Our approach
One portfolio. More than one perspective.
A trading algorithm follows a set of rules. A portfolio brings multiple systems together—with deliberate decisions about how much each one can trade.
Start with the basics
An algorithm can identify an opportunity, size a position and decide when to exit. Automation makes a rule repeatable; it does not make the rule profitable or remove the need for supervision.
A portfolio combines approaches that may respond differently to the same market. The useful question is how those approaches behave together, especially when conditions become difficult.
From a collection to a portfolio
- Understand each approach. Identify its rules, market exposure, holding periods and assumptions.
- Study the relationships. Compare overlapping trades, shared losses and performance across different conditions.
- Allocate deliberately. Decide how much exposure each component receives and set limits at the account level.
- Test and monitor. Evaluate the combined result, document changes and investigate deviations from the plan.
Why combine algorithms?
A system that thrives in a sustained trend may struggle when prices reverse repeatedly. A different approach may behave differently in that environment. Combining them can spread dependence across ideas—but the benefit must be measured, and correlations can change.
More algorithms can also mean more overlapping exposure, trading costs and operational complexity. Counting strategies is not a substitute for evaluating the portfolio.
And what about buying an index?
Buying and holding an equity index is a useful comparison, with its own exposure and cost structure. An algorithmic portfolio should earn its complexity through evidence. We would compare the portfolio and an appropriate S&P 500 or Nasdaq benchmark over matching periods, including losses, costs, leverage and the time needed to recover.
We are not claiming that every algorithmic portfolio outperforms an index. Our v3 comparison will require a completed, reproducible test.
What a fair comparison needs ↗Questions worth asking
Do I need to have traded an algorithm before?
No. Begin with the rules, what can go wrong and what supervision is needed. The Matchmaker explains its terminology as you go.
Can I see the names of every constituent algorithm?
Our portfolios combine proprietary and licensed systems. Individual constituent identities are confidential. Portfolio documentation should instead explain aggregate exposures, construction, limits, version history and the evidence available for that version.
Will the portfolio change with market conditions?
Condition-aware allocation is part of the development direction. The v3 specification will explain any implemented weighting and rebalancing rules once they have been established and tested.
Does a color describe a portfolio’s risk?
A Matchmaker color summarizes stated preferences. It is not a measured risk rating, a return forecast or a financial suitability assessment. Portfolio mappings will need their own evidence.
The Algo Matchmaker / About 5 minutes
Find your trading style.
Know what to explore.
Answer ten questions to discover your color profile, the trading approaches that align with your preferences, and where to take your research next.
Find Your Algo Match ↗Free · No experience needed · Email optionalA guide to your preferences, not a prediction of returns.
Let’s talk
What would you like to explore next?
Questions about our research, portfolios or your next step? Get in touch.