network-layer firewall, intrusion detection with automated IP banning and continuous security patching
Built by Traders,
Run Like a Desk
Two decades of market experience behind every system we operate
We trade the systems we build. Strategy research, execution, risk controls and reporting are engineered in-house and tailored to each market structure, with no white-label trading rails. An AI-augmented engineering process, held to strict testing and review standards, lets us move from research to production quickly without lowering the bar.
From directional and factor strategies to funding arbitrage, prediction markets and wealth consolidation, every product answers real trading and reporting requirements, and every position and fill is reconciled against the venue’s own records.
Get in Touch
For family offices, proprietary trading firms and funds: strategy partnerships, managed accounts and platform access
Contact Us