From Principle to Practice
Ideas become useful only when they survive contact with implementation.
The earlier Compass routes establish the assumptions, the evidence and the environment. Practice is where those ideas become machinery.
A systematic trend programme needs entry rules, exits, position sizing, portfolio construction, diversification, data, rollover, execution and a disciplined operating process. None of those components is especially intelligent on its own.
The intelligence is in the architecture that connects them.
Practice is not the search for the perfect rule. It is the construction of a programme whose deliberately limited components work together well enough to survive uncertainty, remain exposed to opportunity and execute what the philosophy requires.
System Architecture
Where does the intelligence of a systematic programme actually live?
Not in a single indicator. Not in an entry rule that somehow knows what comes next. Not in an exit that identifies the top. Not in a sizing formula capable of measuring everything we mean by risk.
A robust programme distributes the problem. Each component has a limited responsibility, and the architecture determines how those limited components interact.
-
The System Anatomy Series: Introduction
The gateway to the complete architecture: entry, exit, sizing, portfolio construction, ensembles, rollover, execution and process.
-
System Anatomy 1: Entry Logic
The first component opened. Entry does not forecast the trend. It detects a condition and acts.
-
The Paradox of Simplicity
Why apparently unsophisticated rules can be more robust than complicated machinery fitted too closely to the past.
Entries & Exits
How do we participate in a trend we cannot identify in advance, and how do we leave one whose endpoint we cannot know?
The entry has a narrow responsibility: detect the condition and act. False starts are not necessarily evidence of failure. They are part of the cost of being present before a genuine trend has become obvious.
The exit solves a different problem. Failure must be bounded, while success must be given room to remain open. The asymmetry matters. A trend follower does not need to know how far a move will travel in order to keep following it.
The practical principle: enter without requiring certainty, leave when the predefined condition says to leave, and resist the temptation to convert an open-ended payoff into a predetermined profit target.
Position Sizing
How much capital should we expose while waiting for something we cannot predict?
Position sizing translates a trading signal into an economic exposure. It normalises different markets, limits the consequence of any single position and helps determine whether the programme survives the sequence of losses that inevitably precedes some of its largest opportunities.
The objective is not to discover the mathematically perfect position. It is to find one the programme can survive.
-
Position Sizing: The Geometry of Survival
A deeper treatment of sizing as a survival problem rather than a search for maximum short-run return.
-
Dynamic Position Sizing
The implications of changing exposure dynamically, viewed through the requirements of an Outlier Hunting programme.
-
The Cut Back Rule
A defensive mechanism for reducing exposure when closed equity breaches its survival threshold.
-
Position Sizing and Outlier Hunting
Why sizing has to respect fat-tailed outcomes and the asymmetry between ordinary trades and the rare moves that carry the programme.
Portfolio Construction & Ensembles
What happens when fifty individually sensible positions become one portfolio?
Markets are coupled. Correlations change. Several positions can become different expressions of the same underlying event. Portfolio construction therefore cannot be reduced to counting instruments.
For an Outlier Hunter, breadth has another purpose as well. We do not know where the next consequential trend will appear, so the portfolio must create many meaningfully different ways of encountering what the system does next.
-
Casting a Wider Net
Why broad opportunity sets matter when the location of future outliers cannot be known beforehand.
-
Diversification for Trend Following Models
How small variations across systems and markets can create useful differences at portfolio level.
-
From Hurricanes to Financial Markets
An exploration of ensemble thinking and why multiple imperfect models can be preferable to dependence on a single answer.
Rollover & Data
What happens when the market we trade is not the same thing as the historical series we use to represent it?
Futures contracts expire. A long-run history has to be assembled from a succession of individual contracts. That construction process can introduce artificial jumps, altered relationships and information that no trader actually experienced.
The system cannot distinguish genuine market information from an artefact introduced by its own data. The practitioner has to do that work.
Rollover is not administration. It is part of system design. The quality of the historical representation affects entries, exits, sizing, research and every conclusion built downstream from the data.
The full treatment forms part of The System Anatomy Series, alongside the other components of the programme.
Execution & Process
What good is an architecture that disappears the moment the real market becomes uncomfortable?
A backtest executes every rule without hesitation. A human operator does not. Slippage, gaps, missed orders, operational errors, drawdowns and the urge to override the system all appear only when the programme meets reality.
Execution is therefore more than obtaining a fill. It includes maintaining the integrity of the process itself.
-
The Hidden Costs of Certainty
Why the desire to feel comfortable can interfere with the very exposures a systematic process was designed to hold.
-
Trading as a Natural Selection Process
Why survival and adaptability matter more than requiring the market to conform to a fixed forecast.
-
Explore Practice in The Vault
Search the wider ATS archive for system design, implementation, portfolio construction, execution and related practical work.
The System Anatomy Series
System Anatomy is the backbone of this route because it examines one integrated programme from the inside. Seven components are separated for explanation, but the argument is ultimately about how they depend upon one another.
Begin here: The Programme, From the Inside →
-
System Anatomy 1: Entry Logic
The entry cannot know whether a trend is beginning. Its job is to detect the condition and act.
-
System Anatomy 2: Exit Logic
Failure is bounded. Success is left open. The exit follows rather than predicts.
-
System Anatomy 3: Position Sizing
How much capital the programme exposes while waiting for an unknowable sequence of outcomes.
-
System Anatomy 4: Portfolio Construction
How many positions become one portfolio without pretending the markets are independent.
-
System Anatomy 5: The Ensemble Approach
Multiple deliberately limited systems create a broader response than any single rule can provide.
-
System Anatomy 6: Rollover Mechanics
The construction of the historical record and why representation becomes part of the trading machinery.
-
System Anatomy 7: Execution and Process
The architecture only exists in the real world if the trader actually executes it.
The system is not an entry rule with some supporting machinery attached. It is the machinery, arranged so that no single component has to know what the future will be.
← Complex Markets | Start Here
↑ Back to top