The previous version treated sector as universally the most valuable field, described catalyst labels as causes, suggested that a fixed 30–50-trade sample would reveal an edge, recommended increasing size from simple win-rate slices, and hard-coded a broad short-borrow range. Those claims are corrected below. A stock journal can expose a decision worth testing; it cannot prove causation or prescribe size without a defined population, complete costs, uncertainty, and risk controls.
What Actually Matters in a Stock Trading Journal
A useful stock trading journal must preserve the execution ledger and the market context that can change an equity trade: security identity, venue, session, catalyst hypothesis, earnings timing, corporate actions, borrow evidence, costs, and overlapping portfolio exposure. Record what was known before entry separately from what was learned after the outcome. Then compare like with like: the same setup version, direction, holding style, cost treatment, and market window.
| Layer | Minimum record | What it prevents |
|---|---|---|
| Source | Account alias, broker or custodian, venue, source file/API route, import batch | Mixing accounts, currencies, or incomplete histories |
| Execution | Order and fill IDs, side, quantity, timestamps, prices, status, partials | Turning several fills into an unreconciled “trade” |
| Economics | Gross result, commissions, regulatory and venue fees, borrow or locate charges, net result | Calling a gross winner a net winner |
| Decision | Setup version, thesis, trigger, invalidation, planned size and initial risk | Rewriting the plan after seeing the result |
| Stock context | Ticker/security ID, sector taxonomy, event source and timestamp, benchmark, session, corporate-action state | Treating labels or adjusted charts as source facts |
| Portfolio context | Concurrent positions, sector/factor overlap, open risk and account equity timestamp | Reviewing correlated bets as independent trades |
Stock Trading Journal Fields: The Complete Schema
Do not start by collecting every tag a template offers. Start with the fields required to reproduce the trade and answer the next review question. Missing source values stay Not verified; they are not reconstructed from a chart or memory.
| Group | Fields to store | Evidence rule |
|---|---|---|
| Security identity | Raw symbol, stable security identifier when supplied, company name, listing venue, quote currency | Keep the raw symbol; version symbol changes and mappings |
| Time | Order, fill, entry and exit timestamps; source time zone; normalized UTC; regular or extended session | Never infer session from a date without a timestamp and venue rule |
| Order and fills | Order ID, fill IDs, requested and filled quantity, side, order type, fill price, cancellations and rejections | Preserve every fill beneath the grouped review trade |
| Plan | Setup and version, thesis, trigger, invalidation, planned stop, target, size and initial planned risk | Timestamp before or at entry; retrospective fields must be labeled |
| Economics | Gross P&L, commissions, fees, borrow/locate charges, rebates, dividends or adjustments, net P&L | Record value and currency; reconcile components to the account statement |
| Company/event | Event type, source URL or filing ID, announced timestamp, earnings period, guidance state, before/after-event flag | Store the event as context or hypothesis, not proof that it caused the move |
| Market context | Sector classification and version, chosen index/sector benchmark, benchmark window, volatility/liquidity observation | Choose the benchmark before comparing results |
| Review | Plan adherence, execution exception, screenshot, source evidence, lesson, next prospective test | Keep interpretation separate from imported facts |
Day and swing trades can live in one database, but they should not share one undifferentiated statistic. Use the day-trading journal field set for session and execution questions, and the swing-trading journal field set for overnight, thesis, and holding-period questions.
How to Journal Catalysts and Earnings Without Inventing Causation
“Earnings,” “news,” “breakout,” and “sector rotation” are useful retrieval labels, but a label does not prove why price moved. A clean record separates four things: the event that was publicly available, the trader’s pre-trade interpretation, the execution, and the later outcome.
- Identify the source. Save the issuer filing, earnings release, exchange notice, or other named source and its publication timestamp. The SEC’s EDGAR search provides public access to company filings; a headline copied into a note is not a substitute for the filing.
- Freeze the hypothesis. State what the event was expected to change, the trigger that would confirm the setup, and the condition that would invalidate it.
- Mark event timing. Record whether entry occurred before the announcement, during extended hours, at the next regular-session open, or after the initial reaction.
- Store comparable fields. For earnings, keep the reporting period, actual versus the contemporaneous estimate source, revenue, guidance state, and any options-implied-move input only when its source and timestamp are known.
- Review competing explanations. Benchmark movement, sector news, liquidity, and a broader market event may overlap. Describe the association; do not upgrade it to causation from one trade.
A fixed observation count cannot guarantee that an earnings pattern is stable. Report the number of eligible signals, taken trades, exclusions, direction, setup version, event window, costs, and uncertainty. A small cohort can generate a hypothesis; position size should follow the trader’s validated risk policy, not whichever slice has the highest recent win rate.
Overnight Gaps, Extended Hours, and Corporate Actions
The SEC’s investor bulletin notes that extended-hours venues may have lower liquidity, wider spreads, uncertain prices, different order handling, and unlinked markets. Your journal therefore needs the venue and session attached to each fill. “After hours” is not a sufficient execution record.
Define an overnight gap before calculating it. One reproducible choice is the percentage change from the previous regular-session close to the next regular-session open on the same declared price series. Keep raw and adjusted prices separate, name the data vendor, and record the calendar and time zone. If your strategy enters or exits outside regular hours, add those fills rather than pretending the regular open captured the execution.
Corporate actions can change symbol, share count, price scale, cash flows, or ownership history. FINRA lists dividends, stock splits, mergers, and symbol changes among common corporate actions. Preserve the original fills and the adjustment factor or cash event as a separate record. A split-adjusted chart should not silently rewrite the quantity and price that appeared on the broker statement.
How to Track Short Selling
A short record needs more than the direction field. Store whether a locate was required, requested, accepted, rejected, or unavailable; the approved quantity and time; the security’s hard-to-borrow state as supplied; recall or buy-in notices; daily borrow postings; locate charges; and the final net result. Regulation SHO’s locate framework applies before effecting many short sales, but the trader’s actual broker record is the evidence for a specific order.
Do not hard-code a universal borrow-rate range. Rates, locate charges, availability, and broker treatment can change by security and day. If the source provides an annualized rate, preserve the rate, date, notional basis, day-count method, currency, and actual posted charge. If only a posted fee exists, record the fee without inventing a rate. If neither exists, show Not verified.
Net short P&L should reconcile the sale and cover fills with commissions, regulatory/venue fees, borrow or locate charges, dividends or other account adjustments, and currency conversion where applicable. The trading-cost reconciliation guide shows how to keep gross result and each cost component separate.
Sector Exposure and Portfolio Overlap
Sector is valuable when it answers a defined question, not because it is always the “most important” stock field. Version the classification source because companies and taxonomies change. Also record the benchmark used for the comparison and the exact window; SPY direction alone does not explain a stock’s outcome.
A trade-by-trade journal can hide account risk when several positions respond to the same sector, factor, index, event, or issuer relationship. At each decision timestamp, store concurrent positions, long/short direction, market value or another consistent exposure unit, planned open risk, and account equity timestamp. Then flag concentration as a review lead.
Correlation is sample- and window-dependent. It does not prove that two positions are the same bet, and historical correlation can change. The portfolio correlation and overlap workflow explains how to declare the return series, window, missing-data policy, and exposure units before acting.
How to Analyze Stock Trading Journal Data
First reconcile the population: source trade count, fills, quantities, timestamps, gross result, costs, and net result. Then freeze one question. Every breakdown should show its observation count and exclusions beside the result.
| Analysis | Question | Controls | Safe next step |
|---|---|---|---|
| Net expectancy by setup | Did comparable planned trades have different average net outcomes? | Setup version, planned risk, costs, complete eligible population | Create a prospective test; do not resize from one retrospective slice |
| Long versus short | Did direction cohorts differ after all short-specific costs? | Market regime, setup mix, borrow/locate evidence, sample size | Inspect cost and setup composition before blaming direction |
| Sector or benchmark-relative | Did results differ under one stable classification and benchmark rule? | Taxonomy version, benchmark, return window, concurrent exposure | Test the same rule prospectively across new observations |
| Event window | How did a frozen setup behave before or after one event definition? | Source timestamp, announcement/session rule, eligible signals, exclusions | Treat the result as association, not event causation |
| Execution by session | Did fill quality or plan adherence differ in regular and extended hours? | Venue, UTC timestamps, order type, reference price, liquidity state | Review order handling and slippage definition |
| Overnight holdings | What part of net result and risk arrived outside regular hours? | Gap definition, raw/adjusted series, corporate actions, actual fills | Separate overnight policy from intraday setup changes |
A winning slice is not automatically an edge, and a losing slice is not automatically a setup to delete. Check selection bias, multiple comparisons, changing conditions, outliers, and uncertainty. The trading-performance analysis workflow provides the complete sequence from population definition to a versioned decision and rollback rule.
A Stock Journal Workflow That Survives Audit
- Capture the plan before outcome. Save the setup version, event hypothesis, trigger, invalidation, planned size, initial risk, and intended holding style.
- Import or enter source facts. Preserve order and fill IDs, raw symbols, timestamps, quantities, prices, and account identity.
- Attach stock-specific evidence. Add the filing/event source, session, benchmark, corporate-action state, borrow/locate record, and concurrent exposure.
- Reconcile. Match fills and quantities, then reconcile gross P&L and every posted cost to net account P&L.
- Review one declared cohort. Freeze the question, population, metric, exclusions, setup version, and sample count before looking at the answer.
- Change one rule prospectively. Record the change, evaluation window, risk cap, success criterion, and rollback condition.
Where TSB Fits in a Stock Trading Journal
Trader’s Second Brain is our product. Its useful role here is to normalize supported source records, keep Journal evidence beside review fields, expose analytics and Leak Map slices, and let AI Coach discuss the selected stored evidence. It does not supply missing broker facts, prove why a stock moved, or make a strategy profitable.
TSB has processed 600K+ imported trades across the product and recognizes 331 structured source profiles through canonical server truth. The first number is product-scale throughput, not the sample behind this article or any stock-sector conclusion. The second counts parsers and mappings, not guaranteed compatibility with every custom export. A specific analysis must display its own complete observation count and percentages must use that same denominator.
For a stock workflow, test a representative file containing partial fills, fees, short activity, symbol changes, and multiple accounts. Reconcile it against the source before trusting charts. If you want formula-level control or a custom research model, Excel can remain the better working surface; the server-rendered comparison below keeps current offering boundaries separate from this editorial verdict.
Test the source before moving the history
Check the exact broker or file route, preview a representative sample, and verify counts, quantities, timestamps, costs, and net result.
Check the import route See the evidence workflowTSB does not guarantee every custom export, replace the broker or issuer source, reconstruct missing events, or validate a trading edge.
Sources and Verification
Market-structure statements were checked September 11, 2026 against the SEC/Investor.gov extended-hours bulletin, the SEC EDGAR search, FINRA’s corporate-actions overview, and FINRA’s Regulation SHO locate summary. Product-scale counts and provider-card values render from canonical TSB server truth; the production database remained read-only during this rewrite.
