Strategy risk and review: know when the process has changed
Monitor strategy risk through exposure limits, drawdowns, execution quality, model drift, market-regime changes, capacity, and a written process for pausing, changing, or retiring the strategy.
This guide covers:
- Define failure conditions before reviewing performance.
- Why risk controls should be specified as part of the strategy rather than added after losses occur.
- Identify the risk controls that should be in place before capital is committed.
Review these foundations before moving into the details.
Define failure conditions before reviewing performance
A strategy can stop behaving as expected because the market, implementation, data, or the strategy itself changed. Review should separate ordinary variance from evidence that the process no longer matches the assumptions under which it was tested.
- Which risk limits are defined before losses occur?
- How will execution slippage, capacity, model drift, and regime change be monitored?
- What evidence justifies pausing, changing, or retiring the strategy?
01SECTION 01 · 2 MINRisk controls belong inside the strategy
Position sizing, diversification, liquidity limits, leverage limits, stop-order assumptions, and account permissions should be part of the strategy specification rather than added after a loss. Automated execution can follow instructions perfectly and still lose money if the instructions are wrong or the market changes.
Risk controls belong inside the strategy
Position sizing, diversification, liquidity limits, leverage limits, stop-order assumptions, and account permissions should be part of the strategy specification rather than added after a loss. Automated execution can follow instructions perfectly and still lose money if the instructions are wrong or the market changes.
How the mechanism works
Before capital is committed
Know the planned loss and a worse fill if the market gaps.
Know the average volume, spread, and whether the position can be reduced without forcing the price.
Stress-test how financing and volatility change the account equity.
Write the evidence that pauses or retires the strategy instead of endlessly changing parameters.
02SECTION 02 · 2 MINUse the same review sequence every time
Check scheduled events, liquidity, account buying power, open risk, and strategy eligibility.
Use the same review sequence every time
Check scheduled events, liquidity, account buying power, open risk, and strategy eligibility.
How the mechanism works
Before the session
Check scheduled events, liquidity, account buying power, open risk, and strategy eligibility.
Before the order
Record the signal, entry rule, invalidation point, size, order type, and maximum acceptable slippage.
After the trade
Record actual fill, cost, rule compliance, and whether the result came from the intended process or an unplanned decision.
At the review date
Evaluate a series of trades, not one outcome. Look for changing costs, signal decay, concentration, and behavioral overrides.
03SECTION 03 · 2 MINA strategy becomes improvable when every decision leaves a record
A research journal should capture the hypothesis, data version, universe, rule, parameter choice, expected edge, risk, planned execution, and reason for every live position. A post-trade review then compares the actual decision with the written process without changing the original assumptions in hindsight.
A strategy becomes improvable when every decision leaves a record
A research journal should capture the hypothesis, data version, universe, rule, parameter choice, expected edge, risk, planned execution, and reason for every live position. A post-trade review then compares the actual decision with the written process without changing the original assumptions in hindsight.
How the mechanism works
Before
Thesis, setup, entry rule, invalidation, expected holding period, size, maximum loss, event risk, liquidity, and expected costs.
Before the trade, save the signal, thesis, entry rule, size, expected cost, invalidation condition, and maximum loss so the later review has a fixed reference point.
During
Execution price, slippage, partial fills, changes in volatility/liquidity, material new information, and whether any rule was overridden.
During the position, record executions, slippage, rule changes, unusual liquidity, and whether exposure still matches the tested system instead of improvising around short-term P&L.
After
Process quality, return, maximum favorable/adverse excursion, costs, whether the hypothesis was right for the right reason, and one rule to keep or change.
After exit, compare the realized path with the original rules, separate market outcome from process quality, and update the system only after enough evidence accumulates.
Separate a losing trade from a broken process
A strategy can lose money even when it was followed correctly, and it can show a gain despite a poor decision. Review process quality separately from outcome so one lucky or unlucky result does not rewrite the system.
| Review layer | Question |
|---|---|
| Rule adherence | Was the setup, sizing, entry, exit, and risk rule followed as written? |
| Data quality | Was the decision based on timely, reliable information and correct calculations? |
| Execution | Did spread, slippage, liquidity, order choice, financing, or assignment change the result? |
| Environment | Did volatility, correlation, market structure, or another regime input move outside the assumptions? |
| Strategy evidence | Has the long-run expected relationship weakened, or is the latest outcome within normal uncertainty? |
Review the strategy when the process changes, not only when performance disappoints
A strategy can underperform while still operating as designed, or it can appear profitable while the original edge has disappeared. Define the expected environment, turnover, drawdown range, capacity, liquidity, and implementation assumptions before deployment.
For option-based or leveraged strategies, add the specific rights and obligations created by the instrument. Expiration, assignment, margin, and path dependency can change risk faster than a simple historical return series suggests. Strategy review should test whether the original mechanism is still present and executable.
- Separate process failure from ordinary return variability.
- Track implementation costs and liquidity as part of the strategy, not as an afterthought.
- Pause or resize when the environment violates assumptions that were essential to the original edge.
Review the key points
1. What failure conditions should be defined before performance is reviewed?
A strategy can stop behaving as expected because the market, implementation, data, or the strategy itself changed. Review should separate ordinary variance from evidence that the process no longer matches the assumptions under which it was tested.
2. Why risk controls should be specified as part of the strategy rather than added after losses occur.
Position sizing, diversification, liquidity limits, leverage limits, stop-order assumptions, and account permissions should be part of the strategy specification rather than added after a loss. Automated execution can follow instructions perfectly and still lose money if the instructions are wrong or the market changes.
3. Which risk controls should be in place before capital is committed?
Know the planned loss and a worse fill if the market gaps. Know the average volume, spread, and whether the position can be reduced without forcing the price. Stress-test how financing and volatility change the account equity.