A 30-minute weekly review is a decision budget, not a promise that every investigation fits inside 30 minutes. Use the timebox to reconcile the week, inspect coverage, separate process from outcome, choose one supported observation, and version one next action. Put unresolved questions in a deeper-analysis queue instead of answering them with a story.
1. What a Weekly Trading Review Can—and Cannot—Tell You
A daily review can preserve fresh context, while a monthly or quarterly review can compare broader periods. The weekly layer has a different job: combine several sessions under one frozen scope, catch missing data or rule drift early, and carry one well-defined question into the next comparable week.
One week is still a small and path-dependent slice. It can show what happened in the selected accounts and dates. It can show whether a written rule was followed when the necessary evidence exists. It can identify a candidate concentration—for example, that most recorded slippage occurred in one session. It cannot by itself prove that the session caused the slippage, that a strategy has permanently changed, or that next week will behave the same way.
- Reconcile: Can every eligible trade and cash result be traced to its source?
- Describe: What changed inside the frozen weekly scope?
- Test: What one rule or question should the next evidence window evaluate?
This is the compact weekly application of a fuller trade-review method. The weekly habit works when it repeatedly produces traceable decisions—not because 30 minutes has a special statistical property.
2. Freeze the Review Scope Before Opening the Metrics
Start by writing the review contract. If the scope changes while you inspect the results, every comparison becomes negotiable after the outcome is known.
| Scope field | Record before analysis | Failure it prevents |
|---|---|---|
| Window | Exact start/end timestamps and reporting timezone | A session or overnight trade moving between weeks |
| Accounts | Included and excluded account IDs; live, sim and prop separated | Combining unlike risk contracts or duplicate feeds |
| Population | Closed trades, open positions and cancelled orders handled explicitly | A changing denominator or hidden open risk |
| Economics | Gross result, commissions, fees, financing and adjustments available | Calling gross P&L the account result |
| Taxonomy | Setup, session, instrument and rule versions frozen | Retagging losers after the result |
| Evidence state | Last import, duplicates, corrections, missing fields and excluded rows | Treating incomplete coverage as a performance conclusion |
| Baseline | Named compatible period or strategy version | Comparing this week with a convenient but incompatible average |
If the source is incomplete, the review output is an availability report: what is present, what is missing, and what cannot yet be concluded. That is useful work. Filling the gaps with memory so the dashboard looks complete is not.
3. The 30-Minute Evidence Budget
The schedule below is a default operating template. Extend or split it when reconciliation fails, a material risk issue appears, or the week contains a lifecycle event the template cannot resolve. Keep the routine review short by moving the investigation—not the evidence standard—to a separate queue.
| Block | Default time | Question | Saved output |
|---|---|---|---|
| 1. Reconcile | 3 minutes | Is the selected week complete enough to review? | Scope, freshness, duplicates, missingness and exclusions |
| 2. Read the week | 5 minutes | What did the compatible metrics report? | Cost-complete summary plus denominators |
| 3. Inspect decisions | 10 minutes | Which trades require evidence-level review? | Rule exceptions, outliers and representative controls |
| 4. Test one pattern | 7 minutes | Does one candidate survive a baseline check? | Observation, alternatives, confidence and missing fields |
| 5. Version next week | 5 minutes | What one reversible action follows? | Rule version, effective date, test window and rollback |
Do not spend the decision block only on the three biggest winners and losers. Monetary extremes can reveal tail risk, but they are not a representative sample. Review every rule violation and data exception; then inspect material outliers plus one or more ordinary compliant trades as controls. The goal is to understand the record, not to produce a dramatic highlight reel.
4. Read Metrics With Their Denominators
A weekly summary should answer three questions for every number: what population generated it, what evidence is missing, and which costs are included. The companion performance-analysis guide explains the broader measurement contract; this weekly pass keeps only the fields needed for a repeatable decision.
- Net result: realized gross P&L minus the included trading costs, with deposits and withdrawals excluded from trading performance.
- Trade count: the number of eligible logical trades under one stable grouping policy; also retain the underlying execution count.
- Win rate: winning eligible closed trades divided by eligible closed trades. Show the denominator and excluded breakeven/unknown outcomes.
- Average win and loss: conditional means over clearly labeled winning and losing sets. Do not infer a complete edge from the ratio alone.
- Expectancy: a cost-complete descriptive estimate under one declared outcome and grouping policy, accompanied by count, dispersion and missingness.
- Drawdown and exposure: use the account equity convention and timestamps defined by the report; do not substitute the largest closed loss for drawdown.
- Plan adherence: followed observable applicable rules divided by observable applicable rules. Unknown evidence is not compliance and non-applicable rules are not violations.
Do not use universal alarms such as “win rate fell ten points” or “trade count rose 50%” without a compatible baseline and an action rule written in advance. A change may reflect a different setup mix, fewer opportunities, one outlier, a taxonomy edit, missing rows or ordinary variation. Use the trading-statistics framework when the weekly observation needs uncertainty, segmentation or later validation.
5. Separate Process Quality From Trade Outcome
Outcome and execution answer different questions. A winning trade does not prove the decision followed the plan, and a losing trade does not prove it violated the plan. Classify only what the stored evidence supports.
| Recorded outcome | Observable plan state | Weekly interpretation | Next step |
|---|---|---|---|
| Positive | Followed | Compliant positive outcome in this window | Retain; compare with later compatible examples |
| Positive | Violated | Positive outcome does not authorize the breach | Apply the prewritten consequence; preserve the result |
| Negative | Followed | Compliant negative outcome; edge remains a separate question | Retain; escalate only under the strategy-review rule |
| Negative | Violated | Both outcome and recorded execution require review | Address the observable breach without inventing motive |
| Any | Unknown | Plan compliance cannot be scored from available evidence | Repair capture or leave the state unknown |
Avoid labels such as “lucky,” “disciplined,” “revenge trade” or “emotional sizing” unless the decision-time record establishes the relevant facts and the label has an operational definition. The weekly review should name the observable event: size exceeded the active limit, the stop was changed after entry, a required note is absent, or the rule was not applicable.
6. Turn One Observation Into a Reversible Test
The one-observation rule is valuable when it reduces simultaneous changes, not when it promotes the loudest correlation into a permanent rule. Use this sequence:
- State the observation. Name scope, metric, denominator and eligible records.
- Open the examples. Check raw executions, notes, screenshots, rule state and exclusions.
- Compare a baseline. Use the same strategy, taxonomy, costs and market/session definition.
- List alternatives. Data gaps, setup mix, one outlier and execution differences can compete with the first explanation.
- Choose the action class. Keep, monitor, repair data, investigate deeply, or test one plan change.
- Version the change. Record its effective date, eligible population, evaluation window and rollback trigger before the next outcome.
“No late-session entries under plan version 4 until the next review” can be a test if the active plan defines late session, exceptions and rollback. “Be more disciplined” is not testable. Keep the authority in a versioned trading plan so old trades remain attached to the rule that actually governed them.
7. How TSB Makes the Weekly Review Traceable
TSB turns the routine into an evidence chain rather than a motivational recap. Journal holds the exact account and source, executions, fees, tags, notes, screenshots, corrections and exclusions. Reports and deterministic analytics calculate the selected period from compatible stored outcomes and preserve availability. Trading Plan stores the authority and version that existed before the trade.
Coach is the high-leverage review layer over that evidence. Ask it to explain a weekly change, compare a rule with the selected executions, locate concentration by setup/session/source, challenge a causal story, or turn one supported observation into a bounded next test. It can refuse a personalized conclusion when the selected evidence is insufficient, and it does not need to invent motive, silently recalculate an official metric or promise improvement to be powerful. The result stays inspectable because the answer remains tied to the chosen scope and evidence.
TSB has processed 600K+ imported trades across its import history, and the canonical registry recognizes 328 exact broker, exchange, platform and prop-export profiles. Those are product-scale facts—not user counts, proof of profitability, a minimum sample requirement, or trades analyzed by Coach.
Reconcile this week Open the weekly report Review the evidence with Coach
8. Feed the Monthly and Quarterly Review
Save the same compact record every week: scope, availability, cost-complete summary, plan version, one observation, alternatives considered, action, effective date and unresolved queue. Do not infer that four weeks or a fixed number of trades automatically establishes a trend. The required evidence depends on the decision, frequency, variability, dependence and cost of being wrong.
The monthly review can then compare compatible weekly records, identify repeated data failures or rule exceptions, and decide which questions deserve strategy-level analysis. Quarterly review should examine version changes and out-of-sample behavior rather than simply average unlike weeks.
Stop the 30-minute routine and create a separate investigation when positions or account state do not reconcile, a material rule or risk breach is open, a broker correction changes prior results, taxonomy drift invalidates the comparison, or the proposed change would alter strategy, size or risk authority.
Quick Reference: Weekly Review Checklist
- Freeze dates, timezone, accounts, eligible population, cost policy and baseline.
- Confirm import freshness, duplicates, corrections, exclusions and missing fields.
- Read net result, count, win/loss distributions, exposure and adherence with denominators.
- Inspect all violations and exceptions, material outliers and ordinary compliant controls.
- Write one bounded observation, competing explanations and confidence.
- Choose keep, monitor, repair, investigate or one versioned reversible test.
- Save the unresolved queue and the exact evidence needed next.
The Bottom Line
The 30-minute habit works when it protects cadence without lowering the evidence standard. Reconcile first, read every metric with its denominator, inspect decisions rather than outcome theater, and promote only one supported observation into a reversible next step. If the week cannot answer the question, save the missingness and escalate the investigation. In TSB, Journal, Reports, Trading Plan and Coach keep that chain connected from source record to next-week test.
Disclosure: Trader’s Second Brain is our product. Its Journal, report, deterministic-analytics, Trading Plan, Coach evidence gates and canonical public-truth values were checked against the local codebase on September 10, 2026. The 30-minute allocation is an editorial workflow template, not a universal threshold or performance claim. This article is educational and does not provide individualized investment, tax, legal or mental-health advice. See our editorial methodology.
