Systematic Liquid Alternatives

Rules-based research in continuously traded markets

Signal research, validation and ongoing model monitoring applied to the most liquid digital assets — with the discipline that systematic investing requires and the humility that model-driven approaches demand.

Capability overview

Systematic Digital Asset Strategies

Digital assets trade continuously, across many venues, with a participant base that differs markedly from established markets. Those conditions produce recurring patterns in price behaviour that a rules-based process can be designed to identify.

Systematic implementation removes discretionary bias from day-to-day decisions. It does not remove risk. Models can degrade, market regimes can change faster than a research cycle, and a rule that held historically may simply stop working. Our governance framework is built around that possibility rather than around confidence that it will not occur.

Opportunity set

Continuous, fragmented markets

Markets that never close, distributed across venues with differing liquidity, produce microstructure behaviour that differs from markets with defined sessions and a dominant venue.

A distinctive participant mix

The balance of participants in digital assets differs from established markets, and that composition shapes how prices respond to flow and to stress.

Liquidity concentrated in few assets

A small number of digital assets account for the substantial majority of liquidity. Concentrating research there prioritises tradability and reduces reliance on thin markets.

Our approach

Assessment of signal research

Signal research is assessed on quality, stability, drawdown behaviour and execution robustness — not on historical return alone.

Guarding against overfitting

A model that fits history closely is easy to produce and frequently worthless. Validation emphasises out-of-sample behaviour and resistance to small changes in assumptions.

Machine learning, and what governs it

Where machine-learning techniques are used in signal research and validation, what matters is the governance around them and the human oversight that remains. They are a method of research, never a guarantee of performance.

Complementary model families

Approaches operating at different horizons and drawing on different signal architectures are combined so that the outcome depends less on any single model continuing to work.

Research framework

What we study when assessing this area.

Signal quality, and whether apparent edge survives realistic execution assumptions
Stability of behaviour across differing market regimes
Drawdown characteristics, including depth, duration and recovery
Sensitivity to transaction costs, slippage and available liquidity
Out-of-sample and walk-forward evidence rather than in-sample fit
Correlation between model families, and whether it is stable
Indicators of model decay and the thresholds that trigger review

Model governance

Model validation

Models are assessed before deployment against defined criteria, with explicit attention to the risk that historical fit does not reflect a genuine, repeatable effect.

Ongoing model monitoring

Live behaviour is compared against research expectations on a continuing basis. Divergence is treated as a signal to investigate rather than as noise to be tolerated.

Human oversight

Systematic does not mean unattended. Defined oversight governs model behaviour, exposure and the circumstances in which a model is reduced or withdrawn.

Implementation principles

Conceptual principles. Product-specific portfolio information is restricted.

Rules-based position sizing

Exposure is determined by predefined rules rather than by discretionary conviction, and adjusted according to the same rules as conditions change.

Defined exposure limits

Leverage, concentration and aggregate exposure operate within limits set in advance.

Execution discipline

Implementation prioritises the most liquid segments of the market, because a signal that cannot be executed at reasonable cost is not an investable signal.

Risk considerations

Capital is at risk. Investors may lose part or all of the amount invested.
Strategies in this capability may be implemented through Contracts for Difference (CFDs). CFDs are leveraged derivative instruments and carry significant risk, including losses that may exceed the initial margin.
Implementation through derivatives does not involve direct ownership of the underlying digital assets.
Digital asset markets are highly volatile and can move sharply and without warning.
Systematic models may degrade, may fail to adapt to a change in market regime, or may underperform for extended periods.
Historical model research, including out-of-sample testing, does not ensure future results.
Past performance is not a reliable indicator of future results.

Institutional considerations

What an allocator should understand when evaluating this type of strategy.

Ask what has been discarded

A systematic manager's research discipline is better evidenced by the models they rejected and why than by the ones they kept.

Compare live behaviour with research expectation

Divergence between researched and realised behaviour is one of the most informative diagnostics available when assessing a systematic approach.

Understand the leverage and the instrument

Where implementation uses leveraged derivatives, the instrument itself is a material part of the risk and should be assessed alongside the strategy.

Establish capacity honestly

Systematic approaches that depend on short-horizon effects frequently have capacity limits, and those limits should be understood before allocating.

Related research

Product-specific materials

Product-specific materials are available to eligible investors through controlled investor access.

Eligible Investor Access