Risk control sits at the centre of the trading system. The objective is not simply to generate returns, but to do so within a structure that limits downside, caps exposure, reduces emotion, and removes manual interference wherever possible.
The system is built around controlled compounding. That means accepting that not every trade will win, but making sure that losses remain defined, gains are protected where possible, and the overall account curve can grow without relying on oversized risk.
Account Structure
The trading setup uses separate accounts for long and short exposure. This provides a clearer structure for execution, monitoring, and risk control.
Returns are assessed on the combined balance across the active accounts, rather than viewing either side in isolation. This gives a more accurate view of overall system performance, because the long and short accounts are intended to operate as parts of the same strategy framework.
Following the Drift Protocol security incident in April 2026, trading was paused and the execution infrastructure was rebuilt. The event reinforced an important risk-management point: strategy risk and platform risk are separate.
A trading system can control entries, exits, stop losses, and exposure, but it must also account for venue, custody, and operational risk. As a result, the system has moved to a new execution setup and the capital allocation process is being treated more cautiously.
The goal is to avoid over-concentration on any one venue and to place greater emphasis on operational resilience, not just trading fees. Low fees are useful, but they are not the whole risk equation.
Exposure and Leverage
In the earlier phase of development, the system operated with materially higher exposure. At that time, it was effectively deploying the full account size through the split-account structure by using higher leverage.
Losses were still capped with predefined stop-loss levels, but week-to-week volatility was significantly higher.
That early period was useful because it made the trade-off very clear: higher leverage can accelerate gains, but it also increases volatility and puts more pressure on decision-making.
As the system matured, exposure was deliberately reduced. The current approach is designed around lower leverage, smaller position allocations during testing, and a more selective signal set.
Weekly Return Profile
The weekly return graph below reflects the system’s development over time. Earlier results were more volatile because exposure was higher and the signal set was still being refined. Over time, the return profile became smoother as leverage was reduced, signal selection improved, and manual interference was removed.
The reduction in volatility is intentional. The aim is not to maximise every short-term move, but to build a return profile that can compound more consistently over time.
The current maximum drawdown is:
Max Drawdown: -10.73%This figure is important because it shows the largest peak-to-trough decline recorded during the tracked period. Return is only half the story; the quality of a trading system also depends on how much drawdown it experiences while producing those returns.
Stop Losses and Profit Protection
Losses are controlled in several layers. Each strategy operates with a predefined stop loss, set before entry and not widened once a trade is live.
The system also uses staged trailing profit protection. This allows profitable trades to lock in gains once certain profit levels are reached, while still leaving room for larger moves to develop.
This creates an asymmetric structure: losses are capped, while profitable trades can continue if the market keeps moving in the right direction.
Protective Execution
Execution risk is managed directly. Protective stop logic is placed on the exchange where supported, while the bot also monitors positions locally and can send an emergency reduce-only market order if required.
This belt-and-braces approach is designed to reduce dependence on any single mechanism. If the local bot is running, it can actively manage the position. If the local machine or connection fails, the exchange-side protective stop provides an additional layer of protection.
No execution setup can remove risk completely, especially during fast market moves, but the goal is to reduce avoidable failure points wherever possible.
Circuit Breakers
Each strategy has its own circuit breaker. If a strategy reaches its defined loss threshold, it is automatically disabled until manually reviewed.
Above that, the system also has a master circuit breaker, which can stop trading entirely if losses accumulate across strategies beyond an acceptable level.
These controls are designed so that no single trade, and no short run of poor trades, can cause disproportionate damage to the account. The purpose of the automation is not only to execute trades, but to enforce discipline under pressure.
Strategy Selection
Another important part of the process is selectivity. The current focus is on higher-confidence strategies that trade less frequently but with tighter statistical support.
Every live strategy must first be assessed across a meaningful historical period and through changing market conditions. The aim is quality over quantity: fewer trades, better filters, tighter control.
Strategies can be added, reduced, or removed based on evidence from ongoing analysis. The aim is to make those changes systematically, not emotionally.
Treasury and Capital Allocation
Treasury management is also part of the risk framework. Net trading profits are retained within the company treasury and reinvested into trading capital to support controlled compounding over time.
Treasury is held primarily in cash-equivalent form, such as USDC, and is used to support short-duration systematic trading rather than long-term speculative holding.
Following the Drift incident, platform concentration is being treated more carefully. The longer-term plan is to reduce dependence on any single venue and build a more resilient operational structure as the system grows.
Summary
This framework does not eliminate risk. No trading system can do that.
What it does provide is a structured process for controlling exposure, capping losses, protecting gains, managing capital responsibly, reducing platform concentration, and shutting down automatically when conditions move outside acceptable bounds.
The goal is controlled compounding: steady improvement, disciplined execution, and risk management that remains active before, during, and after every trade.